Pediatric epilepsy surgery : preoperative assessment and surgical treatment [Second ed.] 9781626238169, 1626238162

Table of contents :
Pediatric Epilepsy Surgery: Preoperative Assessment and Surgical Treatment 2nd Edition
Title Page
Copyright
Dedication
Contents
Videos
Foreword
Preface
Acknowledgments
Contributors
Chapter 1 Basic Considerations of Pediatric Epilepsy Surgery
Introduction
Historical Evolution
Epilepsy Surgery in Children
Unique Characteristics of and Special Considerations for Children
Age Group
Epileptic Syndromes and Pathologies Specific to Children
Semiology and Electrophysiological Characteristics
The Effect of Seizures on the Developing Brain
Functional Plasticity
Medical Intractability
Timing of Surgery
Goals of Surgery
Surgical Procedures
Present Status and Future Considerations
Conclusion
Chapter 2 A Historical Review of Epilepsy Surgery and Its Application in Children
Introduction
The Beginning: England
Germany: Cortical Stimulation, Mapping and Electroencephalography
The Montreal Neurological Institute: The Beginning of a New Era
The Quest for a Surgical Treatment for Temporal Lobe Epilepsy
Exploring the Epileptogenic Network: Stereoencephalography
Recording Seizures: Long-term Electroencephalography/Video Monitoring
New Frontier: Pediatric Epilepsy Surgery
Pediatric Guidelines, Task Forces, and Meetings
Conclusion
Chapter 3 Epidemiology of Intractable Epilepsy in Children
Introduction
Definitions of Intractable Epilepsy
Intractability that Comes and Goes
Predicting Intractable Epilepsy
Epilepsy Syndromes
Mortality
Conclusion
Chapter 4 Genetics in Epilepsy Surgery
Introduction
Neurobiological Pathways in Genetic Epilepsy
Genetic Principles in Nonstructural and Structural Epilepsies
Nonstructural Epilepsies
Structural Epilepsies
Challenges of Genetics in Epilepsy Surgery
Chapter 5 Surgical Neuropathology of Pediatric Epilepsy
Introduction
Malformations of Cortical Development
Malformations of Cortical Development Due to Abnormal Neuronal and Glial Proliferation or Differentiation
Malformations of Cortical Development Due to Abnormal Neuronal Migration
Malformations of Cortical Development Due to Abnormal Cortical Organization
Neoplastic Lesions
Glioneuronal Tumors
Glial Tumors
Hamartomas
Glioneuronal and Glial Hamartomas
Hypothalamic Hamartomas
Meningioangiomatosis
Hippocampal (Mesial Temporal) Sclerosis
Vascular Malformations
Arteriovenous and Cavernous Vascular Malformations
Sturge–Weber Syndrome
Acquired Lesions
Vascular and Traumatic Lesions
Inflammatory Lesions
Chapter 6 Epilepsy and Brain Plasticity
Introduction: Seizure Susceptibility and Transient Plasticity in the Developing Brain
Etiology of Seizure Onset and Long-term Plasticity
Seizures Beget Seizures?
Neurogenesis in Epilepsy
Cell Death and Axonal Sprouting
Inflammation and Epilepsy
Lesional Epilepsy and Mirror Focus Development
Possible Clinical Implications
Conclusion
Chapter 7 Effects of Seizures and Their Comorbidities on the Developing Brain
Comorbidities in Epilepsy
Cognitive Deficits
Psychiatric Comorbidities
Effects of Early-Life Seizures
The Impact of Treatment
Conclusion
Chapter 8 Ethical Considerations in Pediatric Epilepsy Surgery
Introduction
Ethical Issues Surrounding Access to Epilepsy Surgery
Ethical Issues in the Presurgical Evaluation of Children with Intractable Epilepsy
Informed Consent for Epilepsy Surgery
Informed Consent in Children
Informed Consent for Research and the Therapeutic Misconception
Ethical Issues in Operative Decision-making
Surgery in Eloquent Cortex
Clinical Translation of Surgical Innovations
Conclusion
Chapter 9 Infantile and Childhood-Onset Catastrophic Epilepsy Syndromes
Introduction
Catastrophic Epilepsies with an Onset During Infancy and Childhood
Neonatal Period (~1 Month of Age)
Infant Period (Aged 1–24 Months)
Childhood Period (Aged 24+ Months)
Chapter 10 Epilepsy Surgery for Congenital or Early Brain Lesions
Introduction
Types of Congenital or EarlyAcquired Lesions Causing Epilepsy
Hemispheric Lesions
Focal Lesions
Nonlesional MRI
Impact of Early Lesions
Early Lesions: Role of Adaptive Plasticity
Early Lesions: Electroencephalography Manifestations
Timing of Surgery
Selection of Surgical Candidates
Types of Surgeries for Early Lesions
Outcome
Conclusion
Chapter 11 Intractable Epilepsy in Children and Selection of Surgical Candidates
Introduction
When to Refer Children for Epilepsy Surgery?
Process of Evaluation for Selection of Surgical Candidates
The Presurgical Clinical Evaluation
Seizure Semiology and Neurological Examination
Physical Examination
Presurgical Evaluation Techniques and Modalities
Noninvasive Electrophysiology Monitoring
Structural Neuroimaging
Functional Neuroimaging
Neuropsychological Evaluation
Epilepsy Surgery Conference and Invasive Monitoring
Surgical Treatment and Outcome
Postsurgical Follow-up
Chapter 12 Clinical Semiology in Preoperative Assessment
Introduction
Localization-Dependent Seizure Semiology
Frontal Lobe Seizures
Temporal Lobe Seizures
Posterior Cortex Seizures
Seizures with Insular Origin
Seizure Semiology of Multilobar and Hemispheric Epilepsies
Epileptic Spasms: When Age Overcomes Localization
Does Etiology Play a Role in Childhood Seizure Semiology?
Peri-ictal Lateralizing Signs in Children
Semiology Pitfalls in Childhood
Nonepileptic Events Mimicking Childhood Epilepsy Semiology
Focal Semiology in Nonsurgical Childhood Epilepsy Syndromes
Extratemporal Epilepsies with Temporal Semiology
Temporal Lobe Epilepsies with Extratemporal Semiology
Inconcordant/Incongruent Data During Presurgical Assessment
Chapter 13 Preoperative Neuropsychological and Cognitive Assessment
Introduction
Neuropsychological Assessment of Children
Standardized Psychological Tests
Domains of Neuropsychological Assessment
General Cognitive Ability
Language
Memory
Visual–Spatial
Executive Functions
Motor and Function
Psychosocial Adjustment
Academic Skills
Adaptive Function
Epilepsy in Children and Its Effect on Neuropsychological Function
Goals of Neuropsychological Assessment in Pediatric Epilepsy Surgery
Special Considerations Regarding Neuropsychological Assessment of Children with Epilepsy
Neuropsychological Data for Localizing and Lateralizing the Epileptogenic Area
Language
Memory
Visual–Spatial
Executive Functions
Special Tests for Localization and Lateralization of Cognitive Function
Advice to Parents and Teachers Regarding Potential Postoperative Issues
Follow-up Assessment and Neuropsychological Outcome
Conclusion
Acknowledgment
Chapter 14 Electroencephalography and Noninvasive Electrophysiological Assessment
Introduction
Electroencephalography Interpretation
Temporal Lobe Epilepsy
Extratemporal Lobe Epilepsy
Activation Procedures
Functional Imaging and EEG
EEG Monitoring and SPECT Scan
EEG Monitoring and PET Imaging
EEG and Magnetoencephalography
EEG and Functional MRI
Pitfalls of Localization of Epileptiform Activity
Seizure Semiology
EEG and Epileptogenic Lesions
False Lateralization/Localization
Normal EEG
Emerging Noninvasive Neurophysiology Methods
EEG Source Imaging
Infraslow Activity
High-Frequency Oscillations
Functional Near-Infrared Spectroscopy-EEG
Chapter 15 Invasive Electrophysiological Monitoring
Introduction
Pragmatic Considerations
Indications
Inconclusive Preoperative Data
Normal or Nonspecific Structural Neuroimaging Findings
Multiple or Large Structural Lesions
Divergent Preoperative Data
Epileptogenic Region in Proximity to Eloquent Cortex
Technical Aspects
Spatial Coverage
Type of Electrodes
Surgical Insertion
Risks
Recording
Interpretation
Ictal Onset Zone
Irritative Zone
Functional Deficit Zone
Conclusion
Chapter 16 Extraand Intraoperative Electrocortical Stimulation
Introduction
Electrocortical Stimulation Basics
Pediatric Aspects of the Presurgical Evaluation
A Historical Note
Physiology of Electrocortical Stimulation
Extraoperative Electrocortical Stimulation
General Principles and Techniques
Safety Issues and Complications in Pediatric Patients
Special Considerations in Children
The Effect of Pathology on Electrocortical Stimulation Mapping
Electrocortical Stimulation with Depth Electrodes in Children
Electrocortical Stimulation-Induced Responses
Studying More Specific Brain Anatomy with Stimulation
Extraoperative Electrocortical Stimulation Compared with Other Noninvasive Functional Mapping Techniques
Intraoperative Electrocortical Stimulation
Conclusion
Acknowledgment
Chapter 17 Stereoelectroencephalography
Introduction
Historical Evolution of Stereoelectroencephalography
Basic Principles of Stereoelectroencephalography
Anatomo-Electro-Clinical Correlations and Epileptogenic Network Concepts
Propagation Theory
Stereoelectroencephalography Indications
Stereoelectroencephalography Implantation Strategy
Stereoelectroencephalography Recording
Stereoelectroencephalography Data Analysis and Interpretation
Interictal Activities
Ictal Activities
Postictal Activities
Stimulation
Defining the Epileptogenic Zone
Special Considerations for Pediatric Patients
Conclusion
Chapter 18 Magnetoencephalography
Introduction
Basic Principles of Magnetoencephalography and Magnetic Source Imaging
Magnetoencephalography Spike Sources
Lesional Epilepsy
Focal Cortical Dysplasia
Tuberous Sclerosis Complex
Extratemporal Lobe Epilepsy
Insular Epilepsy
Temporal Lobe Epilepsy
Nonlesional Epilepsy
Recurrent or Residual Seizures after Surgery
Functional Mapping
Chapter 19 Structural Brain Imaging in Pediatric Epilepsy
Introduction
Magnetic Resonance Techniques
Epileptogenic Substrates
Low-Grade Epilepsy-Associated Tumors
Common Malformations of Cortical Development
Other Malformations of Cortical Development
Hypothalamic Hamartoma
Hippocampal Sclerosis
Rasmussen’s Encephalitis
Other Causes for Partial or Catastrophic Epilepsies in Children
Conclusion
Chapter 20 Functional Magnetic Resonance Imaging in Pediatric Epilepsy Surgery
Introduction
Instrumentation and Methods
Basic Principles of Functional Magnetic Resonance Imaging
Experimental Designs
Response Monitoring
Assessment of Laterality of fMRI Activation
Applications for Presurgical Evaluation
Preoperative Assessment of Language Lateralization
Assessment of Language Lateralization: fMRI and IAT
Localization of Language Cortex: fMRI and ECS
Preoperative Assessment of Memory Lateralization
Localization of Epileptic Discharge Activity Using fMRI
Conclusion
Chapter 21 Application of Positron Emission Tomography and Single-Photon Emission Computed Tomography in Pediatric Epilepsy Surg
Introduction
Rationale for PET and SPECT
Principles and Techniques of PET
Principles and Techniques of SPECT
The Role of PET and SPECT in Pediatric Epilepsy Surgery
Epileptogenic Region
Functional Status of the Rest of the Brain
Postsurgical Evaluation
Experience with Other PET Tracers in Epilepsy
Multimodality Imaging with PET-MR
Conclusion
Chapter 22 Multimodality Imaging and Coregistration
Introduction
Coregistration
Newer Imaging Technologies
Magnetic Resonance Imaging
Functional Magnetic Resonance Imaging
Diffusion Tensor Imaging
Magnetoencephalography
Intraoperative Imaging
Future Perspectives
Conclusion
Acknowledgment
Chapter 23 Cerebral Cortex: Embryological Development and Topographical Anatomy
Introduction
The Human Cerebral Cortex
Embryology and Cortical Development of the Cerebral Cortex
The Cerebral Sulci and Gyri
Frontal Lobe
Central Lobe
Insula and the Peri-insular Region
Temporal Lobe
Parietal Lobe
Occipital Lobe
Addendum: Topographical Anatomy of the Cerebral Cortex on MRI
Frontal Lobe
Parietal Lobe
Temporal Lobe
Insular Region
Occipital Lobe
Acknowledgment
Chapter 24 Tractographic Anatomy of White Matter
Introduction
Limbic System
Fornix
Mammillothalamic Tract
Cingulum
Thalamic Radiations
Projection System
Thalamocortical Sensory Projections
Extrathalamic Projections
Optic Radiations
Auditory Projections
Corticocerebellar Projections
Descending Motor Projections
Commissural Fibers
Anterior and Posterior Commissures
Corpus Callosum
Association System
Arcuate Fasciculus
Superior Longitudinal Fasciculus
Inferior Fronto-Occipital Fasciculus
Uncinate Fasciculus
Inferior Longitudinal Fasciculus
Middle Longitudinal Fasciculus
Cranial Nerves
Conclusion
Chapter 25 Localization of Motor Cortex and Subcortical Pathways Using Functional Magnetic Resonance Imaging and Diffusion Tenso
Introduction
Functional MRI for Localization of the Motor Cortex
Indications
Blood Oxygen Level Dependent fMRI
Stimulus Presentation
Functional Anatomy of Sensorimotor System
Common Sensorimotor Paradigms
Bilateral Complex Finger Tapping
Passive Movement or Sensory Paradigms
Resting State fMRI
Validation in Literature
Diffusion Tensor Imaging
Rationale for Presurgical White Matter Mapping
Imaging Technique
Basic Anatomy of Corticospinal Tract
Essential Considerations and Limitations with BOLD fMRI and DTI Techniques
Special Considerations and Challenges in Young Children
Conclusion
Chapter 26 The Wada Test: Lateralization of Language and Memory
Introduction
Special Considerations for Pediatric Wada Testing
Wada Language Testing in Pediatrics
Wada Memory Testing in Pediatrics
Conclusion
Chapter 27 Language Lateralization and Localization: Functional Magnetic Resonance Imaging
Introduction
fMRI in Children: Challenges and Limitations
fMRI Language Task Selection
Language Laterality by fMRI
New Applications of fMRI in Language Mapping in Children
Conclusion
Chapter 28 Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia
Introduction
Anatomy and Applied Physiology
Somatosensory-Evoked Potentials
Short-Latency Somatosensory-Evoked Potentials
Localization of the Primary Somatosensory and Motor Cortices with the SSEP Phase Reversal Technique
Motor Evoked Potentials
Direct Cortical and Subcortical Electrical Motor Stimulation
The Multi-Pulse Train High-Frequency Technique: Taniguchi’s Method
Direct Subcortical Stimulation
Transcranial Electrical Stimulation
Physiology, Stimulation, and Recording
Anesthetic Considerations
Applications and Interpretation during Lesion Resection in Proximity to Eloquent Cortex
Navigated Transcranial Magnetic Motor Stimulation
Future Directions
Chapter 29 Cortical Stimulation and Mapping
Introduction
Language Mapping
Indications for Language Mapping
Anesthesia for Language Mapping
Language Mapping Technique
Motor Mapping
Indications and Contraindications for Motor Mapping
Awake versus Asleep Motor Mapping
Anesthesia for Awake Motor Mapping
Anesthesia for Asleep Motor Mapping
Importance of Temperature and Other Physiological Parameters
Motor Mapping Technique
Complications Avoidance
Seizures
Registration
Patient Intolerance
Anesthesia
Conclusion
Chapter 30 Subcortical Mapping During Intracranial Surgery in Children
Introduction
Tractography
Direct Cortical Stimulation: Traditional Penfield’s Technique
Short Train Technique and Monitoring of Motor Evoked Potentials
Subcortical Stimulation
Cortico-Cortical Evoked Potentials
Conclusion
Chapter 31 Anesthetic Considerations and Postoperative Intensive Care Unit Care in Pediatric Epilepsy Surgery
Introduction
Physiological Differences in Pediatrics
Preoperative Evaluation and Preparation
Anesthetic Management
Premedication
Induction of Anesthesia
Airway Management
Positioning
Vascular Access
Maintenance of Anesthesia
Intraoperative Fluid and Electrolyte Management
Anesthetic Considerations for Specific Procedures
Anesthesia for Placement of Grids and Strips
Stereotactic-Guided Ablations
Resection of Seizure Focus
Awake Craniotomy
Corpus Callosotomy
Hemispherectomy
Vagal Nerve Stimulator
Postoperative Management
Conclusion
Chapter 32 Pediatric Awake Craniotomy
Introduction
Patient Selection and Evaluation
Anesthetic Evaluation
Neuropsychological Evaluation
Psychological Preparation
Intraoperative Technique
Operating Room Setup
Anesthetic Technique
Induction and Initial Asleep Phase
Regional Anesthesia for Awake Craniotomy
Maintenance of Anesthesia
Electroencephalography for Depth of Anesthesia and Detection of Afterdischarges
Awake Phase
Intraoperative Complications
Conclusion
Chapter 33 Implantation of Strip, Grid, and Depth Electrodes for Invasive Electrophysiological Monitoring
Introduction
Indications
Invasive Monitoring Techniques
Subdural Strip and Grid Electrodes
Depth Electrodes
Advantages and Limitations of Invasive Monitoring Techniques
Surgical Technique
Subdural Strip Electrode Placement
Subdural Grid Electrode Placement
Intraparenchymal Depth Electrode Placement
Postoperative Management
Complications
Conclusion
Chapter 34 Stereoelectroencephalography in Children: Methodology and Surgical Technique
Introduction
History and Basic Principles of the Stereoelectroencephalography Methodology
Choosing SEEG as the Appropriate Method for Extraoperative Invasive Monitoring
How to Select SEEG Trajectories: Planning the Implantation?
Limbic Network Explorations
Frontal–Parietal Network Explorations
Posterior Quadrant Network Explorations
The “Nuts and Bolts” of the SEEG Implantation Technique in Children
Morbidity and Seizure Outcome in the Pediatric Population
Conclusion
Chapter 35 Mesial Temporal Sclerosis in Pediatric Epilepsy
Introduction
Epidemiology
Histopathology
Pathophysiology
MTS and Febrile Seizures
Human Herpes Virus 6
Dual Pathology
Differential Diagnosis
Presentation
Evaluation
Electroencephalography and Electrocorticography
Magnetic Resonance Imaging
Positron Emission Tomography and Single Proton Emission Tomography
Magnetoencephalography
Treatment
Antiepileptic Medications
Laser Interstitial Thermal Therapy
Resective Surgery
Conclusion
Chapter 36 Anteromesial Temporal Lobectomy
Introduction
Historical Evolution of the Surgical Technique
Surgical Technique
Positioning the Patient
Scalp Incision
Craniotomy
Neocortical Resection
Mesial Temporal Resection
Complications
Outcome
Chapter 37 Selective Amygdalohippocampectomy
Introduction
Patient Selection and Preoperative Evaluation
Surgical Techniques
Pterional Transsylvian–Transamygdalar Selective Amygdalohippocampectomy
Paramedian Supracerebellar–Transtentorial Selective Amygdalohippocampectomy
Surgical Considerations
Conclusion
Chapter 38 Surgical Management of Lesional Temporal Lobe Epilepsy
Introduction
Vascular Malformations
Pathological Substrate
Tumors
Mechanism of Seizures
Surgical Strategy
Extent of Resection
Lesion
Epileptogenic Zone
Mesial Temporal Structures
Dual Pathology
Location
Special Considerations in Vascular Malformations
Cavernous Hemangioma
Arteriovenous Malformations
Outcome
Tumors
Cavernous Hemangioma
Arteriovenous Malformation
Conclusion
Chapter 39 Surgical Management of MRI-Negative Temporal Lobe Epilepsy
Introduction
Patients and Methods
General Workflow at the “Claudio Munari” Center
MRI-Negative Temporal Lobe Epilepsy Case Series at “Claudio Munari” Epilepsy Surgery Center
Results
Illustrative Case
Continued)
Discussion
Conclusion
Chapter 40 Surgical Management of Insular–Opercular Epilepsy in Children
Introduction
Insular–Opercular Surgical and Functional Anatomy
The Insula
The Opercula
Insular Function and Connectivity
Surgical Pathology
Patient Selection and Surgical Indications
Surgical Candidate Identification and Selection
Preoperative Evaluation
Indication for Invasive Monitoring
Options for Invasive Investigation
Surgical Considerations for Invasive Investigation
Open Method
SEEG Method
Insular–Opercular Resective Surgery
Patient Positioning, Incision, and Craniotomy
Transsylvian Selective Insulectomy
Subpial Operculoinsulectomy
Alternative Procedures
Outcome
Seizure Outcome
Surgical Morbidity
Neuropsychological Outcome
Conclusion
Chapter 41 Focal Cortical Dysplasia: Histopathology, Neuroimaging, and Electroclinical Presentation
Introduction
Histopathology and Classification of Focal Cortical Dysplasia
Neuroimaging of Focal Cortical Dysplasia
Electrographic Signature and Functional Status of Focal Cortical Dysplasia
Electroclinical Epilepsy Syndromes Associated with Focal Cortical Dysplasia
Surgical Treatment and Outcomes of Epilepsy due to Focal Cortical Dysplasia
An Illustrative Case
Phase 1
Phase 2
Phase 3
Chapter 42 Surgical Approaches in Cortical Dysplasia
Introduction
Case 1: Well-Defined Focal Cortical Dysplasia Distant from Eloquent Cortex
Case 2: Poorly Defined Lesion near Eloquent Cortex
Case 3: Discrete Deep Lesion Close to Eloquent Cortex
Case 4: Deep Lesion with Unclear Margins, Away from Eloquent Cortex
Case 5: Deep Lesion, Clear
Margins
Outcomes
Conclusion
Introduction
Epilepsy and Neurodevelopment
Chapter 43 Tuberous Sclerosis Complex
Unknown
Identification of Epilepsy Surgical Candidates and Challenges in Tuberous Sclerosis Complex
Presurgical Evaluation of Tuberous Sclerosis Complex
Imaging Techniques
Neurophysiology
Combined Modalities
Other Experimental Modalities
Controversies in Pediatric Epilepsy Surgery in Tuberous Sclerosis Complex
Tuber-to-Tuber Propagation
Intraversus Extratuber Onset of Seizures
Minimally Invasive or Large Resection Volume?
Can Extensive Preoperative Workup Preclude the Need for Invasive Monitoring?
Chapter 44 Resective Epilepsy Surgery for Tuberous Sclerosis Complex
Introduction
Epilepsy Syndromes in Tuberous Sclerosis Complex: Why Is It Difficult to Treat?
The History of Resective Surgery
Future Surgical Management Strategies
An Illustrative Case
Acknowldgement
Chapter 45 Extratemporal Resection and Staged Epilepsy Surgery in Children
Introduction
Unique Considerations in Pediatric Extratemporal Epilepsy Surgery
Presurgical Evaluation
Noninvasive Modalities
Functional Mapping
Surgical Techniques
Implantation of Subdural Electrodes
Stereoelectroencephalography
Seizure Focus Resection
Stereotactic Laser Ablation
Results
Conclusion
Chapter 46 Supplementary Sensorimotor Area Surgery
Introduction
Anatomy
Supplementary Sensorimotor Area Function
SMA Syndrome
Functional Mapping
Seizures Arising from the SSMA
Seizure Outcomes following Surgical Resection
Chapter 47 Rolandic Cortex Surgery
Introduction
Historical Background of Rolandic Surgery
Clinical and Pathological Features of Rolandic Epilepsy: Idiopathic and Lesional
Evaluation of Patients for Surgery
Presurgical Evaluation
Locating the Epileptogenic Zone: The Case for Extraoperative Seizure Mapping
Subdural Grid Placement
Localization of Primary Motor and Somatosensory Cortex
Resective Surgery
Surgical Complications
Neurological Outcomes
Seizure Outcomes
Conclusion
Chapter 48 Anterior Peri-insular Quadrantotomy
Introduction
Indications
An Illustrative Case
Preoperative Assessment
Anesthetic Considerations
Intraoperative Functional Mapping
Operative Technique of Periinsular Anterior Quadrantotomy
Step 1: Suprainsular Window
Step 2: Anterior Callosotomy
Step 3: Intrafrontal Disconnection
Step 4: Frontobasal Disconnection
Closure
General Considerations and Complications
Outcomes
Acknowlgement
Chapter 49 Posterior Peri-insular Quadrantotomy
Introduction
Indications
Case Series
Illustrative Case
Preoperative Assessment
Anesthetic Considerations
Intraoperative Functional Mapping
Anatomical Localization of Perirolandic Cortices
Electrophysiological Localization of Perirolandic Cortices
Electrocorticographical Localization and Monitoring of Perirolandic Cortices
Operative Technique of Peri-insular Posterior Quadrantotomy
Step 1: Infrainsular Window
Step 2: Parieto-occipital Disconnection
Closure and Early Postoperative Management
Complications
Outcomes
Acknowledgment
Chapter 50 Tailored Extratemporal Resection in Children with Epilepsy
Introduction
Patient Selection and Presurgical Evaluation
Noninvasive Evaluation
Invasive Evaluation
Surgery
Presurgical Planning
Types of Resection
Technical Aspects
Sites of Resections
Histological Findings
Outcomes
Seizure Outcome
Functional Outcome
Complications
Two Illustrative Cases
Case One
Case Two
Chapter 51 Surgical Management of MRINegative Extratemporal Lobe Epilepsy
Introduction
Seizure Semiology
Hemispheric Lateralization
Frontal Lobe Seizures
Parietal-Occipital Onset
Insular Seizures
Confirming Localization
Advanced Imaging and Electrophysiology
Positron Emission Tomography
Single Photon Emission Computed Tomography
Magnetoencephalography
Electrocorticography
Stereotactic EEG
Seizure Outcomes
Chapter 52 Surgical Management of Hypothalamic Hamartomas
Anatomic Features
Molecular and Genetic Characteristics
Clinical Manifestations
Epilepsy
Precocious Puberty
Behavioral, Cognitive, and Psychiatric Disorders
Neuroimaging Findings
Computed Tomography
Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy
Electrographic Findings
Surgical Decision Making
Surgical Techniques
Microsurgical Approaches
Staged Approaches and Repeat Surgery
Stereotactic Gamma Knife Radiosurgery
Stereotactic Laser Ablation
Conclusion
Chapter 53 Hemispherectomy and Hemispherotomy Techniques in Pediatric Epilepsy Surgery
Introduction
Hemispheric Epilepsy Surgery: From Resection to Disconnection
Epilepsy Syndromes Associated with Hemispheric Lesions
Infantile Spasms
Hemiconvulsion-Hemiplegia-Epilepsy Syndrome
Sturge-Weber Syndrome
Hemimegalencephaly
Cortical Dysplasia
Rasmussen’s Encephalitis
Porencephalic Cyst
Preoperative Assessment
Medical Intractability
Clinical Status
Physical Examination
Electroencephalographic Assessment
Structural Imaging
Functional Imaging
Wada Test
Neuropsychological Evaluation
Surgical Planning
Surgical Approaches
Anatomical Hemispherectomy
Hemidecortication
Functional Hemispherectomy
Hemispherotomy
Vertical Approach: Transventricular Vertical Hemispherotomy
Special Considerations Regarding the Hemispheric Surgical Approaches
Complications
Outcomes
Conclusion
Chapter 54 Multifocal Epilepsy and Multilobar Resections
Introduction
Indications
Epidemiology
Workup
Surgical Strategies
Pathology
Outcomes
Conclusion
Chapter 55 Anatomical Hemispherectomy
Introduction
Historical Background
Indications
Preoperative Evaluation
Surgical Technique
Anesthesia and Perioperative Monitoring
Operative Positioning and Opening
Hemispherectomy
Closure
Complications
Critical Analysis of Outcome
Chapter 56 Hemidecortication for Intractable Epilepsy
Introduction
History and Development
Patient Selection and Indications
Rasmussen’s Encephalitis
Malformation of Cortical Development
Vascular Injury
Sturge-Weber
The Differences between Hemidecortication and Other Techniques
Presurgical Preparation
Surgical Techniques—Pitfalls
Postoperative Management
Loss of Brain Volume
Postoperative Medications
Drains
Morbidity and Mortality
Meningitis
Hydrocephalus
Outcome and Prognosis
Conclusion
Chapter 57 Functional Hemispherectomy at the University of California, Los Angeles
Introduction
History and Rationale of the UCLA Functional Hemispherectomy
Presurgery Evaluation
Operative Technique and Perioperative Care
Preincision
Opening
Lateral Resection
Mesial Dissections under the Operating Microscope
Closure
Immediate Postoperative and ICU Care
Long-term Follow-up
UCLA Cerebral Hemispherectomy Cohort
Characteristics Based on Etiology
Operation Time and Expected Blood Loss
Postoperative Cerebrospinal Fluid Shunts
Operative-Related Complications and Reoperations
Outcome: Seizure Freedom, Use of Antiepilepsy Drugs, and Cognitive Development
Long-term Studies: Motor Skills, Language, and Rehabilitation
Assessment and Conclusion
Acknowledgment
Chapter 58 Transsylvian Hemispheric Deafferentation
Introduction
Indication and Presurgical Evaluation
Advantages and Limits of the Transsylvian Keyhole Deafferentation
Presurgical Management
Surgical Technique
Craniotomy and Transsylvian Exposure
Temporomesial Disconnection and Unco-amygdalohippocampectomy
Exposure of Whole Ventricular System
Basal and Mesial Disconnection
Postoperative Management
Our Series and Seizure Outcome
Complications
Conclusion
Acknowledgment
Chapter 59 Vertical Parasagittal Hemispherotomy
Introduction
Historical Background
Description of the Surgical Technique of Vertical Parasagittal Hemispherotomy
Approach and Paracentral Resection
Posterior Callosotomy
Interrupting the Hippocampal Tail
Laterothalamic Disconnection
Anterior Callosotomy
Frontal Disconnection
Complications
Results
Incomplete Disconnection
Conclusion
Acknowledgment
Chapter 60 Peri-insular Hemispherotomy
Introduction
Preoperative Evaluation
Surgical Procedure
The Suprainsular Window
Corpus Callosotomy
The Infrainsular Window
Modifications
Surgical Outcome
Complications
Chapter 61 Endoscope-Assisted Hemispherotomy
Introduction
Principles of Endoscope-Assisted Hemispherotomy
Surgical Technique
Preoperative Workup
Procedure
Postoperative Care
Outcomes Following Surgery
Complications
Conclusion
Acknowledgment
Chapter 62 Corpus Callosotomy
Introduction
Corpus Callosotomy
Histological Refinement of Corpus Callosotomy
Anatomical/Imaging Basis and Rationale for Corpus Callosotomy
Case Demonstration of Anterior 1/2 Callosotomy in a Prepubertal Children with LGS and Multiple Seizure Types
Patient Selection Criteria, Surgical Indications, and Preoperative Assessment
Surgical Technique
Complications and Disconnection Syndromes
The Taipei VGH Experience
Conclusion
Chapter 63 Endoscope-Assisted Corpus Callosotomy with Anterior, Hippocampal, and Posterior Commissurotomy
Introduction
Indications
Principles of Endoscope-Assisted Corpus Callosotomy and Anterior, Middle, and Posterior Commissurotomy
Surgical Technique
Presurgical Workup
Procedure
Results
Seizure Outcomes
Neuropsychological Outcomes
Control Cohort
Complications
Discussion
Conclusion
Acknowledgment
Chapter 64 Endoscopic Disconnection of Hypothalamic Hamartomas
Introduction
Indication of Endoscopic Transventricular Disconnection
Description of the Endoscopic Transventricular Disconnection Technique
Surgical Results
Conclusion
Chapter 65 Multiple Subpial Transections in Children with Refractory Epilepsy
Introduction
Fundamentals of the Multiple Subpial Transection Technique
Indications for the Multiple Subpial Transection Technique
Focal Seizures Located in the Eloquent Cortex
Landau-Kleffner Syndrome
Syndrome of Malignant Rolandic-Sylvian Epilepsy
Other Indications
Presurgical Evaluation
Surgical Technique
Complications
Discussion
Conclusion
Chapter 66 Hippocampal Subpial Transection
Introduction
Rationale of Hippocampal Subpial Transection
Surgical Exposure of the Hippocampus
Electrocorticography over the Hippocampus and Amygdala
Hippocampal Transection
Seizure Outcome
Postoperative Verbal Memory
Pediatric Data and Case Report
An Illustrative Case
Other Series
Chapter 67 Vagal Nerve Stimulation
Introduction
Anatomy and Physiology of Vagal Nerve Stimulation
Preclinical History of Vagal Nerve Stimulation
Clinical Trials for Vagal Nerve Stimulation in Epilepsy
Vagal Nerve Stimulation in Pediatric Epilepsy
Operative Technique for Vagal Nerve Stimulation Placement
Special Considerations and Complications
Vagal Nerve Stimulation Hardware and Stimulation Parameters
Conclusion
Chapter 68 Cortical and Deep Brain Stimulation
Introduction
Brain Stimulation Regions
Subcortical Structures
Cerebral Cortex
Responsive Neurostimulation
Transcranial Magnetic Stimulation
Surgical Technique
Use in Pediatric Subjects
Conclusion
Chapter 69 Radiosurgical Treatment for Epilepsy
Introduction
Mesial Temporal Lobe Epilepsy
Dose
Target
Patient Selection
Complications
Current Indications
Hypothalamic Hamartomas
HH Classification and Treatment Strategy
Effect of GKS on Behavior and Cognitive Functions
Limits and Strengths of Radiosurgery in HH
Cavernous Malformations
Prognostic Factors
Management Strategy
Target Definition
Dose Selection
Radiosurgical Callosotomy
Callosotomy Techniques
Gamma Knife Callosotomy
Indications
Extent of Callosotomy
Dose Prescription
Conclusion
Chapter 70 Stereotactic Laser Ablation for Hypothalamic Hamartomas
Introduction
Indications
Preoperative Workup
Choice of Laser System and Application
Imaging and Ablation
Postoperative Imaging
Postoperative Care
Outcomes
Complications
Conclusion
Chapter 71 MRI-Guided Laser Thermal Therapy in Pediatric Epilepsy Surgery
Introduction
MRIgLITT Technology
Commercially Available MRIgLITT Systems
Surgical Technique
MRIgLITT Advantages
MRIgLITT in Pediatric Epilepsy
Conclusion
Chapter 72 Surgical Failure and Reoperation
Introduction
Surgical Failure
Defining Surgical Failure
Acute Postoperative Seizures
Prediction of Surgical Failure
Causes of Surgical Failure
Evaluation for Reoperation
Reoperation
Repeat Localization in Reoperation
Reoperation in Focal Cortical Resections
Reoperation after Epileptogenic Lesional Resection
Reoperation in Temporal Lobectomy
Evolving Concepts in Goals of Pediatric Epilepsy Surgery
A Changing Landscape in Pediatric Epilepsy Surgery
New Techniques in Reoperation and Palliation
Conclusion
Chapter 73 Postoperative Seizure Control
Introduction
Outcome with Respect to Seizure Control after Surgery
Seizure Control after Surgery for a Structural Lesion
Seizure Control after Temporal Lobe Surgery
Seizure Control after Extratemporal Resection
Seizure Control after Hemispherectomy
Seizure Outcome in Relation to Pathology of the Epileptogenic Substrate
Seizure Outcome after Reoperation
Seizure Control after Surgical Treatment of Epilepsy in Children without an Obvious Structural Lesion
Factors Affecting Postoperative Seizure Control
Presurgical Factors
Surgical Factors
Postsurgical Factors
Medical Management after Epilepsy Surgery
Conclusion
Chapter 74 Postoperative Neuropsychological and Psychosocial Outcome
Introduction
Neuropsychological Outcome
Resection from the Temporal Lobe
Extratemporal Resections
Hemispheric Resections
Long-term Outcomes
Psychosocial Outcome
Psychiatric and Behavioral Outcomes
Health-related Quality of Life
Predictors of Neuropsychological and Psychosocial Outcomes
Future Directions and Conclusion
Index
Access Code

Citation preview

To access the additional media content available with this e-book via Thieme MedOne, please use the code and follow the instructions provided at the back of the e-book.

Pediatric Epilepsy Surgery Preoperative Assessment and Surgical Treatment Second Edition

Oğuz Çataltepe, MD Professor of Neurosurgery and Pediatrics; Director, Pediatric Neurosurgery and Epilepsy Surgery Department of Neurosurgery University of Massachusetts Medical School and Medical Center Worcester, Massachusetts, USA

George I. Jallo, MD Professor of Neurosurgery and Pediatrics; Director Institute for Brain Protection Science Johns Hopkins All Children’s Hospital St. Petersburg, Florida, USA

796 illustrations

Thieme New York • Stuttgart • Delhi • Rio de Janeiro

Library of Congress Cataloging-in-Publication Data Names: Cataltepe, Oguz, editor. | Jallo, George I., editor. Title: Pediatric epilepsy surgery: preoperative assessment and   surgical treatment / [edited by] Oguz Cataltepe, George I. Jallo. Description: Second edition. | New York: Thieme, [2019] |   Includes bibliographical references and index. Identifiers: LCCN 2018051853| ISBN 9781626238169   (hard- cover: alk. paper) | ISBN 9781626238176 (eISBN) Subjects: | MESH: Epilepsy—surgery | Child | Preoperative   Care—methods | Neurosurgical Procedures—methods | Post  operative Care—methods Classification: LCC RJ496.E6 | NLM WL 385 | DDC 618.92/853059—dc23 LC record available at https://lccn.loc.gov/2018051853

Important note: Medicine is an ever-changing science undergoing continual development. Research and clinical experience are continually expanding our knowledge, in particular our knowledge of proper treatment and drug therapy. Insofar as this book mentions any dosage or application, readers may rest assured that the authors, editors, and publishers have made every effort to ensure that such references are in accordance with the state of knowledge at the time of production of the book. Nevertheless, this does not involve, imply, or express any guarantee or responsibility on the part of the publishers in respect to any dosage instructions and forms of applications stated in the book. Every user is requested to examine carefully the manufacturers’ leaflets accompanying each drug and to check, if necessary in consultation with a physician or specialist, whether the dosage schedules mentioned therein or the contraindications stated by the manufacturers differ from the statements made in the present book. Such examination is particularly important with drugs that are either rarely used or have been newly released on the market. Every dosage schedule or every form of application used is entirely at the user’s own risk and responsibility. The authors and publishers request every user to report to the publishers any discrepancies or inaccuracies noticed. If errors in this work are found after publication, errata will be posted at www. thieme.com on the product description page. Some of the product names, patents, and registered designs referred to in this book are in fact registered trademarks or proprietary names even though specific reference to this fact is not always made in the text. Therefore, the appearance of a name without designation as proprietary is not to be construed as a representation by the publisher that it is in the public domain.

©2020. Thieme. All rights reserved. Thieme Medical Publishers New York 333 Seventh Avenue New York, New York 10001 USA +1 800 782 3488 [email protected] Thieme Publishers Stuttgart Rüdigerstrasse 14, 70469 Stuttgart, Germany +49 [0]711 8931 421, [email protected] Thieme Publishers Delhi A-12, Second Floor, Sector-2, Noida-201301 Uttar Pradesh, India +91 120 45 566 00, [email protected] Thieme Publishers Rio de Janeiro, Thieme Publicações Ltda. Edifício Rodolpho de Paoli, 25º andar Av. Nilo Peçanha, 50 – Sala 2508, Rio de Janeiro 20020-906 Brasil +55 21 3172-2297 Cover design: Thieme Publishing Group Typesetting by DiTech Process Solutions, India Printed in USA by King Printing Company, Inc. ISBN 978-1-62623-816-9 Also available as an e-book: eISBN 978-1-62623-817-6

54321

This book, including all parts thereof, is legally protected by copyright. Any use, exploitation, or commercialization outside the narrow limits set by copyright legislation, without the publisher’s consent, is illegal and liable to prosecution. This applies in particular to photostat reproduction, copying, mimeographing, preparation of microfilms, and electronic data processing and storage.

Dedicated to my wife, Şule, and to our children, Deniz and Arda. – Oğuz Çataltepe

Dedicated to Maxwell, Nicholas, and Alexis, who remind me all about love, trust, and living life. – George I. Jallo



Contents Videos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xii

Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xiii

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xv

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xvi

Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xvii

M. Gazi Yaşargil

Part I. Introduction to Epilepsy in Children 1

Basic Considerations of Pediatric Epilepsy Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .



2

A Historical Review of Epilepsy Surgery and Its Application in Children . . . . . . . . . . . . . . . . . . . . . .

 10

Epidemiology of Intractable Epilepsy in Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 21



3

4

5

6

7

8

9

10

Oğuz Çataltepe and George I. Jallo

Deniz Çataltepe and Oğuz Çataltepe

Peter Camfield and Carol Camfield

Genetics in Epilepsy Surgery

3

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  28

Maurits W. C. B. Sanders, Floor E. Jansen, Bobby P. C. Koeleman, and Kees P. J. Braun

Surgical Neuropathology of Pediatric Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 36

Epilepsy and Brain Plasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 47

Effects of Seizures and Their Comorbidities on the Developing Brain . . . . . . . . . . . . . . . . . . . . . . . . .

 53

Ethical Considerations in Pediatric Epilepsy Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 57

Infantile and Childhood-Onset Catastrophic Epilepsy Syndromes . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 61

Thomas W. Smith

Shilpa D. Kadam and Michael V. Johnston Susan Lee Fong and Carl E. Stafstrom

George M. Ibrahim and Mark Bernstein Hirokazu Oguni

Epilepsy Surgery for Congenital or Early Brain Lesions

Ahsan Moosa Naduvil Valappil, Tobias Loddenkemper, and Elaine Wyllie

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  70

Part II. Preoperative Assessment Section IIa. Preoperative Clinical and Neuropsychological Assessment 11

12

13

Intractable Epilepsy in Children and Selection of Surgical Candidates . . . . . . . . . . . . . . . . . . . . . . . .

 81

Clinical Semiology in Preoperative Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 92

Preoperative Neuropsychological and Cognitive Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 103

Çiğdem Inan Akman and James J. Riviello Jr. András Fogarasi

Katrina M. Boyer

vii

Contents

Section IIb. Preoperative Electrophysiological Assessment 14

Electroencephalography and Noninvasive Electrophysiological Assessment . . . . . . . . . . . . . . . .

 110

Invasive Electrophysiological Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 127

Extra- and Intraoperative Electrocortical Stimulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 134

17 Stereoelectroencephalography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 145

18 Magnetoencephalography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 153



15

16

Çiğdem Inan Akman and James J. Riviello Jr.

Prasanna Jayakar, Ann Hyslop, and Ian Miller Ingrid Tuxhorn

Ika Noviawaty and Patrick Chauvel

Hiroshi Otsubo, Kota Kagawa, and O. Carter Snead III

Section IIc. Preoperative Neuroimaging 19

Structural Brain Imaging in Pediatric Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 161

Functional Magnetic Resonance Imaging in Pediatric Epilepsy Surgery . . . . . . . . . . . . . . . . . . . . . .

 171

21 Application of Positron Emission Tomography and Single-Photon Emission Computed Tomography in Pediatric Epilepsy Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 180

22

 194



20



Charles Raybaud and Elysa Widjaja

Torsten Baldeweg and Frédérique Liégeois

Stephen J. Falchek, Ajay Kumar, and Harry T. Chugani

Multimodality Imaging and Coregistration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Prashin C. Unadkat, Walid Ibn Essayed, John M. K. Mislow, and Alexandra J. Golby

Part III. Surgical Anatomy and Mapping Techniques 23

Oğuz Çataltepe

Cerebral Cortex: Embryological Development and Topographical Anatomy . . . . . . . . . . . . . . . . .

 205

24

Tractographic Anatomy of White Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 228

25 Localization of Motor Cortex and Subcortical Pathways Using Functional Magnetic Resonance Imaging and Diffusion Tensor Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 240

26

The Wada Test: Lateralization of Language and Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 251

Language Lateralization and Localization: Functional Magnetic Resonance Imaging . . . . . . .

 255

Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia . . . . . .

 259

 Sandip S. Panesar, David T. Fernandes-Cabral, Antonio Meola, Fang-Cheng Yeh, Maximiliano Nunez, and Juan C. Fernandez-Miranda



27

28

29

Sathish Kumar Dundamadappa and Mohit Maheshwari

David W. Loring and Gregory P. Lee Cristina Go and Elizabeth Donner

James L. Stone and Bartosz Grobelny



Doris D. Wang, John D. Rolston, and Mitchel S. Berger

Cortical Stimulation and Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

277

30

Subcortical Mapping During Intracranial Surgery in Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 284

 Francesco Sala and Davide Giampiccolo

viii

Contents

Part IV. Surgical Treatment of Epilepsy Section IVa. Anesthesia 31 Anesthetic Considerations and Postoperative Intensive Care Unit Care in Pediatric Epilepsy Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

301

32

 307



Sulpicio G. Soriano and Michael L. McManus

Pediatric Awake Craniotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Gaston Echaniz, Michael Tan, Ibrahim Jalloh, Samuel Strantzas, and Tara Der

Section IVb. Intracranial Electrode Placement for Invasive Monitoring 33 Implantation of Strip, Grid, and Depth Electrodes for Invasive Electrophysiological Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 315

34

 323



Oğuz Çataltepe and Julie G. Pilitsis

Stereoelectroencephalography in Children: Methodology and Surgical Technique . . . . . . . . .

Robert A. McGovern and Jorge A. Gonzalez-Martinez

Section IVc. Temporal Lobe Epilepsy and Surgical Approaches 35

Rafael Uribe, George I. Jallo, and Caitlin Hoffman

Mesial Temporal Sclerosis in Pediatric Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 333

36

Anteromesial Temporal Lobectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 341

37 Selective Amygdalohippocampectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 351

38

Surgical Management of Lesional Temporal Lobe Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 364

Surgical Management of MRI-Negative Temporal Lobe Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 376

 Oğuz Çataltepe

39

Uğur Türe, Ahmet Hilmi Kaya, Berrin Aktekin, and Canan Aykut Bingöl Oğuz Çataltepe

Francesco Cardinale, Piergiorgio d’Orio, and Michele Rizzi

Section IVd. Extratemporal Lobe Epilepsy and Surgical Approaches 40

Surgical Management of Insular–Opercular Epilepsy in Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 386

41 Focal Cortical Dysplasia: Histopathology, Neuroimaging, and Electroclinical Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Olesya Grinenko and Imad M. Najm

 404

42

Surgical Approaches in Cortical Dysplasia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 414

43 Tuberous Sclerosis Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

424

44

Resective Epilepsy Surgery for Tuberous Sclerosis Complex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 432



Alexander G. Weil and Sanjiv Bhatia

 Jeffrey Bolton, Sanjay P. Prabhu, Eun-Hyoung Park, Scellig S. Stone, and Joseph R. Madsen

45

Jurriaan M. Peters and Mustafa Şahin

Jeffrey S. Raskin, Daniel J. Curry, and Howard L. Weiner



Daxa M. Patel, Howard L. Weiner, and Robert J. Bollo

Extratemporal Resection and Staged Epilepsy Surgery in Children . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 437

46

Supplementary Sensorimotor Area Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 447

 Jarod L. Roland and Matthew D. Smyth

ix

Contents

47 Rolandic Cortex Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 452

48



Ibrahim Jalloh and James T. Rutka



Giulia Cossu, Mahmoud Messerer, Sebastien Lebon, Etienne Pralong, Margitta Seeck, and Roy Thomas Daniel

Anterior Peri-insular Quadrantotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 459

49

Posterior Peri-insular Quadrantotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 465

Tailored Extratemporal Resection in Children with Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 472

51 Surgical Management of MRI-Negative Extratemporal Lobe Epilepsy . . . . . . . . . . . . . . . . . . . . . . . .

Jarod L. Roland and Matthew D. Smyth

 481

52

Surgical Management of Hypothalamic Hamartomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 489

 Giulia Cossu, Mahmoud Messerer, Sebastien Lebon, Etienne Pralong, Krothapalli Srinivasa Babu, Margitta Seeck, and Roy Thomas Daniel

50

Alessandro Consales and Massimo Cossu

 Neena I. Marupudi and Sandeep Mittal

Section IVe. Hemispheric Surgery Techniques 53 Hemispherectomy and Hemispherotomy Techniques in Pediatric Epilepsy Surgery . . . . . . . .

497

54



Oğuz Çataltepe



Rafael Uribe, George I. Jallo, and Caitlin Hoffman

Multifocal Epilepsy and Multilobar Resections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 506

55

Anatomical Hemispherectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 510

Hemidecortication for Intractable Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 519

57 Functional Hemispherectomy at the University of California, Los Angeles . . . . . . . . . . . . . . . . . . .

 530

58

Transsylvian Hemispheric Deafferentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 542

Vertical Parasagittal Hemispherotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 553

Peri-insular Hemispherotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 562

Endoscope-Assisted Hemispherotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 572



56

59

60

61

Concezio Di Rocco, Kostas N. Fountas, and Luca Massimi

Nir Shimony and George I. Jallo

Sandi Lam and Gary W. Mathern Johannes Schramm

Georg Dorfmüller, Mikael Levy, and Sarah Ferrand-Sorbets Robert J. Bollo and Nicholas M. Wetjen

P. Sarat Chandra, Jitin Bajaj, Heri Subianto, and Manjari Tripathi

Section IVf. Other Disconnective Procedures 62

Corpus Callosotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 581

63 Endoscope-Assisted Corpus Callosotomy with Anterior, Hippocampal, and Posterior Commissurotomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

589

64

 597

 Tai-Tong Wong, Shang-Yeong Kwan, Kai-Ping Chang, Min-Lan Tsai, and Kevin Li-Chun Hsieh



x

P. Sarat Chandra, Jitin Bajaj, Heri Subianto, and Manjari Tripathi

Endoscopic Disconnection of Hypothalamic Hamartomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Georg Dorfmüller and Sarah Ferrand-Sorbets

Contents

65

Multiple Subpial Transections in Children with Refractory Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . .



Zulma S. Tovar-Spinoza and James T. Rutka

66

Hippocampal Subpial Transection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Hiroyuki Shimizu

 603

 608

Section IVg. Neuromodulation Procedures 67 Vagal Nerve Stimulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 613

68

 619



Jarod L. Roland, David D. Limbrick Jr., and Matthew D. Smyth

Cortical and Deep Brain Stimulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Edward E. Woodward, Nir Shimony, and George I. Jallo

Section IVh. Radiosurgery and Ablative Procedures 69

70

71

Radiosurgical Treatment for Epilepsy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 625

Stereotactic Laser Ablation for Hypothalamic Hamartomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 640

MRI-Guided Laser Thermal Therapy in Pediatric Epilepsy Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 645

Jean Régis, Hussein Hamdi, and Patrick Chauvel Daniel J. Curry and Nisha Gadgil

Zulma S. Tovar-Spinoza and James T. Rutka

Part V. Postoperative Course and Outcome 72

Surgical Failure and Reoperation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 655

73 Postoperative Seizure Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

667

74

Postoperative Neuropsychological and Psychosocial Outcome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

 676

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

683

 Matthew C. Davis and Jeffrey P. Blount

Seema Adhami and Chellamani Harini Klajdi Puka and Mary Lou Smith

xi



Videos Video 9.1 Epilepsy of infancy with migrating focal seizures due to KCNT1 mutation. Hirokazu Oguni

Video 9.2

Lennox-Gastaut Syndrome and generalized tonic seizures. Hirokazu Oguni

Video 12.1  Epileptic spasms. András Fogarasi

Video 12.2

Lateralization of the seizure onset zone.

Video 16.1

Motor mapping: primary motor hand area.

Video 16.2

Speech mapping.

Video 16.3

Motor mapping.

Video 32.1

Awake craniotomy and cortical stimulation.

András Fogarasi Ingrid Tuxhorn Ingrid Tuxhorn Ingrid Tuxhorn Tara Der

Video 36.1 Anterior temporal lobectomy and Amygdalohippocampectomy. Oğuz Çataltepe

Video 53.1

Peri-insular hemispherotomy.

Video 61.1

Endoscopic-assisted hemispherotomy.

Video 62.1

 nterior Corpus Callosotomy: an anterior ½ callosotomy was performed via the interhemispheric A approach on the left side.

Oğuz Çataltepe

P. Sarat Chandra

Tai-Tong Wong

Video 63.1 The technique of endoscopic complete corpus callosotomy with anterior, hippocampal and posterior commissurotomy. P. Sarat Chandra

xii



Foreword Many diseases of the central nervous system with diverse etiology affect our patients of all ages with acute, subacute, or chronic symptoms and deficits, which are most distressing. Since ancient times up until the present day, rational, irrational, and alternative modes of treatment have been researched, analyzed, and applied. The passion to discover the cause of disease and to explore treatment possibilities and alternative methods has engaged the intellects of the medical field since ancient times, and continues today. On 4 January 1953, at University Hospital Zürich in ­Switzerland, the very first day of my professional career as a neurosurgeon, Professor Hugo Krayenbühl performed surgery on a young patient with intractable temporal epilepsy. A right-sided anterior two thirds temporal lobectomy was the planned procedure. In the adjoining room with glass windows, P ­rofessor Rudolf M. Hess, a neurologist and EEG specialist, trained by Dr. W. A. Cobb in London and G. Walter in Bristol, directed and guided the intraoperative EEG recordings. Two nurses, one resident, and one technician assisted him in the coordination of EEG leads and connections. A very long overhead metal tube originating from the enormous EEG machine and a carrier for the connections curved over the window into the operating room and hung over the open wound to accept the leads from electrodes placed on the exposed surface of the brain. The surgery was time-consuming and progressed slowly. After one millimeter removal of cortex, an EEG recording checked the status of electrical activities in the temporal region. Professor Hess delivered short messages: “Still spikes around…!”, until, finally, with a sense of happy relief: “No more spikes!” Professor Krayenbühl turned and spoke to us, his observers: “The treatment of epilepsy is an obligation for neurosurgery.” The impact of this challenging remark on my first day made a deep impression on me and I contemplated that cooperation between neurophysiology and neurosurgery could be a rational approach to pursue. Whether a coincidence or my destiny, now sixty-five years later, I am honored to write this Foreword for the second edition of Pediatric Epilepsy Surgery by Oğuz Çataltepe and George I. Jallo. This volume of 74 chapters covering specialized aspects of epilepsy, with contributions by experts in this field, has been conducted into a captivating and thought-provoking symphony by the two editors. Advances in mathematics, basic sciences, scientific technology, and the flourishing medical industry have all contributed to developments in the past 200 years. Since the 1950s, a surge in research and the resulting accomplishments have empowered unforeseen breakthroughs in several fields of the sciences and offered the potential to progress towards a more profound research in the broadest fields of the neurosciences. This particular volume explores the progress and achievements in all the fields of neurosciences pertaining to epilepsy, including the impact of neurosurgery,

e.g., the events leading to diagnosis, diligent consideration of indications for the various neurotherapies, the concerns in pre-, intra-, and postoperative nursing, and the importance of neuropharmacology, neuropsychology, neurophysiology, speech therapy, and social care. Wilhelm Sommer (1852-1900), in Erkrankung des Ammonshorns als aetiologisches Moment der Epilepsie, Archiv für Psychiatrie und Nervenkrankheiten (1880;10:631-675), documented changes in Ammon’s horn on the autopsy of brains of 36 epilepsy patients and deducted that this finding could explain the epileptic factor. He gave credits to the observations of G. B. Morgagni and J. E. Greding made in the eighteenth century, and also to Th. H. Meynert, who noted changes in the hippocampus of epilepsy patients in the nineteenth century. Sommer expressed his hope that the results of his study would be confirmed by others. In the year 1899, E. Bratz demonstrated a normal hippocampus on a woodcut, compared it with a hippocampus with gross atrophy and microscopic cell loss from the brain of an epilepsy patient with Ammon’s horn sclerosis (Ammonshornbefunde der Epileptischen. Archiv für Psychiatrie und Nervenkrankheiten. 1899;31:820-836). The availability of modern visualization technology and advances in histopathology are processes that confirm the diagnosis of hippocampal sclerosis, cortical dysplasia, gliosis, and micro-infarctions. The second half of the nineteenth century ushered in a period of major progress in medicine and surgery, for instance, new concepts for hospitals, appointment of licensed nurses, and especially the courageous attempts to alleviate the burden and stress of patients with intractable seizures (see the timeline discussed later). Selective amygdalo-hippocampectomy was pioneered by Paul Niemeyer in 1954, and was developed into the pterional transsylvian approach by myself, which, in the meantime, has also been complemented by the paramedian supracerebellar transtentorial approaches by Uğur Türe. Both are effective in treating patients with medial basal temporal epilepsy. These approaches are surgical explorations in a small, compact, highly functional area of the central nervous system. Successful approach and elimination of the seizure focus is dependent on knowledge of this complex anatomy, and competence in micro-neurosurgical skills and techniques. Laboratory training offers the ideal environment for cadaver dissection to learn ­cisternal, vascular, and parenchymal neuroanatomy, and to perfect micro-technical skills. Selection of treatment modalities relevant to the particular seizure problems of an individual patient is a determining factor in achieving cessation of epileptic events or at least in reducing seizure frequency and severity, and therefore improving quality of life. A center with a multidisciplinary team established within a hospital and dedicated to the treatment, study, and research of epilepsy and its medical and social repercussions has the resources to offer more qualified service to these patients.

xiii

Foreword The range of topics, information, and data in these chapters, when combined together, represent the current opportunities and indicate the future research and treatment landscape in the realm of epilepsy. The insights, judgments, and balanced assessments and conclusions have the potential to stimulate young colleagues to accept

the challenge of preventing these debilitating and frightening episodes called epilepsy. M. Gazi Yaşargil, MD Professor of Neurosurgery Yeditepe University Hospital Istanbul, Turkey

Timeline of the endeavors of neuroscientists and neurosurgeons to develop effective therapies for the various epilepsy types and the etiology of this disease:

1879

W. Macewen

Cortical resection following brain trauma; a 7-year-old boy

1880

J. L. Corning

Electrostimulation of the cervical carotid artery

1882

W. Alexander

Bilateral ligation of the vertebral arteries

1886

Removal of scar tissue

1893–1912

V. Horsley S. H. Jackson F. Krause

1925–1935

O. Foerster

1930–1980

Partial temporal lobe resection and several other excisions

1954

W. Penfield H. Jasper M. Baldwin T. Rasmussen W. Feindel P. Niemeyer

1973

M.G. Yaşargil

Pterional anterior transylvian approach for ­amygdalo-hippocampectomy

2012

U. Türe

Paramedian supracerebellar transtentorial approach for amygdalohippocampectomy

Cerebral Hemispherectomy 1938

K. G. McKenzie

1950

R. A. Krynauw

1970

T. Rasmussen

1992

J. Schramm

1993

J. G. Villemure

Callosotomy 1940 1962

W. P. Van Wagenen et al. J. E. Bogen et al.

1970

A. J. Luessenhop et al.

1993

A. R. Wyler

Cortical Resection of Dysplasia 1971

D. C. Taylor et al.

2003

R. I. Kuzniecky

2004

W. J. Hader et al.

Vagus Nerve Stimulation

xiv

1990

J. K. Penry, J. C. Dean

1998

A. Handforth et al.

Faraday stimulation. First accurate image of motor strip. Removal of scar tissue First use of EEG for epilepsy surgery

Transventricular selective amygdalo-hippocampectomy



Preface The surgical management of epilepsy is often complex and presents exceptional challenges to both clinicians and neurosurgeons. This challenge is especially paramount in children with epilepsy. There are many excellent epilepsy surgery textbooks, but very few focus on the surgical management of epilepsy in children. Since the inception of the first edition of this book, our intention was to fill the void in this field. We conceived this book as a comprehensive pediatric epilepsy surgery textbook aiming to integrate both pediatric neurosurgery and epilepsy surgery perspectives. It is our sincere hope that this revised and updated second edition of this book will bring us even closer to reaching this goal. In the first edition, we had attempted to guide the reader through this complex territory by reviewing the current state-of-the-art preoperative assessment modalities and surgical techniques with extensive details, with each topic elaborated upon by respected experts. Nine years have passed since the publication of the first edition and our aim has not changed. However, we felt obligated to revise this book to be able to stay up-to-date on new concepts and developing techniques in this area. Additionally, the second edition is significantly expanded to provide a more comprehensive and complete coverage of the surgical management of pediatric epilepsy. Naturally, the new edition of this book differs considerably from the first edition. There are 37 all-new chapters in this second edition, and almost all the other chapters have been meticulously updated or indeed re-written. The book starts with two chapters reviewing pediatric epilepsy surgery and its historical development. Part I is devoted to pediatric epilepsy and includes chapters discussing epidemiology, neuropathology, brain plasticity, the effect of seizures on the developing brain, ethical considerations, and infantile catastrophic epilepsy syndromes. Part II provides extensive coverage of preoperative assessment under several sections and includes 12 chapters on clinical and neuropsychological assessment, electrophysiology, and neuroimaging. Part III is a new addition to this edition as it aims to review surgical anatomy and mapping techniques. Brain mapping techniques with different modalities, such as fMRI and cortical/subcortical stimulation techniques, have been more frequently used in surgical epilepsy cases over the past 10 years. However, this remains an area of limited experience for many neurosurgeons and centers due to the scarcity of resources and expertise. Standards and approaches also differ center to center. Therefore, a new section in this second edition was devoted to this significant topic, and we are grateful to have contributions from the most respected experts in this field. This part opens with two chapters reviewing the topographic anatomy of the cerebral cortex and white matter with numerous illustrations to enhance the understanding of the content. We believe that it is essential for the neurosurgeon, epileptologist, and electrophysiologist to have a thorough understanding of cortical and white matter anatomy. This allows for a better appreciation of the advantages and limitations of cortical stimulation and current mapping techniques. Other chapters of this section review the most common cortical/subcortical

localization and mapping techniques such as fMRI, WADA test, and magnetic and electrical cortical and subcortical stimulation modalities. Part IV constitutes the backbone of the book and includes 41 chapters to cover all the topics related to the surgical treatment of pediatric epilepsy with extensive details. This section has been expanded significantly in the second edition with many new chapters written by the most respected experts in these areas to define surgical pathologies, anesthetic techniques, and state-of-the-art surgical approaches. Part IV includes eight sections that review anesthetic considerations and approaches, intracranial invasive diagnostic techniques, surgical approaches to temporal and extratemporal epilepsy, multilobar and hemispheric surgical techniques, various disconnection, neuromodulation, and ablation procedures. We invited the most experienced authors to write full chapters dedicated to all major variations of these surgical procedures to facilitate a high-level understanding of these surgical techniques. The book ends with Part V that focuses on postoperative course and outcome. This section includes chapters on surgical failure and re-operation, postoperative seizure control, and, finally, postoperative neuropsychological and psychosocial outcomes. The readers may notice that some chapters in the book may have a certain degree of overlap. Although this is somewhat inevitable in multi-author textbooks, we also purposefully kept some of these overlapping parts of the chapters to give the reader a better overview of differing points of view and practices on that specific topic. We also felt that removing all overlaps would disturb the completeness and integrity of the respective chapters. This is a pediatric epilepsy surgery volume and, of course, our main goal is to reach the epilepsy surgeons and epileptologists who are managing children with intractable epilepsy. We also believe that this textbook is a useful reference book, thanks to our distinguished contributors, even for the experienced subspecialists in epilepsy surgery. It is also intended to be a reference book for everyone on the medical team caring for children with epilepsy, including both adult and pediatric neurosurgeons and neurologists, residents and fellows, clinical neuropsychologists, electrophysiologists, neuroradiologists, and medical students who are involved in the assessment and surgical management of epilepsy patients. This book is a compilation of chapters that were written by the foremost experts in their field around the globe from 14 different countries within North America, Europe, and Asia. They represent a wide range of disciplines and experiences. They were given considerable flexibility and independence in preparing their chapters to be comprehensive and state-of-the-art. We felt honored to assemble and edit this vast body of work that provides a remarkable depth to the text. We are very grateful to our contributors. It is our sincere hope that this new edition of our book will fill the void in this field and be instrumental in improving the care of both children and adult patients with epilepsy. Oğuz Çataltepe, MD George I. Jallo, MD xv



Acknowledgments We thank our colleagues who work with us, especially the epileptologists. These clinicians are the most essential part of the multidisciplinary team that manage epilepsy patients. We are also thankful to our fellows, residents, and medical students for their continuous engagement and insight. Most of all, we would like to thank our pediatric patients and their parents who entrust us with their most precious treasures. This is an exceptional honor and huge responsibility for all of us. We would also like to acknowledge our friend, colleague, and one of the contributors of this book, Dr. Sanjiv Bhatia, who passed away unexpectedly in May 2018. He

xvi

was a respected member of Nicklaus Children’s Hospital, University of Miami, and pediatric neurosurgery community. Dr. Bhatia was a passionate advocate for children with epilepsy and he will be remembered for his significant contributions to pediatric epilepsy surgery. We are grateful to the Thieme editorial and production team members, particularly Timothy Hiscock. Their professionalism, expert assistance, and guidance enabled us to complete this book. Oğuz Çataltepe, MD George I. Jallo, MD



Contributors Seema Adhami, MD, MRCP Associate Professor Pediatrics and Neurology; Chief Division of Pediatric Neurology University of Massachusetts Medical School and Medical  Center Worcester, Massachusetts, USA Çiğdem Inan Akman, MD Associate Professor Department of Neurology; Director Department of Pediatric Epilepsy Columbia University Medical Center Morgan Stanley Children’s Hospital New York—Presbyterian Hospital New York, USA Berrin Aktekin, MD Professor Department of Neurology Yeditepe University School of Medicine Kadikoy, Istanbul, Turkey Krothapalli Srinivasa Babu, PhD Senior Scientist Neurophysiology Laboratory Department of Neurological Sciences Christian Medical College Vellore, Chennai, India Jitin Bajaj, MD Assistant Professor Department of Neurosurgery NSCB Government Medical College Jabalpur, Madhya Pradesh, India Torsten Baldeweg, MD, CCST Professor Developmental Cognitive Neuroscience UCL Institute of Child Health Holborn, London, England Mitchel S. Berger, MD, FAANS, FACS Professor and Chair Department of Neurological Surgery University of California San Francisco San Francisco, California, USA Mark Bernstein, MD, MHSc (Bioethics), FRCSC The Greg Wilkins-Barrick Chair in International   Surgery Professor Department of Surgery University of Toronto Toronto Western Hospital Toronto, Ontario, Canada

Sanjiv Bhatia, MD Pediatric Neurosurgeon Chief, Department of Surgery Nicklaus Children’s Hospital Brain Institute; Affiliated Professor Department of Neurosurgery University of Miami Miller School of Medicine Miami, Florida, USA

†

Canan Aykut Bingöl, MD Professor Department of Neurology Yeditepe University School of Medicine Kadikoy, Istanbul, Turkey Jeffrey P. Blount, MD Professor Division of Neurosurgery Chief Section of Pediatric Neurosurgery Children’s of Alabama Hospital The University of Alabama at Birmingham Birmingham, Alabama, USA Robert J. Bollo, MD, MS, FACS, FAAP, FAANS Associate Professor Department of Neurosurgery Surgical Director, Pediatric Epilepsy Program University of Utah School of Medicine Primary Children’s Hospital Salt Lake City, Utah, USA Jeffrey Bolton, MD Instructor Department of Neurology Boston Children’s Hospital Boston, Massachusetts, USA Katrina M. Boyer, PhD Assistant Professor of Psychology Harvard Medical School Staff Neuropsychologist Director, Neuropsychology and Epilepsy Program Epilepsy Center Boston Children’s Hospital Boston, Massachusetts, USA Kees P. J. Braun, MD PhD Professor of Child Neurology Brain Center Rudolf Magnus University Medical Center Utrecht Utrecht, The Netherlands Carol Camfield, MD Professor Emeritus of Child Neurology Department of Pediatrics Dalhousie University The IWK Health Center Halifax, Nova Scotia, Canada

xvii

 Contributors Peter Camfield, MD Professor Emeritus of Child Neurology Department of Pediatrics Dalhousie University The IWK Health Center Halifax, Nova Scotia, Canada Francesco Cardinale, MD, PhD Neurosurgeon Department of Neurosurgery “C. Munari” Epilepsy Surgery Center Azienda Ospedale Niguarda Ca’ Granda Milan, Italy Oğuz Çataltepe, MD Professor of Neurosurgery and Pediatrics; Director, Pediatric Neurosurgery and Epilepsy Surgery Department of Neurosurgery University of Massachusetts Medical School and   Medical Center Worcester, Massachusetts, USA

Massimo Cossu, MD Neurosurgeon Department of Neuroscience “C. Munari” Epilepsy Surgery Center Azienda Ospedale Metropolitano Niguarda Milan, Italy Daniel J. Curry, MD Associate Professor Department of Neurosurgery Baylor College of Medicine Director of Functional Neurosurgery and Epilepsy  Surgery Texas Children’s Hospital Houston, Texas, USA

Deniz Çataltepe, BA, MPhil Department of History and Philosophy of Science University of Cambridge Cambridge, UK; Medical Student University of Massachusetts Medical School Worcester, Massachusetts, USA

Piergiorgio d’Orio, MD Neurosurgeon Department of Neurosurgery “C. Munari” Epilepsy Surgery Centre Azienda Ospedale Niguarda Ca’ Granda Milan, Italy

P. Sarat Chandra, MD Professor Department of Neurosurgery All India Institute of Medical Sciences AIIMS National Brain Research Center New Delhi, Delhi India

Roy Thomas Daniel, MD Professor of Neurosurgery Department of Clinical Neuroscience Section of Neurosurgery University Hospital of Lausanne Lausanne, Switzerland

Kai-Ping Chang, MD Attending Physician Department of Pediatrics Taipei Veterans General Hospital Taipei, Taiwan Patrick Chauvel, MD Staff Neurologist Epilepsy Center Neurological Institute Cleveland Clinic Cleveland, Ohio, USA Harry T. Chugani, MD Professor Department of Neurology New York University School of Medicine Langone’s Comprehensive Epilepsy Center New York, USA Alessandro Consales, MD Neurosurgeon Department of Neurosurgery Giannina Gaslini Children’s Research Hospital Genoa, Italy

xviii

Giulia Cossu, MD Staff Neurosurgeon Department of Neurosurgery University Hospital of Lausanne Lausanne, Switzerland

Matthew C. Davis, MD Resident Department of Neurosurgery Section of Pediatric Neurosurgery Children’s of Alabama The University of Alabama at Birmingham Birmingham, Alabama, USA Tara Der, MSc, MD, FRCPC Assistant Professor Department of Anesthesia and Pain Medicine ; Director of Neurosurgical and Craniofacial   Anesthesia Program University of Toronto The Hospital for Sick Children Toronto, Ontario, Canada Elizabeth Donner, MD, FRCP(C) Associate Professor Department of Pediatrics; Director of Comprehensive Epilepsy Program Division of Neurology University of Toronto The Hospital for Sick Children Toronto, Ontario, Canada

Contributors Georg Dorfmüller, MD, PhD Professor and Chair Department of Pediatric Neurosurgery Rothschild Foundation Hospital Paris, France

Susan Lee Fong, MD Resident Department of Pediatric Neurology Johns Hopkins University School of Medicine Baltimore, Maryland, USA

Sathish Kumar Dundamadappa, MD Associate Professor Department of Radiology Director of Neuroradiology University of Massachusetts Medical School and   Medical Center Worcester, Massachusetts, USA

Kostas N. Fountas, MD Head and Assistant Professor Department of Neurosurgery University of Thessaly School of Medicine Larissa, Greece

Gaston Echaniz, MD, FACS Staff Anesthesiologist Department of Anesthesiology Vall d’Hebron Hospital Autonomous University of Barcelona Barcelona, Spain Walid Ibn Essayed, MD Clinical Fellow Department of Neurosurgery Brigham and Women’s Hospital Boston, Massachusetts, USA Stephen J. Falchek, MD Pediatric Neurologist and Clinical Neurophysiologist; Chief, Division of Pediatric Neurology Department of Neurology Nemours/Alfred I. DuPont Hospital for Children Sidney Kimmel Medical College at Thomas Jefferson  University Wilmington, Delaware, USA David T. Fernandes Cabral, MD Professor Department of Neurological Surgery UPMC Presbyterian Pittsburgh, Pennsylvania, USA Juan C. Fernandez-Miranda, MD, FACS Professor Department of Neurosurgery and Medicine; Surgical Director Brain Tumor, Skull Base, and Pituitary Centers Stanford University Medical Center Palo Alto, California, USA Sarah Ferrand-Sorbets, MD Pediatric Neurosurgeon Department of Pediatric Neurosurgery Rothschild Foundation Hospital Paris, France András Fogarasi, MD, PhD Professor Head of Neurology Department Bethesda Children’s Hospital Budapest, Hungary

Nisha Gadgil, MD Resident Department of Neurosurgery Baylor College of Medicine Houston, Texas, USA Davide Giampiccolo, MD Neurosurgery Resident Institute of Neurosurgery University Hospital University of Verona Verona, Italy Cristina Go, MD, ABPN-Neurology, CSCN-EEG Associate Professor Department of Pediatrics Division of Neurology University of Toronto The Hospital for Sick Children Toronto, Ontario, Canada Alexandra J. Golby, MD Professor of Neurosurgery and Radiology Harvard Medical School Haley Distinguished Chair in the Neurosciences Department of Neurosurgery Brigham and Women’s Hospital Boston, Massachusetts, USA Jorge A. Gonzalez-Martinez, MD, PhD Staff Neurosurgeon Epilepsy Center Neurological Institute Cleveland Clinic Cleveland, Ohio, USA Olesya Grinenko, MD, PhD Neurologist Epilepsy Center Neurological Institute Cleveland Clinic Cleveland, Ohio, USA Bartosz Grobelny, MD Clinical Fellow in Stereotactic and Functional  Neurosurgery Department of Neurosurgery Emory University Hospital Atlanta, Georgia, USA

xix

 Contributors Hussein Hamdi, MD Neurosurgeon Department of Neurological Surgery Functional and Stereotaxy Unit Tanta University Tanta, Egypt; Department of Stereotactic and Functional Neurosurgery Aix Marseille University Timone University Hospitals Marseille, France Chellamani Harini, MD Instructor Department of Neurology Harvard Medical School Boston Children’s Hospital Boston, Massachusetts, USA Caitlin Hoffman, MD Assistant Professor Department of Neurosurgery NewYork–Presbyterian Hospital Weill Cornell Medical Center New York, USA Kevin Li-Chun Hsieh, MD Attending Physician Department of Medical Imaging Taipei Medical University Hospital Taipei, Taiwan Ann Hyslop, MD Director of Neurocritical Care Program Department of Pediatric Neurology Brain Institute Nicklaus Children’s Hospital Miami, Florida, USA George M. Ibrahim, MD Assistant Professor Division of Neurosurgery University of Toronto Fohio Toronto, Ontario, Canada George I. Jallo, MD Professor of Neurosurgery and Pediatrics; Director Institute for Brain Protection Science Johns Hopkins All Children’s Hospital St. Petersburg, Florida, USA Ibrahim Jalloh, MA, PhD, FRCS (SN) Consultant Neurosurgeon Cambridge University Hospitals NHS Foundation Trust Cambridge, United Kingdom Floor E. Jansen, MD, PhD Neurologist Department of Pediatric Neurology University Utrecht Medical Center Brain Center Utrecht, The Netherlands

xx

Prasanna Jayakar, MD, PhD Adjunct Professor Chairman, Nicklaus Children’s Brain Institute Nicklaus Children’s Hospital Florida International University Miami, Florida, USA Michael V. Johnston, MD Professor Department of Neurology and Pediatrics; Director Neuroscience Laboratory Kennedy Krieger Institute Johns Hopkins University School of Medicine Baltimore, Maryland, USA Shilpa D. Kadam, PhD Assistant Professor Department of Neurology Hugo W. Moser Research Institute at Kennedy Krieger   Institute Johns Hopkins University School of Medicine Baltimore, Maryland, USA Kota Kagawa, MD Division of Neurology Department of Pediatrics University of Toronto The Hospital for Sick Children Toronto, Ontario, Canada; Department of Neurosurgery Hiroshima University Hospital Hiroshima, Japan Ahmet Hilmi Kaya, MD Professor Department of Neurosurgery Yeditepe University School of Medicine Kadikoy, Istanbul, Turkey Bobby P. C. Koeleman, PhD Associate Professor Department of Genetics Center for Molecular Medicine University Medical Center Utrecht, The Netherlands Ajay Kumar, MD, PhD, DNB Assistant Professor Department of Pediatrics, Neurology and Radiology Wayne State University School of Medicine PET Center Children’s Hospital of Michigan Detroit, Michigan, USA Shang-Yeong Kwan, MD Neurologist Section of Epilepsy Neurological Institute Taipei Veterans General Hospital Taipei, Taiwan

Contributors Sandi Lam, MD Associate Professor Department of Neurosurgery Baylor College of Medicine Texas Children’s Hospital Houston, Texas, USA

Mohit Maheshwari, MD Associate Professor of Radiology Department of Radiology Children’s Hospital of Wisconsin Medical College of Wisconsin Milwaukee, Wisconsin, USA

Sebastien Lebon, MD Pediatric Neurologist Department of Pediatrics Unit of Pediatric Neurology and Neurorehabilitation University Hospital of Lausanne Lausanne, Switzerland

Neena I. Marupudi, MD, MS Assistant Professor Department of Neurosurgery Wayne State University School of Medicine Children’s Hospital of Michigan-Detroit Medical Center Detroit, Michigan, USA

Gregory P. Lee, PhD Professor Department of Clinical Neuropsychology Barrow Neurological Institute Phoenix, Arizona, USA

Luca Massimi, MD Assistant Professor Department of Neurosurgery Catholic University School of Medicine Rome, Italy

Mikael Levy, MD Neurosurgeon Department of Neurosurgery Rabin Medical Center Petah Tikva, Israel

Gary W. Mathern, MD Professor Departments of Neurosurgery, Psychiatry, and ­   BioBehavioral Medicine University of California David Geffen School of Medicine Los Angeles, California, USA

Frédérique Liégeois, PhD Lecturer Developmental Cognitive Neuroscience Unit UCL Institute of Child Health Holborn, London, England David D. Limbrick Jr., MD, PhD Professor and Neurosurgeon Department of Neurosurgery Washington University School of Medicine St. Louis Children’s Hospital St. Louis, Mississippi, USA Tobias Loddenkemper, MD Associate Professor Department of Neurology Harvard Medical School Boston Children’s Hospital Boston, Massachusetts, USA David W. Loring, PhD Professor Department of Neurology; Director of Neuropsychology Department of Neurology Emory University Atlanta, Georgia, USA Joseph R. Madsen, MD Associate Professor Department of Neurosurgery Harvard Medical School Boston Children’s Hospital Boston, Massachusetts, USA

Robert A. McGovern, MD Assistant Professor Department of Neurosurgery University of Minnesota Medical School Minneapolis Veterans Affairs Health Care System Minneapolis, Minnesota, USA Michael L. McManus, MD, MPH, FAAP Associate Professor Department of Anesthesia Harvard Medical School Children’s Hospital Boston Boston, Massachusetts, USA Antonio Meola, MD, PhD Clinical Assistant Professor Department of Neurosurgery Stanford University Medical Center and Medical School Stanford, California, USA Mahmoud Messerer, MD Staff Neurosurgeon Department of Neurosurgery University Hospital of Lausanne Lausanne, Switzerland Ian Miller, MD Director of Epilepsy and Neurophysiology Program Department of Neurology Brain Institute Nicklaus Children’s Hospital Miami, Florida, USA

xxi

Contributors John M. K. Mislow, MD, PhD Neurosurgery Resident Department of Neurosurgery Brigham and Women’s Hospital Boston, Massachusetts, USA

†

Sandeep Mittal, MD, FRCSC, FACS Professor and Chair Department of Neurosurgery; Wayne State University School of Medicine Detroit Medical Center and Karmanos Cancer Institute Detroit, Michigan, USA Imad M. Najm, MD Director Department of Neurology Epilepsy Center Neurological Institute Cleveland Clinic Cleveland, Ohio, USA Ika Noviawaty, MD Assistant Professor Department of Neurology University of Massachusetts Medical School and   Medical Center Worcester, Massachusetts, USA Maximiliano Nunez, MD Neurosurgeon Department of Neurosurgery El Cruce Hospital Buenos Aires, Argentina Hirokazu Oguni, MD Professor Department of Pediatrics Tokyo Women’s Medical University Shinjuku-ku, Tokyo, Japan Hiroshi Otsubo, MD Associate Professor Department of Pediatrics; Director of Neurophysiology Lab Department of Neurology University of Toronto The Hospital for Sick Children Toronto, Ontario, Canada Sandip S. Panesar, MD, MSc Postdoctoral Fellow Department of Neurological Surgery Stanford University Medical Center Stanford, California, USA Eun-Hyoung Park, PhD Instructor Department of Neurosurgery Boston Children’s Hospital Boston, Massachusetts, USA

xxii

Daxa M. Patel, MD Pediatric Neurosurgeon Joe DiMaggio Children’s Hospital Hollywood, Florida, USA Jurriaan M. Peters, MD, PhD Assistant Professor of Neurology Department of Neurology Division of Epilepsy and Clinical Neurophysiology Harvard Medical School Boston Children’s Hospital Boston, Massachusetts, USA Julie G. Pilitsis, MD PhD Chair and Professor of Neurosurgery, Neuroscience,   and Experimental Therapeutics Department of Neuroscience and Experimental   Therapeutics Albany Medical College Albany Medical Center Albany, NY, USA Sanjay P. Prabhu, MBBS, FRCR Assistant Professor Department of Radiology Harvard Medical School Boston Children’s Hospital Boston, Massachusetts, USA Etienne Pralong, MD Staff Neurologist Department of Neurosurgery University Hospital of Lausanne Lausanne, Switzerland Klajdi Puka, HBSc PhD Candidate Department of Epidemiology and Biostatistics Western University London Ontario, Canada Jeffrey S. Raskin, MD Assistant Professor Department of Neurological Surgery Indiana University School of Medicine Indianapolis, Indiana, USA Charles Raybaud, MD, FRCPC Professor Department of Neuroradiology University of Toronto The Hospital for Sick Children Toronto, Ontario, Canada Jean Régis, MD Professor and Director Department of Stereotactic and Functional Neurosurgery Aix-Marseille University Timone University Hospital Marseille, France

Contributors James J. Riviello Jr., MD Professor Department of Pediatrics; Department of Neurology Baylor College of Medicine; Associate Section Head Department of Epilepsy, Neurophysiology,   and Neurocritical Care Section of Neurology and Developmental  Neuroscience Texas Children’s Hospital Houston, Texas, USA

Maurits W. C. B. Sanders, MD Department of Child Neurology Brain Center Rudolf Magnus University Medical Center Utrecht Utrecht, The Netherlands

Michele Rizzi, MD Neurosurgeon and Research Assistant Department of Neurosciences “C. Munari” Epilepsy Surgery Centre Azienda Ospedale Niguarda Ca’ Granda Milan, Italy

Margitta Seeck, MD, PhD Professor of Neurology Epilepsy Unit Department of Clinical Neurosciences University Hospital of Geneva Geneva, Switzerland

Concezio Di Rocco, MD Professor and Director Pediatric Neurosurgery International Neuroscience Institute Hannover, Germany

Hiroyuki Shimizu, MD Professor Emeritus Shimizu Clinic Kohenjiminami, Suginamiku, Japan

Jarod L. Roland, MD Assistant Professor Department of Neurological Surgery University of California San Francisco UCSF Benioff Children’s Hospital San Francisco, California, USA

Johannes Schramm, MD Professor Department of Neurosurgery University of Bonn Bonn University Medical Center Bonn, Germany

Nir Shimony, MD Pediatric Neurosurgeon Department of Neurosurgery Geisinger Commonwealth School of Medicine   and Medical Center Dannville, Pennsylvania, USA

John D. Rolston, MD, PhD Director of Functional Neurosurgery Department of Neurosurgery University of Utah Salt Lake City, Utah, USA

Thomas W. Smith, MD Professor of Pathology and Neurology Department of Pathology University of Massachusetts Medical School and   Medical Center Worcester, Massachusetts, USA

James T. Rutka, MD, PhD, FRCSC, FACS, FAAP, FAANS RS McLaughlin Professor and Chair Department of Neurosurgery and Surgery Hospital for Sick Children University of Toronto Toronto, Ontario, Canada

Mary Lou Smith, PhD Professor Department of Psychology Hospital for Sick Children University of Toronto Toronto, Ontario, Canada

Mustafa Şahin, MD, PhD Professor of Neurology Department of Neurology Harvard Medical School Boston Children’s Hospital Boston, Massachusetts, USA

Matthew D. Smyth, MD Appoline Blair Professor of Neurological Surgery and  Pediatrics Department of Neurosurgery Washington University School of Medicine St. Louis Children’s Hospital St. Louis, Mississippi, USA

Francesco Sala, MD Professor of Neurosurgery Department of Pediatric Neurosurgery Institute of Neurosurgery University Hospital University of Verona Verona, Italy

O. Carter Snead III, MD Professor Department of Pediatrics and Medicine (Neurology); Head, Division of Neurology; University of Toronto The Hospital for Sick Children Toronto, Ontario, Canada

xxiii

Contributors Sulpicio G. Soriano, MD Professor Department of Anesthesiology Harvard Medical School BCH Endowed Chair in Pediatric Neuroanesthesia Boston Children’s Hospital Boston, Massachusetts, USA Carl E. Stafstrom, MD, PhD Professor Department of Pediatric Neurology Johns Hopkins University School of Medicine Baltimore, Maryland, USA James L. Stone, MD Professor Department of Neurosurgery and Neurology New York University Langone Medical Center New York, USA Scellig S. Stone, MD, PhD, FRCSC Assistant Professor Department of Neurosurgery Harvard Medical School Boston Children’s Hospital Boston, Massachusetts, USA Samuel Strantzas MSc, DABNM, REP Associate Clinical Neurophysiologist Division of Neurosurgery University of Toronto The Hospital for Sick Children Toronto, Ontario, Canada Heri Subianto, MD Fellow in Epilepsy and Functional Neurosurgery Department of Neurosurgery All India Institute of Medical Sciences AIIMS National Brain Research Center New Delhi, India Michael Tan, MBChB, FANZCA Pediatric Anesthesiologist and Honorary Lecturer Department of Pediatric Anesthesiology Starship Children’s Hospital and Department of  Anesthesiology University of Auckland Grafton, Auckland, New Zealand Zulma S. Tovar-Spinoza, MD Associate Professor Department of Neurosurgery and Pediatrics; Director Pediatric Neurosurgery SUNY Upstate Medical University Syracuse, New York, USA Manjari Tripathi, MD Professor Department of Neurology All India Institute of Medical Sciences (AIIMS) AIIMS National Brain Research Center New Delhi, India

xxiv

Min-Lan Tsai, MD Pediatric Neurologist Department of Neurosurgery Taipei Medical University Taipei, Taiwan Uğur Türe, MD Professor and Chairman Department of Neurosurgery Yeditepe University School of Medicine Kadikoy, Istanbul, Turkey Ingrid Tuxhorn, MD Professor Department of Pediatric Neurology Case Western Reserve University Pediatric Epilepsy Rainbow Babies and Children’s   Hospital Lakewood, Ohio, USA Prashin C. Unadkat, MBBS Postdoctoral Research Fellow Surgical Planning Laboratory and Golby Laboratory Brigham and Women’s Hospital Boston, Massachusetts, USA Rafael Uribe, MD, MHS Neurosurgery Resident Department of Neurosurgery Weill Cornell Medicine New York-Presbyterian Hospital/Weill Cornell   Medical Center New York, USA Ahsan Moosa Naduvil Valappil, MD Staff Neurologist Section of Pediatric Epilepsy Epilepsy Center Neurological Institute Cleveland Clinic Cleveland, Ohio, USA Doris D. Wang, MD, PhD Clinical Fellow in Stereotactic and Functional  Neurosurgery Department of Neurological Surgery University of California San Francisco San Francisco, California, USA Alexander G. Weil, MD, FRCSC, FAANS, FACS Assistant Professor Division of Pediatric Neurosurgery Sainte Justine University Hospital University of Montreal Montreal, Quebec, Canada Howard L. Weiner, MD, FACS, FAAP, FAANS Professor and Vice-Chairman Department of Neurosurgery Baylor College of Medicine; Chief of Neurosurgery Texas Children’s Hospital Houston, Texas, USA

Contributors  Nicholas M. Wetjen, MD Neurosurgeon Department of Neurological and Spinal Surgery The Iowa Clinic Des Moines, Iowa, USA Elysa Widjaja, MRCP, MD, FRCR Associate Professor Department of Neuroradiology University of Toronto The Hospital for Sick Children Toronto, Ontario, Canada Tai-Tong Wong, MD Professor Chief Department of Neurosurgery Taipei Medical University Taipei, Taiwan

Edward E. Woodward, Medical Student University of South Florida Morsani College of Medicine Tampa, Florida, USA Elaine Wyllie, MD Professor The Cleveland Clinic Lerner College of Medicine; Staff Neurologist Epilepsy Center Neurological Institute Cleveland Clinic Cleveland, Ohio, USA Fang-Cheng Yeh, MD, PhD Assistant Professor Department of Neurological Surgery University of Pittsburgh Medical Center Pittsburgh, Pennsylvania, USA

xxv

Part I Introduction to Epilepsy in Children

I

1  Basic Considerations of Pediatric Epilepsy Surgery 

03

2  A Historical Review of Epilepsy Surgery and Its Application in Children

10

3  Epidemiology of Intractable Epilepsy in Children

21

4  Genetics in Epilepsy Surgery

28

5  Surgical Neuropathology of Pediatric Epilepsy

36

6   Epilepsy and Brain Plasticity

47

7  Effects of Seizures and Their Comorbidities on the Developing Brain

53

8  Ethical Considerations in Pediatric Epilepsy Surgery

57

9  Infantile and Childhood-Onset Catastrophic Epilepsy Syndromes

61

10 Epilepsy Surgery for Congenital or Early Brain Lesions

70

1

Basic Considerations of Pediatric Epilepsy Surgery Oguz Çataltepe and George I . Jallo

Introduction Epilepsy is one of the most common neurological disorders in children. The incidence of pediatric epilepsy is reported to be 33 to 82 cases per 100,000 per year in developed countries.1-3 Almost half of the children with epilepsy have high rates of learning disabilities, mental retardation , developmental delay, psychiatric / behavioral difficulties , and psychosocial problems. Therefore, preventing cognitive and developmental stagnation or decline in this patient population is as important as achieving seizure freedom. Approximately, 20% of the children with epilepsy continue to suffer from medically intractable seizures and surgery is often the single remaining treatment option for them, not only to control seizures but also to prevent and improve comorbid conditions.2 4-9 Comprehensive care of epilepsy in children is challenging and needs specialized knowledge and expertise in both the medical and surgical management of this condition. Well-coordinated , collaborative relationships between the medical and surgical teams in a multidisciplinary environment are one of the key factors for successful management of pediatric epilepsy patients. The number of centers offering surgical management for childhood epilepsy has increased remarkably within the last two decades and epilepsy surgery has become an established treatment option worldwide. We have also witnessed an increased number of publications on pediatric epilepsy surgery related topics by dedicated working groups and the subcommissions of some international organizations in recent years.510-12

-

Historical Evolution Although surgical intervention in epilepsy patients has a relatively long history, many decades elapsed before epilepsy surgery became an established treatment option for adult epilepsy patients. Pediatric epilepsy surgery has followed an even more hesitant course. It was performed in only a handful of centers and remained a last resort option in the management of children with epilepsy. For a long time, even children with intractable focal epilepsy were rarely referred to the specialized epilepsy surgery centers. Historical reluctance to perform pediatric epilepsy surgery stemmed from several legitimate concerns and

limitations such as unavailability of epilepsy surgery centers in many locations, limitation of diagnostic technics , poor surgical results, and general lack of knowledge or misinformation about pediatric epilepsy surgery. Limited data regarding the long- term effect of epilepsy surgery on children as well as the belief that many childhood seizures were benign and had favorable outcomes also contributed to this reluctance toward pediatric epilepsy surgery.2 71013 As a result , it took almost two decades after the publication of initial pediatric epilepsy surgery series to see an increase in the number of centers offering surgical treatment for childhood epilepsy.14-17 Although the initial surgical series included mostly older children , age of the patients gradually decreased and even reports about infantile epilepsy surgery series started to be seen in medical journals.6-9,12,18,19 The accumulated data and reported results from surgical series were encouraging and clearly showed that surgical intervention in children with intractable epilepsy dramatically improved outcomes, and provided not only seizure reduction and freedom but also improved behavioral , language, and cognitive functions as well as quality of life.20-22 Thus, in a relatively short period of time, pediatric epilepsy surgery procedures were transformed from a rarely performed interventions into an established management option for children with intractable epilepsy. These developments have led to worldwide establishment of many dedicated pediatric epilepsy surgery programs. As a result, pediatric epilepsy surgery became a subspecialty interest for many pediatric neurosurgeons.

Epilepsy Surgery in Children The main principles of epilepsy surgery from preoperative assessment to the surgical techniques were primarily developed for adult epilepsy patients. These principles were later extrapolated and applied to pediatric epilepsy patients.23 Although there are many common features in the surgical management of pediatric and adult epilepsy patients, there are also many differences. Understanding these differences is of critical importance in the management of pediatric epilepsy patients.5 24 Pediatric epilepsy patients have very diverse epileptic disorders and some

-

4

I  Introduction to Epilepsy in Children of them exhibit d ­ ifferent, at times unique, electrophysiological and semiological characteristics. They are also more vulnerable to neurodevelopmental and psychosocial problems that must be taken into account when considering surgery. Also, having expertise in perioperative and neurosurgical techniques in the management of children, especially infants, is critical in the surgical management of pediatric epilepsy patients. Among pediatric epilepsy patients, infants and young children constitute an extremely challenging subset. Awareness of age-related characteristics as well as special paradigms and issues related to the management of this patient group is an important prerequisite for ensuring good surgical outcomes. Surgical management of medically intractable seizures in children is characterized by a number of uniquely challenging problems and requires a special experience and expertise in many areas such as preoperative assessment, selecting the surgical candidates, and defining the most appropriate surgical strategy for each patient. A subcommission for pediatric epilepsy surgery was formed by the International League Against Epilepsy (ILAE) in 1998 to meet all these challenges and to formulate appropriate standards for epilepsy surgery in childhood. In 2003, the subcommission organized a meeting to address the following questions:5 1. Are the unique characteristics of children with epilepsy and their syndromes sufficiently different to justify dedicated pediatric epilepsy centers? 2. Is adequate information available to propose guidelines regarding patient selection and surgical treatment for pediatric epilepsy surgery patients? The subcommission agreed that the “neurobiological aspects of epilepsy are unique to children, especially the young, and as such require specific pediatric epilepsy expertise. Collectively these features justify the unique approach necessary for dedicated pediatric epilepsy surgery centers.”5

„„ Unique Characteristics of and Special Considerations for Children Although many aspects of pediatric epilepsy and its surgical management are similar to those in adults, there are significant differences and characteristics unique to children, especially infants and young children. These differences might be critical in many stages throughout the management of children with epilepsy, starting from preoperative assessment and extending to the surgical intervention itself. We will briefly review these areas to provide a general perspective on the subject.

Age Group While age group is not a significant topic in the treatment of adult epilepsy patients, it dominates the ­ discussion in many areas of pediatric epilepsy surgery, including

­reoperative assessment, surgical indications, timing p of the surgery, surgical approaches, and outcome measures. Rapid brain m ­ aturation during early infancy and childhood is responsible for complex evolution of clinical symptoms, seizure semiology, electroencephalography (EEG), and neuroimaging findings.5,​24 This complexity makes the assessment and interpretation of the clinical, electrophysiological, and imaging findings very challenging. Naturally, lower levels of cognitive maturation and language development in children as well as their cooperation during studies can be an obstacle or limiting factor for certain preoperative tests, such as the Wada or functional MRI. Even structural MRI findings in very young children, such as young infants, may be problematic because of incomplete myelinization and repeated images with certain intervals might be needed in young children. Also, harmful and sometimes catastrophic effects of seizures on the developing brain of young children are a major concern, and constitute strong indication for early surgical intervention in some infants. Again, young age might be a limiting factor in the utilization of some surgical techniques in children, such as cortical stimulation and mapping under local anesthesia as well as depth electrode placement for stereoelectroencephalography (SEEG).

Epileptic Syndromes and Pathologies Specific to Children The causes of seizures in children amenable to surgery are quite diverse (Table 1.1): perinatal injuries, cortical dysplasia, developmental brain tumors such as dysembroplastic neuroepithelial tumor (DNET) and ganglioglioma, rare lesions and conditions such as hypothalamic hamartoma, Rasmussen’s encephalitis, Sturge–Weber syndrome, etc.1,​4,​25 Some of these are exclusively seen in children and rarely or never occur in adults. Furthermore, each of these disorders is seen at a differing frequency based on the age and they have unique diagnostic and surgical challenges. Cortical malformations are the most common neuropathological substrate in children (23–78%) and according to an ILAE survey cortical malformations constitute 60% of pathologies in Table 1.1  Etiology/substrates in pediatric epilepsy surgery patients (2004 ILAE survey) Etiology/subtrate

Frequency (%)

Cortical dysplasia

42.4

Tumor

19.1

Atrophy/stroke

9.9

Hippocampal sclerosis

6.5

Gliosis

6.3

Tuberous sclerosis

5.1

Hypothalamic hamartoma

3.6

Rasmussen’s encephalitis

2.7

Vascular

1.5

Source: Harvey AS, Cross JH, Shinnar S, the Paediatric Epilepsy Surgery Survey Taskforce, et al. Defining the spectrum of ­international practice in paediatric epilepsy surgery patients. Epilepsia. 2008;49:146–155.

1  Basic Considerations of Pediatric Epilepsy Surgery Table 1.2  Age distribution of etiology/substrates in pediatric epilepsy surgery patients (2004 ILAE survey) Etiology/substrate

Age 0–4 (%)

Age 4–8 (%)

Age 8–12 (%)

Age 12–18 (%)

Cortical dysplasia

60

45

32

32

Tumor

10

20

24

25

Atrophy/stroke

7

8

14

12

Hippocampal sclerosis

1

5

9

12

Gliosis/normal

5

5

8

8

Source: Harvey AS, Cross JH, Shinnar S, the Paediatric Epilepsy Surgery Survey Taskforce, et al. Defining the spectrum of international practice in paediatric epilepsy surgery patients. Epilepsia. 2008;49:146–155.

pediatric epilepsy surgery cases operated on younger than 4 years age (Table 1.2).25 Again, developmental brain tumors, such as ganglioglioma and DNET, are much more common in children and they frequently associate with adjacent areas of focal cortical dysplasia in a range of 20 to 50%.26 Although mesial temporal sclerosis (MTS) in children is seen much less frequently (6.5–38%) than adults, dual pathologies associated with MTS occur much more frequently in children.1,​4–5,​8,​25

Semiology and Electrophysiological Characteristics In childhood, intractable seizures can be quite atypical and poorly defined compared to relatively well-defined clinical and electrophysiological characteristics of epilepsy syndromes in adults. The clinical and electrophysiological presentation of intractable, localization-related epilepsy might be heterogeneous and wide-ranging in childhood. Both unilateral focal and hemispheric lesions in children may present with generalized clinical seizure manifestations and EEG patterns.5,​24 Some of them present as evolving, age-dependent electroclinical syndromes such as the cases with infantile epileptic encephalopathy that evolve to West syndrome and then Lennox–Gastaut syndrome.27 Again some complex early-onset epilepsy syndromes such as West syndrome and Lennox–Gastaut syndrome manifest themselves as bilateral or generalized epileptiform discharges on EEG despite originating from focal cortical lesions such as focal cortical dysplasia or cortical tubers, and they respond to focal resective surgeries very well.28–31 The immaturity of brain and cortical pathways plays a central role in this atypical clinical and electrophysiological expression of seizures. Electrophysiological evaluation of cortical activity in infants and young children might be extremely difficult because of poorly defined “normal” and “abnormal” EEG patterns of the immature brain, the absence of well-defined epileptiform discharges, rapidly spreading ictal activity, and great variability of electrophysiological seizure patterns. These characteristics make the localizing value of EEG findings for children very controversial compared to adults. As a result, defining the epileptogenic zone in the immature brain is a daunting task in many cases and needs to be handled with a great deal of expertise.

The Effect of Seizures on the Developing Brain The cumulative harmful effects of frequent seizures on the developing brain can be catastrophic. In addition to frequent clinical seizures, continuous postictal state and frequent interictal epileptiform discharges may cause an irritable, dysfunctional cortex and, in some cases, secondary epileptogenesis. Consequently, intractable seizures and ­ their ­deleterious effect on the developing brain may result in cognitive decline, debilitating behavioral problems, aggression, attention deficit disorder, and hyperactivity.32 Although spontaneous remission of seizures is possible, the risk of permanent neurologic, psychosocial, and cognitive impairment and the impact of the adverse effects of antiepileptic drugs (AEDs) during this crucial period of brain development is significant. Early-onset epilepsy and long duration of seizures are associated with poor neurodevelopmental outcomes.33 Successful seizure control may facilitate cognitive development and help to reduce the behavioral and/or psychological burden of epilepsy on the child and family. Although this has yet to be proven, there are studies showing improved quality of life and psychosocial outcome as well as enhanced cognitive development following earlier surgical treatment and seizure cessation.5,​21–22 Therefore, the seizure control should not be the only consideration when making decisions regarding epilepsy surgery in children. Potentially harmful effects of seizures on the developing brain as well as the chance of neurodevelopmental improvement after a successful surgical intervention should be also considered.

Functional Plasticity There is a significant amount of accumulated data regarding the functional plasticity of the young brain based on experimental animal studies, observation of epilepsy patients following surgical resection, and from patients with cerebral insults at a young age. This data shows that young children have a much greater potential for recovery and a significant capacity for the reorganization of neurologic functions following cerebral injuries, including surgery.13,​20,​34–36 This is especially evident in the recovery and relocation of speech-related functions in children younger than 5 years of age.36 The functional plasticity of the

5

6

I  Introduction to Epilepsy in Children young brain also makes children more vulnerable to the deleterious effects of repeating seizures, which can result in deviant or delayed development and trigger permanent changes in developing neural circuitry. Therefore, it is of the utmost importance that pediatric epilepsy surgery teams appreciate the functional plasticity and recovery potential of the young brain and take these characteristics into consideration when making surgical decisions.5

Medical Intractability The medical intractability criterion in children is also different than those in adults. Earlier identification of medical intractability in pediatric epilepsy patients, as opposed to adult epilepsy patients, is feasible in many cases because certain pediatric epilepsy syndromes or seizure etiologies imply intractability by their very own nature, and these children do not need long trial periods with various major AEDs.11 The developing brain is also much more vulnerable to the adverse effects of AEDs, making the risk/benefit assessment for long-term trials of AEDs more critical in children than in adults. On the other hand, epilepsy in childhood is often not a fixed condition, and while it may evolve toward intractability in some cases, it may remit or stop spontaneously in others. Therefore, while decisions regarding medical intractability can be made easily and quickly in some well-defined pediatric epilepsy syndromes, in other cases, great caution must be exercised before deciding if the patient is indeed a surgical candidate.

Timing of Surgery Timing of surgery is another unique aspect of pediatric epilepsy. As discussed earlier, the cumulative harmful effects of epilepsy on early brain development is a major concern in the treatment of pediatric epilepsy patients. Although consensus is still lacking, there is an increasing amount of data in the literature supporting the benefits of early surgical interventions in catastrophic epilepsy. Because certain pediatric epilepsy syndromes are inherently ­ medically refractory, there is no need to “prove” medical intractability before embarking on a surgical course of action in these c­ ases. The harmful effects of prolonged seizures and AEDs on synaptogenesis, brain development, and cognitive/­ psychosocial development support the argument for ­early surgery in pediatric epilepsy patients. Even if clinical ­seizures are ­successfully controlled with medical treatment, frequent interictal discharges can still cause changes in synaptogenesis and cytoarchitecture in immature brains and may create a secondary epileptogenic focus. The potential for recovery from postoperative deficits is highest during the period of high synaptic and dendritic density (ages of 3–7 years), when the plasticity of the brain peaks.18 Surgery performed within this time frame may help to hasten recovery, and anticipated postoperative impairments might be milder. In well-selected patients, early surgical intervention may prevent the negative cognitive, psychosocial, and developmental effects of

seizures.2,​5 Thus, early surgical intervention helps children develop without further psychosocial harm and, in many cases, can maximize their developmental potential.32 Nevertheless, there are also significant concerns regarding early surgical intervention in pediatric epilepsy patients. The possibility of spontaneous remission is one of the main arguments against early surgical intervention since some childhood seizure disorders spontaneously remit in adulthood. Other concerns voiced against early surgical intervention include the possibility of eventually achieving seizure control with AEDs, as well as the morbidity and mortality risks associated with surgical intervention in infants and young children.

Goals of Surgery The goals of epilepsy surgery in children are somewhat different from those for adults because of the critical significance of harmful effects of seizures on the developing brain. Therefore, the goals of pediatric epilepsy surgery are not only to control the seizures, but also to prevent the possible harmful consequences of uncontrolled seizures on the immature brain; to control continued interictal activity resulting in permanent cognitive, behavioral, and psychosocial problems; to prevent secondary epileptogenesis; and to avoid the adverse effects of AEDs on the developing brain.32 However, despite the general acceptance of the potential benefits of seizure control on the cognitive, behavioral, and psychological development of the child, definitive data on this matter are still pending. Therefore, the central goal of pediatric epilepsy surgery remains limited to the attainment of seizure freedom.5

Surgical Procedures The type and frequency of commonly performed surgical procedures for pediatric epilepsy patients are different to a certain degree than those in adults (Table 1.3). Seizures in young children are frequently e ­ xtratemporal and cover large, multilobar cortical areas ­including the eloquent cortex. Therefore, invasive monitoring, cortical mapping, and stimulation studies may be needed more f­ requently in

Table 1.3  Common surgical procedures in pediatric epilepsy surgery patients (2004 ILAE survey) Type of surgery Lobar/focal resections

Frequency (%) 48

Temporal

23.2

Frontal

17.5

Multilobar resections

12.9

Hemispherectomy

15.8

Vagus nerve stimulator

15.8

Corpus callosotomy

3.1

Source: Harvey AS, Cross JH, Shinnar S, the Paediatric Epilepsy Surgery Survey Taskforce, et al. Defining the spectrum of international practice in paediatric epilepsy surgery patients. Epilepsia. 2008;49:146–155.

1  Basic Considerations of Pediatric Epilepsy Surgery Table 1.4  Age characteristics of pediatric epilepsy surgery patients (years) (2004 ILAE Survey) Type of surgery

Age 0–4 (%)

Age 4–8 (%)

Age 8–12 (%)

Age 12–18 (%)

Lobar/focal resections

35

47

49

60

Multilobar resections

20

11

12

10

Hemispherectomy

32

15

10

8

9

25

23

20

Palliative (VNS, corp call)

Source: Harvey AS, Cross JH, Shinnar S, the Paediatric Epilepsy Surgery Survey Taskforce, et al. Defining the spectrum of international practice in paediatric epilepsy surgery patients. Epilepsia. 2008;49:146–155.

children than in adults and their ­management requires a great deal of medical and surgical expertise. Some surgical procedures, such as hemispheric or multilobar resections, disconnections, or multistage resections, are performed much more commonly in pediatric epilepsy patients than in adults. Surgical procedures for hemispheric syndromes are frequently performed in young children, even in infants, but are very rarely performed in adults, and these procedures comprise a large part of the surgical interventions performed at some pediatric epilepsy surgery ­centers. The type of the procedures performed in children with epilepsy also changes significantly based on pediatric age group (Table 1.4). Hemispheric and multilobar procedures constitute 52% of pediatric epilepsy surgery procedures performed in children younger than 4 years of age.25 These are challenging and complex procedures with a higher risk of perioperative complications than any other epilepsy surgery procedure, and require considerable expertise. On the other hand, focal and lobar resections are much commonly performed in older children and constitute 60% of the procedures in children older than 12 years of age.25 Some epileptic lesions are seen much more commonly in pediatric age groups and surgical procedures to treat them are performed much more commonly during childhood. Surgical interventions for hypothalamic hamartoma and multistage cortical resection in patients with tuberous sclerosis are just two examples of this group of lesions. Sturge–Weber syndrome is another condition seen in children and affected patients may require urgent attention for multilobar or hemispheric disconnection procedures because of the syndrome’s potentially deleterious effects, such as developmental delays and progressive hemiparesis. Rasmussen syndrome and Landau–Kleffner syndrome also appear mainly in childhood, and their management requires considerable medical and surgical expertise.5,​24,​37

„„ Present Status and Future Considerations Many advances in structural/functional neuroimaging, EEG/video monitoring technology, invasive diagnostic techniques, perioperative care, and surgical technology have occurred within the last two decades and revolutionized the practice of pediatric epilepsy surgery. The number of pediatric epilepsy surgery centers as well as

the number of young epilepsy patients undergoing epilepsy surgery has exponentially increased within the last decade. The age distribution of these patients has drastically changed and epilepsy surgery in infants has become much more common than before. Much sophisticated noninvasive and invasive neurophysiologic data acquisition techniques have been developed and the availability of outcome data for pediatric epilepsy surgery patients has significantly increased within the last decade. Previously unimaginable improvements in preoperative planning technology, such as three-­dimensional reconstruction and image fusion techniques for multimodality imaging studies, and new tools and techniques in the modern neurosurgical armamentarium, such as neuronavigation systems, intraoperative MRI, SEEG, laser ablation techniques, endoscopic techniques, etc., have enabled neurosurgeons to perform less invasive and more precise interventions in epilepsy patients. Advances in the neurosurgical technology and techniques as well as improving surgical skills and experience of epilepsy surgeons have resulted in increasingly sophisticated surgical procedures. In addition, remarkable improvements in pediatric neuroanesthesiology and pediatric ICU care have had a huge impact on the surgical outcomes of pediatric epilepsy patients. Furthermore, refined pre- and postoperative neuropsychological assessment techniques and improved data accumulation methods have also provided valuable insights into the effect of current surgical interventions on the ­various life domains of the pediatric epilepsy patient. With the advent of all these areas, the spectrum of surgical candidates among children with epilepsy, including MRI-negative cases, has been broadened significantly. These developments have opened a unique window of opportunity for pediatric epilepsy surgery. The number of pediatric epilepsy surgery centers and children undergoing epilepsy surgery has increased dramatically over the past 10 years. The involvement of pediatric neurosurgeons in pediatric epilepsy cases has become a standard practice. As a result, pediatric epilepsy surgery has become an established, safe, and efficacious treatment modality in carefully selected children. However, pediatric epilepsy surgery still faces many hurdles, such as the lack of a consensus for the identification and selection criteria of surgical candidates as well as the lack of guidelines for determining the proper timing of surgery. Although surgical techniques are much more refined and safer than ever before, many new and potentially beneficial areas in pediatric epilepsy surgery

7

8

I  Introduction to Epilepsy in Children remain open to exploration and development, including new ­ neuromodulation procedures, such as deep brain stimulation and the application of ablation and radiosurgery techniques in children. Data from these procedures and similar treatment modalities are still limited or yet to be gathered.

„„ Conclusion Although pediatric epilepsy surgery is a well-established management option in the treatment of this highly vulnerable patient population, the accumulated data are still far from satisfactory in terms of provid-

ing well-defined guidelines and parameters. Because children, especially young children and infants, are still developing human beings, epilepsy is not fixed but is an evolving and complex process in this patient population. Therefore, the selection and referral of young patients for epilepsy surgery constitute a delicate endeavor, one that needs to be handled with great care and expertise. It is necessary to maintain an exquisite balance between avoiding unnecessary surgery and inadvertently causing a patient to experience psychosocial deterioration or the adverse effects from AEDs because of unrealistic expectations of a spontaneous remission. This is the unique challenge that the pediatric epilepsy surgery community now faces and must overcome.

References 1. Guerrini R. Epilepsy in children. Lancet 2006;367(9509):499–524 2. Cross JH. Epilepsy surgery in childhood. Epilepsia 2002;43(Suppl 3): 65–70 3. Camfield P, Camfield C. Incidence, prevalence and ­aetiology of ­seizures and epilepsy in children. Epileptic Disord. 2015;17(2): 117–232 4. Spencer S, Huh L. Outcomes of epilepsy surgery in adults and children. Lancet Neurol 2008;7(6):525–537 5. Cross JH, Jayakar P, Nordli D, et al; International League Against Epilepsy, Subcommission for Paediatric Epilepsy Surgery, Commissions of Neurosurgery and Paediatrics. Proposed criteria for referral and evaluation of children for epilepsy surgery: recommendations of the Subcommission for Paediatric Epilepsy Surgery. Epilepsia 2006;47:952–959 6. Wyllie E. Catastrophic epilepsy in infants and children: identification of surgical candidates. Epileptic Disord 1999;1(4):261–264 7. Duchowny M. Epilepsy surgery in children. Curr Opin Neurol 1995;8(2):112–116 8. Wyllie E, Comair YG, Kotagal P, Bulacio J, Bingaman W, Ruggieri P. Seizure outcome after epilepsy surgery in children and adolescents. Ann Neurol 1998;44(5):740–748 9. Dunkley C, Kung J, Scott RC, et al. Epilepsy surgery in children under 3 years. Epilepsy Res 2011;93(2–3):96–106 10. Harvey AS, Cross JH, Shinnar S, Mathern GW; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epilepsia 2008;49(1):146–155 11. Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 2010;51(6):1069–1077 12. Wilmshurst JM, Gaillard WD, Vinayan KP, et al. Summary of recommendations for the management of infantile seizures: Task Force Report for the ILAE Commission of Pediatrics. Epilepsia 2015;56(8):1185–1197 13. Stafstrom CE, Lynch M, Sutula TP. Consequences of epilepsy in the developing brain: implications for surgical management. Semin Pediatr Neurol 2000;7(3):147–157

17. Rasmussen T. Surgical aspects. In: Lee RG, ed. Topics in Child Neurology. New York: Spectrum Publications; 1977:143–157 18. Adelson PD. Temporal lobectomy in children with intractable seizures. Pediatr Neurosurg 2001;34(5):268–277 19. Duchowny M, Levin B, Jayakar P, et al. Temporal lobectomy in early childhood. Epilepsia 1992;33(2):298–303 20. Depositario-Cabacar DT, Riviello JJ, Takeoka M. Present status of surgical intervention for children with intractable seizures. Curr Neurol Neurosci Rep 2008;8(2):123–129 21. Skirrow C, Cross JH, Harrison S, et al. Temporal lobe surgery in childhood and neuroanatomical predictors of long-term declarative memory outcome. Brain 2015;138(Pt 1):80–93 22. Skirrow C, Cross JH, Cormack F, Harkness W, Vargha-Khadem F, Baldeweg T. Long-term intellectual outcome after temporal lobe surgery in childhood. Neurology 2011;76(15):1330–1337 23. Obeid M, Wyllie E, Rahi AC, Mikati MA. Approach to pediatric ­epilepsy surgery: state of the art, part I: general principles and presurgical workup. Eur J Paediatr Neurol 2009;13:102–114 24. Shewmon DA, Shields WD, Chugani HT, Peacock WJ. Contrasts between pediatric and adult epilepsy surgery: rationale and strategy for focal resection. J Epilepsy 1990;3(Suppl):141–155 25. Harvey AS, Cross JH, Shinnar S, Mathern GW; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epilepsia 2008;49(1):130–155 26. Holthausen H, Fogarasi A, Arzimanoglou A, Kahane P. Structural (symptomatic) focal epilepsies of childhood. In: Bureau M, Genton P, Dravet C, Delgado-Escueta A, Tassinari CA, Thomas P, Wolf P, eds. Epileptic Syndromes in Infancy Childhood and Adolescence. Paris: John Libbey Eurotext Ltd.; 2012:455–505 27. Ohtahara S, Yamatogi Y. Epileptic encephalopathies in early infancy with suppression-burst. J Clin Neurophysiol 2003;20(6):398–407 28. Asarnow RF, LoPresti C, Guthrie D, et al. Developmental outcomes in children receiving resection surgery for medically intractable infantile spasms. Dev Med Child Neurol 1997;39 (7):430–440

14. Falconer MA. Place of surgery for temporal lobe epilepsy during childhood. BMJ 1972;2(5814):631–635

29. Wyllie E, Lachhwani DK, Gupta A, et al. Successful surgery for epilepsy due to early brain lesions despite generalized EEG findings. Neurology 2007;69(4):389–397

15. Goldring S. A method for surgical management of focal epilepsy, especially as it relates to children. J Neurosurg 1978;49 (3):344–356

30. Peltola ME, Liukkonen E, Granström ML, et al. The effect of surgery in encephalopathy with electrical status epilepticus during sleep. Epilepsia 2011;52(3):602–609

16. Green JR, Pootrakul A. Surgical aspects of the treatment of epilepsy during childhood and adolescence. Ariz Med 1982;39(1):35–38

31. Jadhav T, Cross JH. Surgical approaches to treating epilepsy in children. Curr Treat Options Neurol 2012;14(6):620–629

1  Basic Considerations of Pediatric Epilepsy Surgery 32. Berg AT. Paediatric epilepsy surgery: making the best of a tough situation. Brain 2015;138(Pt 1):4–5 33. Berg AT, Smith SN, Frobish D, et al. Longitudinal assessment of adaptive behavior in infants and young children with newly diagnosed epilepsy: influences of etiology, syndrome, and seizure control. Pediatrics 2004;114(3):645–650 34. Obeid M, Wyllie E, Rahi AC, Mikati MA. Approach to pediatric epilepsy surgery: State of the art, Part II: Approach to ­specific epilepsy syndromes and etiologies. Eur J Paediatr Neurol 2009; 13(2):115–127

35. Benifla M, Sala F Jr, Jane J, et al. Neurosurgical management of intractable rolandic epilepsy in children: role of ­ resection in eloquent cortex. Clinical article. J Neurosurg Pediatr 2009;4(3):199–216 36. Hertz-Pannier L, Chiron C, Jambaqué I, et al. Late plasticity for language in a child’s non-dominant hemisphere: a pre- and post-­ surgery fMRI study. Brain 2002;125(Pt 2):361–372 37. Saneto RP, Wyllie E. Surgically treatable epilepsy syndromes in infancy and childhood. In: Miller JW, Silbergeld DL, eds. Epilepsy Surgery. New York: Taylor and Francis; 2006:121-141

9

A Historical Review of Epilepsy Surgery and Its Application in Children Deniz Çataltepe and Oguz Çataltepe

Summary The history of epilepsy surgery dates back to the late Nineteenth Century, with the first epilepsy surgery procedure performed by Victor Horsley in 1886. Horsley’s initial surgery, as well as his subsequent procedures on epilepsy patients, were a culmination of the collaboration between clinical physi ologists and surgeons at the National Hospital for the Paralyzed and Epileptic in Queen Square, London. Around the same time, Fedor Krause started a school of epilepsy surgery in Germany focusing on cortical stimulation and motor cortex mapping. The International League Against Epilepsy, which was established in 1909, brought together these seminal figures as well as the surgeons William Macewen and Harvey Cushing. Following the First World War, while epilepsy surgery techniques were being propagated through this network , the electroencephalography ( EEG ) was invented and was used for intraoperative recordings , localizing epileptogenic cortical areas, and guiding resections One of the EEG’s champions, the German neurologist and neu rosurgeon Otfrid Foerster, inspired Wilder Penfield ’s work in cortical stimulation techniques beginning in 1928. Penfield would go on to establish the Montreal Neurological Institute ( MNI ), which became a pioneer center in the electrophysiologyguided treatment of the epilepsy patient. Although pediatric patients were involved in these epilepsy surgery milestones, the first surgical series to focus on this specific patient popu lation was completed by Murray A. Falconer in London in 1970, with other significant series’ published by Sidney Goldring in St. Louis between 1978 and 198 The first International Pediatric Epilepsy Surgery Symposium took place in 1989, with meetings and workshops continuing to be held regularly to establish guidelines and recommendations for epilepsy surgery in the pediatric population . The history of epilepsy surgery, which is deeply intertwined with the history of modern neuro surgery, encompasses the early foundations of pediatric epilepsy surgery, and this specialty continues to be refined today.

^

^

Keywords: history, epilepsy surgery, pediatric epilepsy surgery, Horsley, Penfield , Falconer

Introduction The history of modern epilepsy surgery starts in the late 19th century with a paradigm shift in our understanding of cortical functions and their clinical correlates. This conceptual change

not only opened the door to understanding the clinical aspects of epilepsy, but also paved the road for the surgical treatment of epilepsy patients and the initiation of modern neurosurgery. In a sense, the early history of epilepsy surgery is also the history of the initial stages of modern neurosurgery.

The Beginning: England The seminal figure in the history of epilepsy and its surgical treatment is John Hughlings Jackson ( 1835-1911 ). Jackson was appointed assistant physician to the National Hospital for the Paralyzed and Epileptic in Queen Square, London in 1862 with the support of Charles Edouard Brown -Séquard. BrownSéquard ’s approach to the physiology of the nervous system significantly influenced Jackson , and he became interested in diseases of the nervous system.1 2 Jackson made a huge impact on neurological sciences with his landmark lectures and publications in a short period of time. In 1864, he delivered a lecture on “ the method of diagnostic neurology based on clinical physiology” at London Hospital. In 1869, he gave the prestigious Goulstonian Lecture at the Royal College of Physicians on “ the study and classification of diseases of the nervous system .” Finally, in 1870 he published his seminal “A Study of Convul sions” by describing the physiology of focal epilepsy and its cortical localization.1 3-7 These lectures and publications were extremely influential in the scientific community and Jackson initiated a paradigm shift within a few years, creating the conceptual framework for neurophysiology and the scientific understanding of epilepsy. He was the key figure in the history of epilepsy, even deemed to be the “father of epilepsy” accord ing to some accounts. His major impact on clinical neurology and neurosurgery came with his “cerebral localization ” theory, which states that certain areas of the brain are related to different functions. Jackson defined the causal relation between abnormal cortical neuronal activity and focal seizures with this concept, and he stated that “a convulsion is but a symptom , and implies only that there is an occasional, an excessive, and a disorderly discharge of nerve tissue on muscles.” 7 Before Jackson, the cortex was thought to be inexcitable tissue and Jackson was the first to claim that seizures originated in the cortex through disorderly electrical discharges. However, this was only based on his clinical observations and it was still a hypothesis. In 1870, Gustav Theodor Fritsch and Eduard Hitzig in Berlin electrically

-

-

2  A Historical Review of Epilepsy Surgery and Its Application in Children stimulated the motor cortex of awake dogs for the first time and showed that cortical stimulation of certain areas produced sim­ ilar responses with a predictable pattern.1 Thus, they provided experimental evidence for Jackson’s claims, and Jackson stated after these observations that “Epilepsy is the name for occasion­ al, sudden, excessive, rapid, and local discharge of grey matter.”8 This development fully opened the door to subsequent cerebral localization and mapping studies. One of the students of this new exciting research area was Victor Horsley (1857–1916). As a researcher, Horsley performed numerous craniotomies on animals, particularly in macaque monkeys and orangutans, to study cortical localization and brain mapping and followed in the footsteps of Fritsch and Hitzig. In 1886, he was appointed as a staff surgeon to perform craniotomies on patients at the National Hospital for the Paralyzed and Epileptic in Queen Square, London. This was also the first formal “brain surgeon” post in history.1,​3–6 Horsley ascribed to the conceptual framework that was developed by Jackson and Ferrier regarding the casual relationship between cortical localization of epileptic discharges and symptomatol­ ogy during seizures. He closely collaborated with the famous neurologists at Queen Square, such as Jackson and Ferrier, and operated on their patients (Fig. 2.1). He performed the first epilepsy surgery procedure in the same year, in 1886, with both Jackson and Ferrier present in the operating theater.6 Horsley’s seminal paper titled “Brain Surgery” was published that same year.9 This paper was a transcription of his lecture in the Sec­ tion of Surgery of the British Medical Association at the 54th Annual Meeting in Brighton. Horsley began his talk by stating

that his presentation was to be “a simple description of that method of operating on the brain which I have adopted as one which successfully meets the various difficulties and dangers of the task … .” Horsley demonstrated that surgery on brain was feasible and he described his surgical technique with minute details under several subtopics: preparation of the patient, anesthetic, treatment of the wound, line of incision, removal of bone, treatment of the dura mater, treatment of brain, and closure of the wound. He continued his lecture by discussing three illustrative cases on which he had operated, all of which were epilepsy patients. Today, it is widely accepted that the first patient he presented, a 22-year-old man, was the first modern epilepsy surgery case. He continued to operate on more epi­ lepsy patients and a year later, in 1887, he published a series of 10 surgically treated epilepsy patients.10 Three of these patients were children of 4, 10, and 18 years of age. While Horsley made a splash with his surgeries and lectures, another British neurosurgeon, William Macewen (1848–1924) was also operating on epilepsy patients. Macewen, who had quite a different background than Horsley and no research experience, was also following localization techniques defined by Jackson. He worked on establishing cortical localization by observing the symptomatology of seizures. His first paper was published in 1879, even before Horsley’s famous lecture.11 The case was on a boy who had seizures after a traumatic brain injury. These were Jacksonian seizures starting from the face and progressing to the arm and leg. Macewen published two other articles that discussed various neurosurgical cases, including some epilepsy surgery cases.12 Macewen was also

Fig. 2.1  Medical staff at Queen Square in 1888: Victor Horsley (back row, first on the left), John Hughlings Jackson (front row, third from the left), David Ferrier (front row, last from the left). (Reproduced with permission from the Queen Square Library and Archives, London.)

11

12

I  Introduction to Epilepsy in Children the first surgeon to perform cortical resections for grossly “nonvisible” lesions based solely on the clinical findings of the patient.13 One such patient was a 7–year-old girl with recurrent seizures.14 This case is one of the first published pediatric epi­ lepsy surgery cases in history, and Macewen’s description of the case demonstrates his approach to symptomatology and cerebral localization concepts as well as his surgical technique for the excision of a cortical lesion: “7–year-old girl repeated seizures: starts with painful sensa­ tion on the right foot great toe with such a severity as to cause her to scream out. Shortly after, that toe was firmly extended in tonic spasm, which lasted about five minutes. Sometimes this ended the attack. More frequently it was followed by clonic con­ tractions of the muscles of the right foot, leg, and thigh, where the convulsions often terminated. Occasionally they extended to the muscles of the trunk, then to those of the right side of the face and right arm, the contractions ceasing in the order of accession. Rare­ ly did they involve the opposite side, and when they did, patient lost consciousness……during the operation the upper portion of the ascending convolutions was exposed, and with the exception of a few tubercular nodules, the size of barley grains, adhering to the vessels over the upper part of the ascending ­frontal, there

was nothing visible on the surface. On careful palpation of the ascending convolutions, there was found in the upper part of the ascending parietal, a circumscribed nodule buried in the cerebral substance, which on exposure by cutting through the grey mat­ ter, was seen to be a tubercular tumor, about the size of a hazel­ nut, which was easily shelled out … there have been no fits for over a year, and the girl is now in excellent health.”14

„„ Germany: Cortical Stimulation, Mapping and Electroencephalography While Horsley and Macewen were busy operating on epilep­ sy patients in England, Fedor Krause (1857–1937) started to operate on epilepsy patients around the same time in Germany. He had a special interest in epilepsy surgery and used electri­ cal stimulation during surgeries on the motor cortex from as early as 1893. He published the first detailed map of the human motor cortex based on cortical stimulation data from his patients (Fig. 2.2). He operated on over 400 e ­pilepsy

Fig. 2.2  Motor cortex mapping based on electrical stimulation responses. (Reproduced with permission from Krause T. Surgery of the brain and spinal cord; Based on personal experiences. Vol II [English Adaptation Max Thorek] New York: Rebman Company; 1912, 291.)

2  A Historical Review of Epilepsy Surgery and Its Application in Children patients and devoted a large part (> 200 pages) of the second volume of his comprehensive surgery textbook to epilepsy and epilepsy surgery.15 He refined the approach pioneered by Horsley and expanded the indications for epilepsy surgery by including grossly “nonlesional” cases. His technique involved using monopolar stimulation to induce seizures, and then localizing the epileptogenic focus for resection.16 He used cortical stimulation extensively and his stimulation data in patients led him to map the human motor cortex at a level of unprecedented detail. Krause was also likely the first surgeon to have performed intraoperative electrical stimulation of the cere­ bral cortex on a child. He reported a 15-year-old patient with seizures since the age of 2, with the seizures being cured after surgery. In this case, he used cortical stimulation to localize the excision area.16 Krause also emphasized the critical significance of the resection of the ictal onset zone, “primares krampfendes Zentrum” (primary spasmodic center), to succeed good postop­ erative seizure control.1 In 1909, the International League Against Epilepsy (ILAE) was founded with an inaugural meeting in Budapest. Many prominent neurosurgeons, including William Macewen, Fedor Krause, and Harvey Cushing attended the meeting and delivered lectures.1 That same year, Sir Victor Horsley was the speaker for the prestigious Linacre Lecture at the University of C ­ ambridge and gave his famous lecture entitled “The Function of the So-called Motor Area of the Brain.”17 One of the cases Horsley

presented in his lecture was a 14-year-old child with “athetoid movements of the left hand which then developed into violent convulsive movements of the whole upper limb.” Although the true nature of these episodes is unclear, the patient underwent surgical intervention with detailed mapping and then a resec­ tion of the motor cortex (Fig. 2.3). Horsley reported that the patient’s spasmodic movements disappeared immediately and he had partial recovery of his arm movements 1 year later. The figures he presented in this lecture neatly document his metic­ ulous cortical stimulation and mapping technique. After the First World War, many veterans in Europe suffered from seizures secondary to various traumatic brain injuries. In Germany, Hans Berger (1873–1941), a neuropsychiatrist at the University of Jena, conducted a series of experiments by placing needle electrodes under the scalps of patients who had skull defects following surgeries during the war period. One of these patients was a 17-year-old boy with a large skull defect, and Berger recorded the first human EEG on this patient in 1924. Initially, he was unsure about the nature of the oscillations he had recorded and subsequently conducted many other experi­ ments, deciding to wait to publish his first paper on the human electroencephalogram until 1929.18–19 Next, another German neurologist and neurosurgeon, Otfrid Foerster (1873–1941), used EEG for intraoperative recordings. In 1934, he published his intraoperative EEG findings by describing the ictal seizure patterns on invasive EEG recordings of 30 patients.20 Around the Fig. 2.3  (a) Horsley’s sketch of operation field made immediately after the operation. The numbers indicate the points stimulated. (b) Outline of the gyrus precentralis removed. (c) Photograph of the resected gyrus precentralis fixed in formol. Abd., abduction; ret., retraction; e.e., elbow extend; w.e., wrist extend; w.f., wrist flexed; ul. ad., ulnar adduction; f.f., fingersflex. (Reproduced with permission from Linacre Lecture, 1909—Victor Horsley. Function of so-called motor area of the brain. British Medical Journal 1909;11:125–132.)

13

14

I  Introduction to Epilepsy in Children same time, Gibbs and Lennox started to use scalp EEGs system­ atically in the United States for diagnostic purposes in epilepsy patients.21 The development of EEG created a new route to localizing the epileptogenic cortical areas,22–23 and some years later, Bailey and Gibbs reported interictal EEG guided resection technique in a series of epilepsy patients.24 Another pivotal figure of this era was Otfrid Foerster (1873–1941) in Breslau, Germany.1,​3,​5,​25 Foerster was originally a neurologist, but started to operate on patients after being frustrated with the postoperative outcomes of his patients who were operated on by other surgeons in his hospital. He was a pioneer in awake craniotomy with cortical stimulation as well as mapping of the human cortex. He used cortical stimulation in almost all brain surgery cases he performed and accumulat­ ed an enormous amount of experience and data that helped him to produce a detailed brain map based on his stimulation results. He also performed the first electrocorticography (ECoG) on a patient. His fame attracted many neurosurgeons from the United States and Britain, including Fulton, Bucy, B ­ ailey, and Penfield, to visit the Breslau Neurological Institute.25 Wilder Penfield worked with Foerster in 1928 to learn awake craniot­ omy and cortical stimulation techniques.1 Penfield participated Sherrington’s physiology laboratory when he was studying at Oxford and already familiar with stimulation studies in nervous ­system. He planned to expand his experience on cortical stimu­ lation by working with Foerster. Penfield and Foerster published papers together including a paper describing an expanded map of the human cortex in 1930 by showing detailed information about functional areas other than the motor cortex, i­ncluding

sensory, acoustic, and visual ­ representations.26–27 Penfield was highly impressed with both Foerster’s surgical technique, mainly his awake surgery and cortical mapping, as well as his “­Institute of Neurology at Breslau” hospital model where neurology and neurosurgery disciplines worked ­ together ­seamlessly. Inspired by the model, Penfield created a similar institution in Montreal, Canada several years later: the Montre­ al Neurological Institute (MNI). This period at the early stage of his career form the groundwork for Penfield as a physiological surgeon of the human brain.28

„„ The Montreal Neurological Institute: The Beginning of a New Era Penfield returned to Montreal with an increased interest and enthusiasm for surgical treatment of epilepsy. He would later describe his time in Germany as “one of the greatest moments” in his life.28 Penfield adapted two of Foerster’s innovative tech­ niques in his own work: awake surgery and cortical stimula­ tion during surgery. Later, Penfield acknowledged Foerster’s influence on him clearly: “I began to operate on epileptogenic cicatrices, following the example of Otfrid Foerster, in 1928, and my indebtedness to him is obvious.”29–30 After returning to Montreal, Penfield operated on his patients using Foerster’s technique and, in 1934, he established the MNI with his ­colleague William Cone (Fig. 2.4).

Fig. 2.4  Faculty and fellows at the Montreal Neurological Institute’s early years. W. Penfield (fourth from right) and W. Cone (third from right) are sitting in the second row. (Reproduced with permission from the Osler Library of the History of Medicine at McGill University.)

2  A Historical Review of Epilepsy Surgery and Its Application in Children One of the patients Penfield operated on in the same year was a 14-year-old child, and he performed the procedure using local anesthesia and cortical stimulation.31 Initially, Penfield used Foerster’s technique to map the motor and sensory areas of his patients in order to protect them from neurological defi­ cits during the surgical excision of the epileptogenic focus. Gradually, however, Penfield and Rasmussen went further by stimulating other cortical regions and identifying a variety of complex and higher cognitive responses including speech, secondary sensory and motor areas, autonomic function, and even “the sites of dreams and memories”.5,​16 Penfield and his colleagues in MNI continued cortical functional mapping studies with intraoperative electrical stimulation over the entire brain, including the insula, and documented their results with min­ ute details (Fig. 2.5). This database covered the entire human cerebral cortex and its functional representations. Based on the data, they collected from more than 400 cortical stimula­ tions, the “homunculus” concept was gradually developed.32–35 ­Penfield presented “the cerebral cortex of man” during the Lane Lectures at Stanford University in 1947.36 Penfield recognized the significance of neurophysiology in a very early stage of his training and saw the value of EEG in the evaluation of epileptic patients quickly and established a laboratory for EEG and neurophysiology at the MNI for Jasper. Close collaboration between Penfield and Jasper made the MNI a pioneer in electrophysiology-guided treatment of epilepsy patients during the second half of 1930s (Fig. 2.6).37 Natural­

ly, the MNI was the first center in the development of longterm invasive EEG monitoring. Penfield described the first case of long-term EEG monitoring with invasive electrodes in 1939.5,​38–40 The patient was a 32-year-old man with traumat­ ic brain injury and intractable epilepsy. The MNI team placed epidural electrodes bilaterally through burr holes over the temporal lobes and performed the first prolonged invasive EEG recording. A series of EEGs were obtained in this patient, but only during the day time, for 3 days. Abnormal activities were localized to the left side over the region with traumatic injury. The patient underwent surgery with intraoperative cortical stimulation and mapping, as well as ECoG for excision of the scar tissue. The patient did not have any improvement after surgery, however. Thereafter, Penfield and Jasper introduced electrocorticography with subdural electrodes in the 1950s. All the studies done in the MNI were published and ushered a new era for the application of EEG in epilepsy patients for both diagnostic purposes preoperatively, and also for the localiza­ tion of the epileptogenic areas during surgery.41–42 Initially, procedures in the MNI were mostly extratem­ poral and lesional, as Penfield was against the resection of normal appearing cortex.28,​42 Gradually, he started to use the ECoG-guided resection technique by collaborating with Jaspers and started to excise surrounding normal cortex around the lesion when the ECoG showed epileptogenic activity over these areas. Later, Penfield’s focus shifted to temporal lobe epilepsy, and the surgical management of

Fig. 2.5  A drawing of intraoperative cortical stimulation results in one of the Penfield’s patient. (Reproduced with permission from the Osler Library of the History of Medicine at McGill University.)

15

16

I  Introduction to Epilepsy in Children

Fig. 2.6  Wilder Penfield and Herbert Jasper at the Montreal Neurological Institute, 1954. (Reproduced with permission from Feindel and Leblanc 2016.36)

temporal lobe epilepsy became the main focus at the MNI. The MNI gradually became the most innovative center in epilepsy surgery under Penfield’s leadership and collaboration with the other pioneers working at MNI such as Jasper, Rasmussen, Feindel, and Gloor. The Wada test was developed at the MNI, as neuropsychological assessments became a standard part of the work up of epilepsy surgery patients. Additionally, Rasmussen revived hemispheric surgeries for patients with multilobar/hemispheric epilepsy by describing the function­ al hemispherectomy technique. Rasmussen also significantly increased the number of pediatric epilepsy surgery cases at the MNI. As a result, the MNI became a very important place for epilepsy surgery and numerous visitors and trainees from all over the world came there to learn.

„„ The Quest for a Surgical Treatment for Temporal Lobe Epilepsy In 1934, Frederick Gibbs, Erna Gibbs, and WG Lennox, the “Boston team,” convincingly recorded spike and wave com­ plexes for the first time in the EEG of petit mal epilepsy patients. This development placed the EEG in a strategic posi­ tion for the diagnosis of epilepsy. Two years later, Gibbs and Lennox described psychomotor seizures as a separate entity and in 1948, they reported the findings of more than 300 patients with “psychomotor seizures” by describing abnormal EEG findings over the anterior temporal lobe.43–​46 In 1944, Frederic and Erna Gibbs moved to Chicago to work with Percival Bailey at the University of Illinois at Chicago.43 In 1947, Gibbs convinced Bailey to perform ECoG-guided ­resections and Bailey operated on 16 patients by January 1948 with an unsatisfactory success rate.42,​43 This prompted Bailey to develop a more radical resection technique and he started to perform anterior temporal lobectomy procedures extending ­ from the sylvian fissure to the occipitotemporal sulcus by

s­paring mesial temporal structures. The posterior borders of his resections reached the level of the central sulcus posterior­ ly. The results were much better.6,​42,​47 At about the same time, Arthur Morris at Georgetown University School of Medicine also operated on patients with psychomotor seizures. His lobectomy technique included the uncus, amygdala, and anterior 2 to 4 cm of the hippocampus in addition to the anterolateral temporal cortex.28,​42 Morris initially used ECoG, but then decided that it was not helpful and started to perform simple “standard tem­ poral lobectomies” without ECoG.42 Wilder Penfield performed his first temporal lobe resection in 1931.28 Initially, his resec­ tions were limited to the temporal neocortex. He operated on 68 cases between 1939 and 1949 and removed the hippo­ campus in only two cases.28,​42 After 1949, Penfield started to remove the uncus, amygdala, and hippocampus routinely along the anterior temporal neocortex anterior to the vein of Labbe and subsequently defined the technique known as the “standard MNI temporal lobectomy.”5,​16,​28,​47–48 While the MNI temporal lobectomy technique was the standard approach in many centers, Murray A. Falconer at Guy’s-Maudsley Hospital in London developed the en bloc anterior temporal lobectomy procedure in temporal lobe epi­ lepsy patients and started to send the entire hippocampus as an en bloc specimen for pathological examinations.49–51 The en bloc removal technique gave pathologists the opportunity to examine epileptogenic tissue within that structure’s anatom­ ical entirety. This approach led to a dramatic impact on our understanding of the pathology of temporal lobe epilepsy. As a result, hippocampal sclerosis was noted to be the most com­ mon histological abnormality in pathological examinations at Guy’s-Maudsley Hospital temporal lobe epilepsy cases.52 In 1963, detailed histological examinations of Falconer’s tempo­ ral lobe specimens revealed that the dominant pathology was mesial temporal sclerosis, which was seen in 47% of cases.51 Another pathological entity, focal dysplasia, was also described by Taylor and Falconer using these specimens in 1971.53 Two major contributions to temporal lobe surgery tech­ nique came from Zurich during the late 1960s. M.G. Yaşargil introduced the operating microscope and microsurgical techniques to neurosurgery in 1967. He also developed an innovative transsylvian approach to treat aneurysms and vascular malformations of the circle of Willis. Later, in 1973, Yaşargil started to apply this technique to perform selective transsylvian amygdalohippocampectomy in patients with mesial temporal epilepsy.54 The application of microneu­ rosurgery techniques using microscopic instruments made a dramatic impact on the quality of surgical results in the treatment of epilepsy patients.

„„ Exploring the Epileptogenic Network: Stereoencephalography One of the fellows at the MNI was a French neurologist, H ­ enri ­Hecaen, who came to the MNI for epilepsy surgery training following the encouragement of two neurosurgical colleagues, G. Mazars and J. Guillaume, at Sainte-Anne Hospital in ­Paris.1 After completing his fellowship at the MNI, Hecaen started an epilepsy service at Sainte-Anne Hospital in Paris and in the early 1950s, epilepsy surgery cases started to be. During this

2  A Historical Review of Epilepsy Surgery and Its Application in Children time, another neurosurgeon in the same hospital, Jean Talairach, was focusing on stereotactic techniques and deep brain stimu­ lation. Jean Bancaud, a neurologist from La Salpêtrière Hospital in ­Paris, realized the potential value of stereotactically placed depth electrodes in EEG recording and joined the Sainte-Anne team to c­ ollaborate with Talairach. Together, they introduced the ­stereotactical implantation of depth electrodes to detect spatiotemporal electrical distribution of EEG activity and ­ named this technique as stereoelectroencephalography (SEEG) in the late 1950s.1,​55–56 Although the first stereotactic depth electrode placement for EEG recording in humans was reported by Hayne and M ­ eyers in 1949, it was not used commonly until the Sainte-Anne team established the methodology.57 The goal was to define the epileptogenic network and correlate it with seizure semiology. SEEG provided information for the three-dimensional analysis of seizure patterns and their propagations and made it possible to monitor intracranial EEG for longer durations of time. Intra­ operative EEG and stimulation studies performed at the MNI were helpful mainly to document the interictal activities, but not much for ictal activities. Talairach and Bancaud were able to record the seizure patterns and their clinical correlates using long-term SEEG monitoring. They also developed the concept of anatomo-electro-clinical correlates to define the epileptogenic network.38 Later, G. Szikla and G. Munari joined Talairach, and A. Bonis and P. Chauvel joined Bancaud, to make Sainte-Anne Hos­ pital another magnet center in epilepsy surgery that attracted visitors and trainees from many other hospitals and countries.

„„ Recording Seizures: Long-term Electroencephalography/Video Monitoring In 1954, Paul Crandall joined the University of California Los Angeles (UCLA) School of Medicine as one of three founding members of the neurosurgery division. Around same time, in 1956, Ross Adey at UCLA developed a radio-­ telemetry ­system for NASA to record and transmit EEG and other phys­ iological data from chimpanzees orbiting the earth.58,​59 Along with Richard Walter, a neurologist with training on bio­ electric signals at MIT, Crandall appreciated the value of this technology and established the first e ­ pilepsy telemetry unit at UCLA. Crandall used telemetry to obtain continuous EEG recording from depth electrodes over long periods of time. Although the Sainte-Anne team in Paris developed EEG recordings using stereotactic depth electrodes, French law permitted recordings for no more than 7 hours. At UCLA, Crandall adapted this technique and used it for long-term ­ EEG monitoring, at about 8 hours each day for several weeks, in 1961.60 The UCLA team published their experience with chronic EEG monitoring with stereotactically implanted depth ­electrodes to record EEG changes occurring at the onset of spon­ taneous seizures in 1963. In 1968, Crandall and his colleagues were the first to record the onset of a seizure in a human being using intracranial electrodes. Video monitoring capabilities were added to the telemetry units and EEG/video monitoring units were established in early 1970s.13,​61–63 Depth electrode placement was not the only invasive electrodiagnostic technique. Subdural electrodes were also

previously used by the MNI team, and a very detailed study with subdural electrodes was published by Ajmane-Marsan and Van Buren in 1958.64 The authors went as far as to describe an implantation plan for temporal lobe epilepsies, but the needed subdural grids were not popularized until the mid-1980s. Lat­ er on, however, it became the most commonly used invasive monitoring technique in the United States.38,​65

„„ New Frontier: Pediatric Epilepsy Surgery Children were always part of the epilepsy surgery series since the beginning of its history, including Horsley’s first epilepsy surgery series in 1887.10 However, the systematic approach to pediatric epilepsy and its surgical management is quite a recent development. Murray A. Falconer was the first vocal advocate of surgical treatment of epilepsy in children (Fig. 2.7). Reportedly, Falconer confided in one of his colleagues that “after establish­ ing the indication and methodology of temporal lobe surgery” he felt that “it is his mission to propagate and sell this knowl­ edge to encourage epileptologists to apply these techniques in pediatric years.”66 Not surprisingly, he became the champion of surgical treatment of temporal lobe epilepsy in children during his tenure years and published a series of papers to report his surgical results in pediatric epilepsy patients.67–69 He presented his findings in meetings and reported that success rates of pediat­ ric epilepsy surgery cases were comparable to those of adults.66,​70 Falconer was one of the first neurosurgeons who published purely pediatric epilepsy surgery series.67–70 Before Falconer, almost all published series included mostly adult patients with occasional pediatric cases who were mostly older children. Even Falconer’s first series in 1970 included only 9 children (of 9 to 14 years of age).67 His next publication, in 1975, included 40 children with temporal lobe epilepsy. Six of these children were

Fig. 2.7  Murray A. Falconer, 1910–1977.

17

18

I  Introduction to Epilepsy in Children younger than 10 years old.69 The duration of follow-up in this series was 1 to 24 years, and Falconer reported good outcomes in 77% of his patients. Soon more reports specifically addressing epilepsy surgery in children began to be published. Rasmussen at the MNI started to focus on pediatric cases and published his experience in 1977.71 In 1978, Green et al published 50 pediat­ ric epilepsy surgery cases with an age range of 2 to 18 years.72 Another pioneer in pediatric epilepsy surgery was Sidney Goldring at Washington University in St. Louis (Fig. 2.8). His first report included 17 patients (2–14 years old) in 1978, then 44 children (5 months to 14 years) in 1984, and, finally, 75 chil­ dren (5 months to 15 years) in 1987.73–75 He initially used ECoG to determine epileptogenic areas in children, like many other neurosurgeons at the time. Awake craniotomy with a child was extremely difficult, however, and only feasible for older chil­ dren.75 After 1971, Goldring started to perform craniotomies to implant epidural electrodes for extraoperative stimulation and mapping to determine the epileptogenic areas in children.73–75 Goldring operated on 95 children with medically intractable epilepsy between 1967 and 1986. The series he published in 1987 is one of the earliest and largest series for invasive mon­ itoring in children with epilepsy.73 In this series, he reported on 75 children, between 5 months to 15 years of age, with intractable epilepsy. He performed EEG/video monitoring with epidural electrodes in 53 of these children. Goldring continued to operate pediatric epilepsy surgery patients until the end of his career with good postoperative outcome and reported a wide range of lesions including Rasmussen’s encephalitis, mesial temporal sclerosis, infantile hemiplegia, cortical dysplasia, tuberous sclerosis, etc. Long-term outcome of 361 patients who underwent epilep­ sy surgery by Dr. Sidney Goldring at Barnes/St. Louis Children’s Hospital from 1967 to 1990 was later reviewed and published in 2012. Mean follow-up duration of this group was 26 years and, remarkably, 48% of these patients were still Engel class I.76 One of the other early pediatric epilepsy surgery series was published by Meyer et al77 in 1986 and included 50 children (7–18 years old). Gradually, the age of reported cases decreased in the early 1990s and eventually, Morrison et al78 published on 45 children with a mean age of 9. Fifty-seven percent of

the patients in this series were younger than 10 years old, 24% being younger than 5 years old. Another large pediatric epilepsy surgery series was published by the MNI by review­ ing their results in the pediatric age group retrospectively.79–80 These patients were mostly operated on by Rasmussen and the report covered a large time period between 1940 and 1980. These patients were typically older children with an average age of 11.7 years (range: 0.6–15 years) at operation. Only 11 of them were 0 to 5 years of age, and 18 of them were 6 to 10 years of age. In the late 1980s, with the advancement of diagnostic technology such as MRI and PET/SPECT, as well as improved anesthesiology and surgical techniques, the surgical treatment of epilepsy slowly became a treatment modality for younger children and infants, as well. The first series reporting epilep­ sy surgery results in infants began to be published in 1988.81–86 Initially, these reports were mostly from UCLA, Miami Children’s Hospital, and Cleveland Clinic, and then increasing numbers of reports and presentations on epilepsy surgery in infants started to come from other countries in Europe, as well as Japan.

„„ Pediatric Guidelines, Task Forces, and Meetings The first epilepsy surgery meeting with international par­ ticipation was “Palm Desert Conference on Surgical Treat­ ment of the Epilepsies,” which was organized by J. Engel in California and held in 1986 to discuss and compare epilepsy surgery programs’ surgical approaches to epilepsy patients.87 Shortly after, the first International Pediatric Epilepsy ­Surgery ­Symposium was organized at Miami Children’s Hospital in 1989. This was followed by another meeting focused on pedi­ atric epilepsy surgery, the Bethel-CCF Epilepsy ­Symposium: Epilepsy ­Surgery in Children, in 1995. The Commission on neurosurgery of the ILAE formed the “Pediatric Epilepsy ­Surgery S ­ ubcommission” in 1998 with a mandate of formu­ lating guidelines and recommendations for epilepsy surgery in childhood. The subcommission organized several meetings and workshops for this purpose, which took place in 1999 in Orlando, 2000 in Cleveland and Los Angeles, and 2003 in France. The subcommission published its report “proposed criteria for referral and evaluation” in 2006.88 Finally, “Task Force for Pediatric Epilepsy Surgery for the ILAE Commissions of Pediat­ rics and Surgical Therapies” was established to ­survey related topics and publish reports.89

„„ Conclusion

Fig. 2.8  Sidney Goldring, 1923–2004. (Reproduced with permission from the Archives of the Department of Neurological Surgery, Washington University in St. Louis School of Medicine, Missouri, USA.)

The history of pediatric epilepsy surgery goes back to the first epilepsy surgery series reported by Victor Horsley in London in 1887. Since then, epilepsy surgeons have operated on chil­ dren in addition to adult patients. It took almost a century after the initial series of operations, however, for the first “pediatric” epilepsy surgery series to be published. It took even longer to develop guidelines for standardizing the surgical approach to children with intractable epilepsy. Today, pediatric epilepsy surgery has become an established treatment modality for a wide range of ages, including infants, thanks to these pioneer physicians and surgeons.

2  A Historical Review of Epilepsy Surgery and Its Application in Children

References 1. Schijns OEM, Hoogland G, Kubben PL, Koehler PJ. The start and development of epilepsy surgery in Europe: a historical review. Neurosurg Rev 2015;38(3):447–461 2. York GK, Steinberg DA. An introduction to the life and work of John Hughlings Jackson with a catalogue raisonné of his writings. Med Hist Suppl 2006;26(26):3–157 3. Magiorkinis E, Diamantis A, Sidiropoulou K, Panteliadis C. Highlights in the History of Epilepsy: The Last 200 Years. Epi­ lepsy Res Treat 2014;582039:13 4. Taylor DC. One hundred years of epilepsy surgery: Sir V ­ ictor Horsley’s contribution. J Neurol Neurosurg Psychiatry 1986;49(5):485–488 5. Feindel W, Leblanc R, de Almeida AN. Epilepsy surgery: historical highlights 1909–2009. Epilepsia 2009;50(Suppl 3):131–151 6. Shorvon S. The evolution of epilepsy theory and practice at the National Hospital for the Relief and Cure of Epilepsy, Queen Square between 1860 and 1910. Epilepsy Behav 2014;31:228–242 7. Hughlings JJ. A study of convulsions. Trans St Andrews Med Grad Assoc 1870;3:162–204 8. Hughlings JJ. On the anatomical, physiological, and pathological investigation of epilepsies. West Riding Lunatic Asylum Medical Reports 1873;3:315–349 9. Horsley V. Brain surgery. BMJ 1886;2:670–675 10. Horsley V. Remarks of ten consecutive cases of operations upon the brain and cranial cavity to illustrate the details and safety of the method employed. BMJ 1887;1(1373):863–865 11. Macewen W. Tumour of the dura mater removed during life in a person affected with epilepsy. Glasgow Med J 1879;12:210 12. Macewen W. Intra-cranial lesions: illustrating some points in connexion with the localisation of cerebral affections and the advantages of aseptic trephining. Lancet 1881;2:544–581 13. Wilson SJ, Engel J Jr. Diverse perspectives on developments in epilepsy surgery. Seizure 2010;19(10):659–668 14. Macewen W. An address on the surgery of the brain and spinal cord. BMJ 1888;2(1441):302–309 15. Krause F. Chirurgie des Gehirns und Ruckenmarks nach eigenen Erfahrungen. Berlin: Urban und Schwarzenberg; 1911 16. Olivier Andre, Boling WW, Tanriverdi T. History of epilepsy surgery. In: Techniques in Epilepsy Surgery: The MNI Approach. Cambridge: Cambridge University Press; 2012:1–14 17. Horsley V. The Linacre lecture on the function of the so-called motor area of the brain. BMJ 1909;2(2533):121–132. (Delivered to the Master and Fellows of St. John’s College, Cambridge, May 6, 1909.) 18. Berger H. Über das Elektroencephalogramm des Menschen. Arch Psychiatr Nervenkr 1929;87(1):527–570 19. Tudor M, Tudor L, Tudor KI. [Hans Berger (1873–1941)— the history of electroencephalography] Acta Med Croatica 2005;59(4):307–313 20. Foerster O, Altenburger H. Elektrobiologische Vorgänge an der menschlichen Hirnrinde. Dtsch Z Nervenheilkd 1934;135: 277–288 21. Gibbs FA, Gibbs EL, Lennox WG. The electro-encephalogram in diagnosis and in localization of epileptic seizures. Arch Neurol Psychiatry 1936;36(6):1225–1235 22. Gibbs FA, Gibbs EL, Lennox WG. Cerebral dysrhythmias of epi­ lepsy. Arch Neurol Psychiatry 1938;39(2):298–314 23. Jasper H, Kershman J. Electroencephalographic classification of the epilepsies. Arch Neurol Psychiatry 1941;45(6):903–943 24. Bailey P, Gibbs FA. The surgical treatment of psychomotor epi­ lepsy. J Am Med Assoc 1951;145(6):365–370 25. Piotrowska N, Winkler PA. Otfrid Foerster, the great neurologist and neurosurgeon from Breslau (Wrocław): his influence on

early neurosurgeons and legacy to present-day neurosurgery. J Neurosurg 2007;107(2):451–456 26. Foerster O, Penfield W. Der narbenzug am und im gehirn bei traumatischer epilepsie in seiner bedeutung für das zustande­ kommen der anfälle und für die therapeutische bekämpfung derselben. Z Ges Neurol Psychiat 1930;125:475–572 27. Foerster O, Penfield W. The structural basis of traumatic epilepsy and results of radical operation. Brain 1930;53(2):99–119 28. de Almeida AN, Teixeira MJ, Feindel WH. From lateral to mesial: the quest for a surgical cure for temporal lobe epilepsy. Epilepsia 2008;49(1):98–107 29. Penfield W. No man alone. Toronto: Little, Brown & Co.;1977:48 30. Penfield W. Epilepsy and surgical therapy. Arch Neuro Psych 1936;36(4):449–484 31. Elder R. Speaking secrets: epilepsy, neurosurgery and patient testimony in the age of the explorable brain, 1934–1960. Bull Hist Med 2015;89(4):761–789 32. Rasmussen T, Penfield W. The human sensorimotor cortex as ­studied by electrical stimulation. Fed Proc 1947;6(1 Pt 2):184 33. Rasmussen T, Penfield W. Further studies of the sensory and motor cerebral cortex of man. Fed Proc 1947;6(2):452–460 34. Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain 1937;60(4):389–443 35. Penfield W, Faulk ME Jr. The insula; further observations on its function. Brain 1955;78(4):445–470 36. Feindel W, Leblanc R. The Wounded Brain Healed: The Golden Age of the Montreal Neurological Institute, 1934–1984. Mon­ treal: McGill-Queen’s University Press; 2016:206 37. Penfield W. The epilepsies: with a note on radical therapy. N Engl J Med 1939;221:209–218 38. Reif PS, Strzelczyk A, Rosenow F. The history of invasive EEG eval­ uation in epilepsy patients. Seizure 2016;41:191–195 39. Almeida AN, Martinez V, Feindel W. The first case of invasive EEG monitoring for the surgical treatment of epilepsy: historical significance and context. Epilepsia 2005;46(7):1082–1085 40. Penfield W, Flanigin H. Surgical therapy of temporal lobe sei­ zures. AMA Arch Neurol Psychiatry 1950;64(4):491–500 41. Jasper H, Pertuisset B, Flanigin H. EEG and cortical electrograms in patients with temporal lobe seizures. AMA Arch Neurol Psy­ chiatry 1951;65(3):272–290 42. Moran N, Shorvon S. The surgery of temporal lobe epilepsy: historical development, patient selection and seizure outcome. In: Shorvon S, Pedley TA, eds. The Epilepsies, 3rd ed. Philadel­ phia, PA: Saunders Elsevier; 2009:294–306 43. Stone JL. Biography: Frederic A. Gibbs, M.D. Surg Neurol 1994;41(2):168–171 44. Gibbs GA, Davis H, Lennox WG. The electro-encephalogram in epilepsy and in conditions of impaired consciousness. Arch Neu­ rol Psychiatry 1935;34(4):1133–1148 45. Gibbs FA, Gibbs EL, Lennox WG. Epilepsy: a paroxysmal cerebral dysrhythmia. Brain 1937;60:377–388 46. Gibbs EL, Gibbs FA, Fuster B. Psychomotor epilepsy. Arch Neurol Psychiatry 1948;60(4):331–339 47. Moran NF. A more balanced and inclusive view of the history of temporal lobectomy. Epilepsia 2008;49(3):543–544 48. Penfield W, Jasper H. In epilepsy and functional anatomy of the human brain. London: J&A Churchill Ltd.; 1954:v–vii 49. Falconer MA. Clinical manifestations of temporal lobe epilepsy and their recognition in relation to surgical treatment. BMJ 1954;2(4894):939–944 50. Falconer MA, Meyer A, Hill D, Mitchell W, Pond DA. Treatment of temporal-lobe epilepsy by temporal lobectomy; a survey of findings and results. Lancet 1955;268(6869):827–835

19

20

I  Introduction to Epilepsy in Children 51. Falconer MA, Serafetinides EA. A follow-up study of surgery in temporal lobe epilepsy. J Neurol Neurosurg Psychiatry 1963;26(2):154–165

71. Rasmussen T. Surgical aspects. In: Blau ME, Rapin I, Kinsbourne M, eds. Topics in Child Neurology (First International Congress of Child Neurology). New York, NY: Spectrum; 1977:143–57

52. Falconer MA. The significance of mesial temporal s­clerosis (Ammon’s horn sclerosis) in epilepsy. Guys Hosp Rep 1968;117(1):1–12

72. Green JR, Sidell AD, Walker ML. Neurosurgery of epilepsy in childhood and adolescent with comments about 50 patients. In: Thompson RA, Green JR, eds. Pediatric Neurology and Neurosur­ gery. New York, NY: Medical Science Book; 1978:151

53. Taylor DC, Falconer MA, Bruton CJ, Corsellis JAN. Focal dysplasia of the cerebral cortex in epilepsy. J Neurol Neurosurg Psychiatry 1971;34(4):369–387

73. Goldring S. Pediatric epilepsy surgery. Epilepsia 1987;28 (Suppl 1):S82–S102

54. Wieser HG, Yaşargil MG. Selective amygdalohippocampectomy as a surgical treatment of mesiobasal limbic epilepsy. Surg Neu­ rol 1982;17(6):445–457

74. Goldring S. A method for surgical management of focal epilepsy, especially as it relates to children. J Neurosurg 1978;49(3):344–356

55. Talairach J, David M, Tournoux P. L’exploration Chirugicale ­Stereotaxique du Lobe Temporale dans L’epilepsie Temporale. Paris: Masson et Cie; 1958

75. Goldring S, Gregorie EM. Surgical management of epilepsy using epidural recordings to localize the seizure focus. Review of 100 cases. J Neurosurg 1984;60(3):457–466

56. Bancaud J, Talairach J, Schaub C, et al. Stereotaxic functional exploration of the epilepsies of the supplementary areas of the mesial surfaces of the hemisphere. Electroencephalogr Clin Neurophysiol 1962;14:788

76. Mohammed HS, Kaufman CB, Limbrick DD, et al. Impact of epi­ lepsy surgery on seizure control and quality of life: a 26-year follow-up study. Epilepsia 2012;53(4):712–720

57. Hayne R, Meyers R, Knott JR. Characteristics of electrical activity of human corpus striatum and neighboring structures. J Neuro­ physiol 1949;12(3):185–195 58. Crandall PH, Walter RD, Rand RW. Clinical applications of stud­ ies on stereotactically implanted electrodes in temporal lobe epilepsy. J Neurosurg 1963;20:827–840 59. Dymond AM, Sweizig JR, Crandall PH, Hanley J. Clinical applica­ tion of an EEG radio telemetry system. Proceedings of the Rocky Mountain Bioengineering Symposium. Fort Collins: Colorado State University; 1971;16–20 60. Engel J, Mathern GW. Paul H. Crandall, MD (1923–2012). Neurol­ ogy 2012;79(2):121–122 61. Engel J Jr, Rausch R, Lieb JP, Kuhl DE, Crandall PH. Correla­ tion of criteria used for localizing epileptic foci in patients considered for surgical therapy of epilepsy. Ann Neurol 1981;9(3):215–224

77. Meyer FB, Marsh WR, Laws ER Jr, Sharbrough FW. Temporal lobec­ tomy in children with epilepsy. J Neurosurg 1986;64(3):371–376 78. Morrison G, Duchowny M, Resnick T, et al. Epilepsy sur­ gery in childhood. A report of 79 patients. Pediatr Neurosurg 1992;18(5–6):291–297 79. Fish DR, Smith SJ, Quesney LF, Andermann F, Rasmussen T. Sur­ gical treatment of children with medically intractable frontal or temporal lobe epilepsy: results and highlights of 40 years’ expe­ rience. Epilepsia 1993;34(2):244–247 80. Quesney LF, Fish DR, Rasmussen T. Extracranial EEG and acute ECoG in children with medically refractory partial seizures. J Epilepsy 1990;3(Suppl 3):55–67S 81. Chugani HT, Shields WD, Shewmon DA, Peacock WJ, Mazziotta JC, Phelps MA. Surgical treatment of intractable neonatal-onset seizures: the role of positron emission tomography. Neurology 1988;38:1178–1188

62. Ives JR, Gloor P. A long term time-lapse video system to document the patients spontaneous clinical seizure synchronized with the EEG. Electroencephalogr Clin Neurophysiol 1978;45(3):412–416

82. Chugani HT, Shields WD, Shewmon DA, Olson DM, Phelps ME, Peacock WJ. Infantile spasms: I. PET identifies focal cortical dysgenesis in cryptogenic cases for surgical treatment. Ann Neu­ rol 1990;27(4):406–413

63. Ives JR, Gloor P. A longterm time-lapse video system to document the patients spontaneous clinical seizure synchronized with the EEG. Electroencephalogr Clin Neurophysiol

83. Chugani HT, Shewmon DA, Shields WD, et al. Surgery for intrac­ table infantile spasms: neuroimaging perspectives. Epilepsia 1993;34(4):764–771

64. Ajmone-Marsan C, Van Buren J. Epileptiform activity in cor­ tical and subcortical structures in the temporal lobe of man. In: Baldwin M, Bailey P, eds. Temporal Lobe Epilepsy. E Springfield, IL: Thomas; 1958:78–108

84. Duchowny MS, Resnick TJ, Alvarez LA, Morrison G. Focal resection for malignant partial seizures in infancy. Neurology 1990;40(6):980–984

65. Wyler AR, Ojemann GA, Lettich E, Ward AA Jr. Subdural strip electrodes for localizing epileptogenic foci. J Neurosurg 1984;60(6):1195–1200 66. Bladin PF. Murray Alexander Falconer and the Guy’s-Maud­ sley Hospital seizure surgery program. J Clin Neurosci 2004;11(6):577–583 67. Falconer MA. Significance of surgery for temporal lobe epilepsy in childhood and adolescence. J Neurosurg 1970;33(3):233–252 68. Falconer MA. Place of surgery for temporal lobe epilepsy during childhood. BMJ 1972;2(5814):631–635 69. Davidson S, Falconer MA. Outcome of surgery in 40 children with temporal-lobe epilepsy. Lancet 1975;1(7919):1260–1263 70. Falconer MA. Reversibility by temporal-lobe resection of the behavioral abnormalities of temporal-lobe epilepsy. N Engl J Med 1973;289(9):451–455

85. Wyllie E, Comair YG, Kotagal P, Raja S, Ruggieri P. Epilepsy sur­ gery in infants. Epilepsia 1996;37(7):625–637 86. Duchowny M, Jayakar P, Resnick T, et al. Epilepsy surgery in the first three years of life. Epilepsia 1998;39(7):737–743 87. Engel J Jr. Surgical Treatment of the Epilepsies. New York, NY: Raven Press; 1987 88. Cross JH, Jayakar P, Nordli D, et al; International League against Epilepsy, Subcommission for Paediatric Epilepsy Surgery. Commissions of Neurosurgery and Paediatrics. Proposed crite­ ria for referral and evaluation of children for epilepsy surgery: recommendations of the Subcommission for Pediatric Epilepsy Surgery. Epilepsia 2006;47(6):952–959 89. Harvey AS, Cross JH, Shinnar S, Mathern GW; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epi­ lepsia 2008;49(1):146–155

Epidemiology of Intractable Epilepsy in Children Peter Camfield and Carol Camfield

Summary The epidemiology of intractable epilepsy in children is complex but potentially important in choosing candidates for epilepsy surgery. The overall incidence of epilepsy in children is estimated to be from about 30 to 80 cases per 100, 000 persons per year, and it is highest in the first year of life. Between ages 11 to 15 years the incidence is similar to adults. The prevalence is higher ( about 3-6 /1 ,000 ). Intractable, pharmacoresistant and resistant epilepsy have varying definitions in the published literature with the risk being between approximately 8 and 20% of children with epilepsy. About one-third or more of children with intractable epilepsy will eventually have a lengthy or permanent remission of their seizures, although others may have periods of seizure control mixed with periods of resistant epi lepsy. Accurate prediction of the outcome of intractability is currently impossible. It is statistically associated neurological deficits, intellectual disability, early high seizure frequency, a few epilepsy syndromes, temporal lobe epilepsy, and causative lesions on MRI. Importantly, about one-third of children with epilepsy and an MRI identified cause will still have remission. It is unclear if intractable epilepsy in children is associated with sudden unexpected deaths in epilepsy ( SUDEP ) even if epilepsy persists into adulthood.

Keywords: epidemiology, intractable, epilepsy, children , outcome

Introduction The basic epidemiology of epilepsy in children has been well described and fairly consistent throughout the world.1 In population - based studies, the reported overall incidence in children less than 16 years of age in developed countries has ranged from 33.3 to 82 cases per 100, 000 persons per year. In less developed countries, the incidence may be somewhat higher. All studies to date have found the incidence to be highest in the first year of life ranging from 81 per 100, 000 in Olmstead County, Rochester, Minnesota 2 to 118 per 100,000 in Nova Scotia3 and up to 130 per 100,000 in Iceland .4 Between 1 and 11 years of age, the incidence is fairly sta ble and does not show much variability.34 Between ages 11 and 15 years, the incidence appears to be lower but again stable at about 21 per 100, 000, an estimate that is similar to the incidence in adults.5 The prevalence of pediatric epilepsy is higher

than the incidence because epilepsy is often a chronic disorder. In Europe, the prevalence of epilepsy in children is estimated to be between 3.2 and 5.1 per 1.000.6 A U.S. national pediatric study estimated that the lifetime prevalence of active epilepsy was 6.3 per 1,000 ( 95% confidence interval , Cl: 4.9-7.8 ).7 Epilepsy was found to be more common in children from impoverished families. In Canada , a national survey showed a prevalence of epilepsy in children to be 5.26 per 1,000.8 Some but not all studies in less developed countries suggest a higher prevalence of epilepsy in children, possibly because of a higher chance of acquired brain disease from causes such as meningitis, cysticercosis, and head injury.9-14 In developing countries, there is a fairly consistent trend toward a higher prevalence in rural compared with urban areas.9 The subject of this book is pediatric epilepsy surgery, and the surgical team’s goal is to improve the outcome when seizures are not controlled by medication or other medical treatments. The seizures are said to be intractable, pharmacoresistant, or just resistant. In this chapter, we emphasize population - based studies of children with epilepsy that have considered intractable epilepsy. These include the studies from Nova Scotia, the Netherlands, Connecticut, Olmstead County in Rochester, Minnesota, and Turku in Finland . Some pertinent studies from each of these cohorts are outlined in Table 3.1 and Table 3.2. We consider the definitions of intractability, the way intracta bility may come and go, predicting intractability, and mortality associated with intractable epilepsy in children.

Definitions of Intractable Epilepsy There are three possible seizure outcomes in childhood epi lepsy: seizure freedom with or without antiepileptic drugs (AEDs ), persistent infrequent seizures , and intractable epilepsy. The distinction between persistent infrequent seizures and intractability is arbitrary. In terms of quality of life, even rare seizures have a significant effect such as precluding a number of professions , limiting certain activities and preventing driving. In a pivotal study of epilepsy surgery in 396 adults, it was noted that “ seizure freedom specifically is required for persistent improvements: those who do not become seizure free do not show a durable improvement in health - related quality of life ( HRQOL).” 32 A study of the HRQOL of 134 adults with refractory epilepsy treated with medication also noted that seizure freedom had a much greater effect than lesser reductions in

22

I  Introduction to Epilepsy in Children Table 3.1  Population-based outcome studies of intractable pediatric onset epilepsy (ages 1 month to 16 years at first seizure)

Terminal remission

Intractable at the end of follow-up

Comments/predictors

245 (150 incidence, 95 30 years prevalence cases) with all types of epilepsy 178 surviving and traced

74% (on or off AEDs)

23%

Predictive factors for drug resistant epilepsy: initial AED failure, status epilepticus, high initial seizure frequency, remote symptomatic etiology

Turku, Finland 201216

102 with drug resistant epilepsy within 2 years of epilepsy onset (42% of original cohort, 59 incidence, 43 prevalence cases)

41 years median (range: 4–55)

49% (on or off AEDs)

51%

Predictive factors for drug resistant epilepsy: initial AED failure, status epilepticus, high initial seizure frequency, remote symptomatic etiology

Nova Scotia 199617

504 incidence cases (excludes absence, infantile spasms, and myoclonic epilepsy)

≥ 5 years (average 8)

54% (off AEDs) 7.7%

Predictive factors for intractability: intellectual disability, poor seizure control in first year of treatment, complex partial seizures

Nova Scotia 200518

260 with focal and/or generalized convulsive seizures who attempted to discontinue AED treatment after ≥ 1 seizure free

6.5 years (average)

79% (off AEDs) 1%

Note: this study applies only to those who discontinue AED treatment, have seizures return that are not controlled (i.e., intractable) Predictors for later intractability: intellectual disability, poor seizure control in first year of treatment, complex partial seizures

Nova Scotia 200519

Complex partial seizures ± secondary generalized n = 108 Secondary generalized only n = 80

28 years (average)

57% (off AEDs) 24% 81% (off AEDs) 3%

Comparison of the two groups was highly statistically significant

Dutch 201020

431, includes all types of epilepsy

15 years

70.9%

8.5%

Predictors of intractability: nonidiopathic etiology, febrile seizures, no 3-month remission, early intractability

Connecticut 200621

527, includes all types of epilepsy

9.7 years (median)

Not stated in this chapter

23.2% (2 drug failure)

Predictors of intractability: epilepsy types: “catastrophic” > focal > idiopathic

Connecticut 200122

599, includes all types of epilepsy

4.8 years (median)

Not stated in this chapter

10% (2 drug failure and > 1 sz/ month over 18 months

Paper addressed “early intractability.” Predictors: syndrome groupings, early seizure frequency, EEG focal slowing

Connecticut 200923

128 with failure of 2 AEDs (23% of total cohort)

11.1 years (average)

22% ≥ 3 years terminal remission

61%

Predictors of remission: idiopathic syndromes and lower frequency of days with seizures

Rochester 201324

381

10.3 years (median)

19% “early” intractability (2 years after diagnosis) 12 years: (average) later 50% were still intractable Overall 9% had “enduring” intractability

Predictor of “enduring” intractability: causative MRI lesion

Rochester 201225

127 (epilepsy onset < 3 years of age)

78 months (median)

35%

Predictors of intractability: age of onset ≤ 12 months, developmental delay, neuroimaging abnormalities and focal EEG slowing

The studies and year published

Number of cases

Turku, Finland 199315

Follow-up

Abbreviations: AEDs, antiepileptic drugs; EEG, electroencephalography.

3  Epidemiology of Intractable Epilepsy in Children Table 3.2  Definitions of intractability

Study

Minimum number of AED efficacy failures

Seizure frequency

Comments

Connecticut21

2

1 seizure per month for ≥ 18 months and < 3 months seizure free during that period (within 3 years of diagnosis)

Requires very careful documentation of seizure free intervals

Nova Scotia17,​19

3

≥ 1 seizure every 2 months in the last year of follow-up

Requires very careful documentation of seizure frequency

Rochester25

2

≥ 1 seizure every 6 months at final follow-up, or epilepsy surgery after failure of ≥ 2 AEDs

May vary with the length of follow-up

Turku, Finland 199315

–

< 1-year remission during follow-up of ≥ 10 years

Requires very long follow-up

Turku, Finland 200926

≥1

< 1-year remission during follow-up of ≥ 10 years despite adequate treatment

Single drug failure is unique

Turku, Finland 201227

1 or 2

Adequate treatment trials of seizures for 2 years after onset of epilepsy, without a 1-year remission

Used to define “incident drug resistant”

Dutch20

–

Failure to be ≥ 3 months seizure free (> 6 months after diagnosis)

Does not account for treatment intensity

2

< 1 year terminal remission

Simple to apply

3

Average of ≥ 1 unprovoked seizure per month during ≥ 2 years

Does not assess treatment intensity

ILAE30

2

Failure of ≥ 2 adequate trials of tolerated AEDs

“Adequate” and “tolerated” are well defined Likely more useful for adults than children

Germany31

–

≥ 1 seizure every day or week

Used by the authors to assess those with intellectual disability; does not include treatment intensity

Scotland28 Hong Kong

29

Abbreviations: AEDs, antiepileptic drugs; ILAE, International League Against Epilepsy.

seizure frequency.33 However, the concept of intractability usually implies both persistent and “frequent seizures” and typically takes into account the intensity of medical treatment. There are many published definitions; see Table 3.2 for examples. Berg et al applied six definitions of intractability to the Connecticut cohort after 92% of patients had at least 5 years of follow-up.21 The rate of intractability ranged from 9 to 24% depending on the definition. The agreement between definitions ranged from a Kappa of 0.45 (moderate agreement) to 0.79 (strong agreement). The three definitions that had the greatest agreement were those from the Connecticut,22 Nova Scotia,17 and Scottish studies;34 the authors suggested that different definitions might be useful for different indications. In a thoughtful review, the ad hoc Task Force of the International League Against Epilepsy (ILAE) Commission on Therapeutic Strategies concluded that “drug resistant epilepsy may be defined as failure of adequate trials of two tolerated and appropriately chosen and used AED schedules (whether as monotherapies or in combination) to achieve sustained seizure freedom.”30 The insistence on “adequate,” “tolerated,” and “appropriately chosen” helps this definition to be particularly clinically relevant. The Task Force did not precisely define “sustained seizure freedom.” This definition was strongly influenced by the sentinel paper of Kwan and Brodie who studied 470 people with newly diagnosed epilepsy aged 9 to 93 years (mostly adults) referred to a single regional center in Scotland.28 During an average follow-up of 5 years (range: 2–16 years),

113 patients discontinued their first AED because it failed to control their seizures. Only 11% of these patients had successful control of their seizures with a second AED. However, in children the failure of one or two AEDs is not so alarming. In the Nova Scotia study 417 patients were assessed for failure/success of a first AED during the first year of treatment.35 If the first AED was successful, then only 4% developed intractable epilepsy during the next 8 years and 61% were able to eventually successfully discontinue AED treatment. If the first AED was unsuccessful for reasons of efficacy and a second AED treatment started, the risk of intractability rose from 4 to 29% (p ≤ 0.0001), but still notably 42% went on to have complete remission. In the Connecticut study, of 613 newly treated children, 128 were unsuccessfully treated with two AEDs.23 Seventy-three of those with two drug failures (57%) eventually had at least 1 year of remission from seizures after more than and equal to three AEDs; however, 50 went on to have at least one further relapse after the remission associated with the third AED. At last contact, 48 (38%) of those who failed two AEDs were in more than and equal to 1 year terminal remission. We conclude that failure of two AEDs for reasons of efficacy is associated with persistent epilepsy in children yet about one-third will eventually have a remission with a third or more AED drugs. Therefore failure of a second AED is not a robust definition of intractability in children: additional factors must be considered. Table 3.1 outlines the definitions used by each of the population-based and several other studies which estimate that

23

24

I  Introduction to Epilepsy in Children the overall rate of intractability lies between about 8 and 16%. It is not clear which definition is most useful. A study from the Canadian province of Saskatchewan examined four ­definitions36—the ILAE,30 the Nova Scotian,17 the Connecticut,22 and the Scottish,28 and each definition was applied to a cohort of 250 primarily adult patients with varying severities of epilepsy. All patients were categorized as intractable or not intractable based on each of the definitions. “Interobserver agreement was moderate using the Connecticut ( j = 0.56) and Scottish ( j = 0.58) definitions and robust for the Nova Scotia definition ( j = 0.69) and ILAE consensus ( j = 0.77) definitions.”36 There was little difference in the patients identified by each scheme.

„„ Intractability that Comes and Goes One of the remarkable findings in several of the p ­ opulationbased studies of childhood epilepsy is that an i­ndividual may have both periods of intractable epilepsy and periods of remission. Usually intractable epilepsy appears early the clinical course and persists; however, in a number of patients intractability may vanish over time or suddenly appear after years without seizures. A clinic-based study was the first to really point this out. In the remarkable study of Huttenlocher et al, 167 children with unequivocal intractable epilepsy (≥ 1 seizures/month for ≥ 2 years, despite “appropriate anticonvulsant agents at maximum tolerated blood levels”) were identified from a clinic population in Chicago.37 Patients were excluded if they had an IQ less than 30, a brain tumor or age of onset greater than 13 years of age. The authors were able to follow-up 155/167 (93%). Ten had epilepsy surgery with 145 remaining to assess the outcome of medical treatment. Good outcome was defined as less than or equal to one seizure/year at the end of follow-up. Since the length of follow-up varied, analysis was carried out with survival curves which indicated that after 18 years of follow-up, only 25% of those with normal intelligence continued to be intractable, although 70% of those with intellectual disability continued to have intractable epilepsy (30% did not). The longer the follow-up, the better was the outcome. For those with normal intelligence, 30% still had more than one seizure/week after 5 years; however, after 14 years only 5% had this seizure frequency. Some patients with etiologies known to do poorly did remarkably well. For example, four of five with unilateral Sturge–Weber disease became seizure free with medical treatment only. Subsequently, several of the population-based studies have confirmed and amplified the findings from Chicago. In the Nova Scotia study at an average of 7.5 years of follow-up after their first seizure, 39 (8%) met the criteria for intractability.17 An average of 5 years later (i.e., 12.5 years after diagnosis), 7/39 (18%) had a terminal remission (seizure free) of at least 1 year. However, early complete seizure control does not guarantee a long-term good outcome. In Nova Scotia there were 296 patients with normal intelligence and a follow-up averaging of 28 ±4 years. There were seven patients (2%) who had a long, seizure-free remission (median: 5 years; range: 5–18 years) and then relapsed with intractable epilepsy until the end of

follow-up. Their epilepsies were focal in five, symptomatic generalized in one, and unclassified in one.38 In the Dutch study, intractability was considered at two points—after 5 years of follow-up after diagnosis and after 15 years.20 For each time point, intractability was considered for the last year of follow-up (i.e., between year 4 and 5 and between year 14 and 15). Follow-up at 15 years was remarkably complete. Of 453 with a 5 year follow-up, 431 (95%) provided questionnaire data after 15 years. At 5 years, 34 (8%) were considered intractable and of these 34, only 19 (56%) were still intractable at 15 years. Of the 379 not intractable at 5 years, 16 (4%) had become intractable by 15 years. The total number with intractability at 15 years was 35 (8%)—the same proportion as at 5 years except about one half of this group were different patients. Thus intractability often vanished between 5 and 15 years or appeared during the same interval. In one of the publications from Connecticut study, 527 patients were described with a follow-up of about 10 years.16 Eighty-two (14%) had intractable epilepsy at some point in their clinical course. In 26 (32%) intractable epilepsy took at least 3 years to appear after treatment had started. Twenty-one percent of those who developed intractable epilepsy went on to an eventual remission. Finally, in the Finnish study 102 incidence patients were selected with new onset focal or generalized tonic-clonic seizures, failure of greater than or equal to two AEDs and ­followup for greater than 10 years.27 In fact, the median follow-up was 40.5 years. After developing drug resistant epilepsy, 3 of 102 (3%) had a 1-year remission, 11 (11%) had at least one 2-year remission, 18 (18%) had at least one 5-year remission, and 52 (52%) had a greater than or equal to 5-year terminal remission. Only 18 (18%) had no subsequent remission of greater than or equal to 1 year. Therefore, it is clear that with prolonged follow-up many patients, perhaps the majority, who have drug resistant or intractable epilepsy will experience significant ­seizure-free periods. Sometimes these seizure free periods last a year or 2 and are followed by relapse, but not infrequently the epilepsy seems to resolve completely. Consideration of epilepsy surgery must take into account that the natural history of a given patient with intractability may eventually include a significant/ worthwhile remission without surgery; however, at present it is impossible to know which patients fall in this category. It is also apparent that patients may be seizure free for many years and then develop intractable epilepsy. There are no absolutes.

„„ Predicting Intractable Epilepsy For an individual child at or close to the time of the diagnosis of epilepsy, it is very difficult to accurately predict the long-term outcome. The Dutch and Nova Scotia cohorts were combined to provide a sample of 1,055 newly diagnosed pediatric patients with a minimum follow-up of 5 years.39 Several statistical models were developed to try to predict terminal remission or no terminal remission based on many clinical variables available at diagnosis or after 6 months of treatment. Remission status was correctly predicted for only 70%. The predictions for 30% were incorrect. It is therefore not surprising that predicting intractability is equally challenging.

3  Epidemiology of Intractable Epilepsy in Children In the Connecticut study, early (within 5 years of diagnosis) intractable epilepsy was associated in multivariate analysis with some epilepsy syndromes, initial high seizure frequency, focal electroencephalography (EEG) slowing and acute symptomatic or neonatal status epilepticus.22 In the Rochester study, early intractability occurred in 19.7%.24 After a median follow-up of 11.7 years, medical intractability persisted in 49% (i.e., 9.6% of the overall cohort). If there was a causative MRI lesion and early intractability, the chance of “enduring” intractable epilepsy was estimated as 91% (95% CI: 77–97%) and for those without an MRI lesion the chance of ‘enduring’ intractable epilepsy was 40% (CI: 26–36%). A separate publication from the Rochester study, examined the 127 children with onset of epilepsy less than 3 years of age and found that 35% had intractable epilepsy which was associated in multivariate analysis with age of onset less than or equal to 12 months, developmental delay, neuroimaging abnormalities, and focal EEG slowing.25 In the Nova Scotia cohort intellectual disability increased the risk of intractable epilepsy by 18 times (95% CI: 8.2–39.6) over those with normal intelligence.17 In the Finnish study, weekly seizures during the first year after diagnosis was associated with an eight-fold increase (hazard ratio 8.2; CI: 1.6–42) in drug resistant epilepsy 40 years later compared to those with early less frequent seizures.26 A prospective study from the Northwest Sector of Hong Kong was devoted to the development and prediction of intractability.29 Patients were recruited from the only referral hospital for that region, although it is unclear if all children with epilepsy are eventually referred to this center. Intractability was defined as at least one seizure/month for 2 years despite treatment with three AEDs. The sample consisted of 309 children with epilepsy who were prospectively recruited and 44 (14.2%) met the criteria for intractability. The overall length of follow-up was not reported; however, for those with intractability, 54% had focal seizures, infantile spasms 23%, mixed seizure types 11%, and not reported 12%. The 44 with intractable epilepsy were compared to the 265 without intractability. On univariate analysis significant predictors were age of onset less than 1 year of age, daily seizures before AED treatment, more than 20 seizures before treatment, infantile spasms and mixed seizure types, symptomatic etiology, abnormal neurological function (including intellectual disability), neonatal seizures, abnormal initial EEG, status epilepticus, greater than or equal to 3 seizures during the first 6 months of treatment. On multivariate analysis, the four factors retained in the model were abnormal neurological status (which includes intellectual disability), daily seizures before treatment and breakthrough seizures in the second 6 months of treatment, and previous febrile seizures (length of the febrile seizures was not stated). We conclude that across most studies there is a strong association on univariate analysis between intellectual disability, neurological abnormalities and intractable epilepsy. Early high seizure frequency has also been frequently associated with intractability. Epilepsy with a known cause is less likely to remit and a causative MRI lesion makes intractability more likely. However, it is important to note that in the Rochester study a causative MRI lesion was still associated with a 32% chance of long term remission without surgery. In addition, in the Connecticut study 33% with a causative MRI lesion had a “complete remission” (≥ 5 years drug free and ≥ 5 years seizure free), again without surgery.40–41

„„ Epilepsy Syndromes Fewer than half of children in the Rochester study had an identifiable epilepsy syndrome.42 It is disappointing that few epilepsy syndromes have an absolute prognosis, that is, virtually always remit or virtually always continue with intractable epilepsy. Rolandic, Panayiotopoulos, and several neonatal or early infantile syndromes are the only syndromes always associated with remission.43 Surprisingly, few of the epileptic encephalopathies virtually always show long-term intractability (e.g., Dravet syndrome and migrating focal epilepsy). One quarter to one half of children with infantile spasms will eventually have a complete remission.44 Even in Lennox Gastaut syndrome a few patients will become completely seizure free.45 The rate of intractability associated with temporal lobe epilepsy in children is of particular interest to the question of epilepsy surgery in children. In population-based studies it is not always easy to be sure that a child’s epilepsy originates from the temporal lobe unless there is a causative lesion on MRI or the seizures are sufficiently frequent to permit video EEG. Nonetheless, it is likely that the temporal lobe is the source of origin for the majority of children with what used to be called complex partial seizures or dyscognitive seizures and most recently focal, nonmotor seizures with impaired awareness.46 In the Nova Scotia study we identified all of the children with normal intelligence, no epilepsy syndrome and focal seizures.19 We compared the 80 with focal seizures with secondary generalization to the 108 with complex partial ± secondary generalization. Follow-up averaged 29 ± 5 years. It is probable that most of those in the complex partial group had temporal lobe epilepsy. Compared with the group with secondarily generalized seizures, the group with complex partial seizures was much more likely to have intractable seizures at the end of follow-up or have undergone epilepsy surgery (36% vs. 5%; p < 0.00001). Clearly, not all children with probable temporal lobe epilepsy will have refractory epilepsy. A community-based sample from the state of Victoria, Australia identified 77 children with newly diagnosed temporal lobe epilepsy who were then followed for a median of 14 years.47 At the end of follow-up 43 (56%) were not seizure free or had undergone temporal lobectomy. Of the 77, 28 had a temporal lesion on MRI and none of these patients had a terminal remission unless they had undergone surgery. Thus, temporal lobe epilepsy is relatively often intractable and an MRI lesion appears to be strongly predictive of persistent or intractable seizures.

„„ Mortality Children with epilepsy have a 5 to 10 times risk of death by early adulthood compared with the general population.48–49 Almost all of this excess mortality is caused by comorbid conditions such as bulbar dysfunction in children with spastic quadriplegia from birth asphyxia: seizures play no causative role.50 A small number of deaths are directly related to epilepsy. Accidents such as falls during a seizure may cause injuries but rarely death. Drowning is a more common cause of death.51 SUDEP is perhaps the most feared “cause,” but it is thankfully very rare in children. A combined analysis of the Connecticut, Rochester, Dutch, and Nova Scotia cohorts (total n = 2,239) suggested that

25

26

I  Introduction to Epilepsy in Children the risk of SUDEP in childhood onset epilepsy was only 10 per 30,000 patient years.48 The major risk factors for SUDEP are generalized tonic clonic seizures (especially during sleep), frequent seizures, and long duration of epilepsy.52 In adults the highest risk of SUDEP is in those under consideration for epilepsy surgery or those who failed to gain seizure control with epilepsy surgery.53 A very striking finding in most studies about SUDEP is that the risk increases with the duration of epilepsy.52 A well-known publication from the Finnish cohort suggested that the risk of SUDEP increased as children with persistent epilepsy became young adults.51 The overall life time risk of SUDEP with childhood onset epilepsy was estimated at 7%. This cohort is complicated because it included at inception both incidence and prevalence cases and there was a high rate of “remote symptomatic epilepsy” (3 times the rate of the Nova Scotia cohort) (unpublished data). In the Nova ­Scotia cohort we have not noted the same risk of later SUDEP (­unpublished data). The risk of SUDEP associated with intractable epilepsy in children has not been documented. The rate of SUDEP was reported in a sample of 1974 children over a 9-year period cared for at a tertiary level, epilepsy referral center in Denmark.53 The center was described as “the only of its kind in Denmark.” There were a total of 43 deaths including 9 cases of SUDEP. When the 9 who died from neurometabolic disorders were excluded, 30 of the remaining 34 patients who died had intellectual disability and 31 had drug resistant epilepsy. As noted above, intellectual disability is well known to be associated with both

severe epilepsy and death from comorbid conditions, so this finding is not unexpected. Of the 9 with SUDEP, 7 had drug resistant epilepsy and 7 had intellectual disability. Epilepsy syndromes of those with SUDEP were not well described but all had generalized tonic clonic seizures “among other seizure types.” Therefore, in this study there seems to be a strong association between death and resistant epilepsy but the association is not specific for SUDEP. The argument that prevention of SUDEP by early surgery in children with intractable epilepsy seems reasonable but remains unproven.

„„ Conclusion An ideal definition of intractability in childhood onset epilepsy has still to be developed. Based on population-based studies, about 10% of children with new onset epilepsy will develop long-lasting intractable epilepsy. Although the intractability is most often early in the clinical course and persistent, in at least one-third, intractability will resolve with medical treatment only. It is unclear if these remissions reflect the natural history of epilepsy or are the result of AED treatment. For some patients intractability may develop after many years of remission from seizures. Intractable epilepsy is associated with a high frequency of initial seizures, neurological or cognitive deficits, and a known etiology. SUDEP is associated with intractable epilepsy but has not been proven to be prevented by epilepsy surgery in children.

References 1. Camfield P, Camfield C. Incidence, prevalence and a ­ etiology of seizures and epilepsy in children. Epileptic Disord 2015;17(2): 117–123 2. Hauser WA, Annegers JF, Kurland LT. Incidence of epilepsy and unprovoked seizures in Rochester, Minnesota: 1935–1984. Epilepsia 1993;34(3):453–468 3. Camfield CS, Camfield PR, Gordon K, Wirrell E, Dooley JM. Incidence of epilepsy in childhood and adolescence: a population-based study in Nova Scotia from 1977 to 1985. Epilepsia 1996;37(1):19–23

11. Lavados J, Germain L, Morales A, Campero M, Lavados P. A descriptive study of epilepsy in the district of El Salvador, Chile,1984–1988. Acta Neurol Scand 1992;85(4): 249–256 12. Burneo JG, Tellez-Zenteno J, Wiebe S. Understanding the burden of epilepsy in Latin America: a systematic review of its prevalence and incidence. Epilepsy Res 2005;66(1–3):63–74 13. Mac TL, Tran DS, Quet F, Odermatt P, Preux PM, Tan CT. Epidemiology, aetiology, and clinical management of epilepsy in Asia: a systematic review. Lancet Neurol 2007;6(6):533–543

4. Olafsson E, Ludvigsson P, Gudmundsson G, Hesdorffer D, Kjartansson O, Hauser WA. Incidence of unprovoked sei­ zures and epilepsy in Iceland and assessment of the epilepsy syndrome classification: a prospective study. Lancet Neurol 2005;4(10):627–634

14. Garofalo NG, Garcia AMG, Concepcion OF, Camfield CS, Camfield PR. Prevalence, syndromes and severity of childhood epilepsy in Cuba. J Pediatr Neurol 2012;10:1–6

5. Wirrell EC, Grossardt BR, Wong-Kisiel LCL, Nickels KC. Incidence and classification of new-onset epilepsy and epilepsy syndromes in children in Olmsted County, Minnesota from 1980 to 2004: a population-based study. Epilepsy Res 2011;95(1–2):110–118

16. Berg AT, Vickrey BG, Testa FM, et al. How long does it take for epilepsy to become intractable? A prospective investigation. Ann Neurol 2006;60(1):73–79

6. Forsgren L, Beghi E, Oun A, Sillanpää M. The epidemiology of epilepsy in Europe: a systematic review. Eur J Neurol 2005;12(4):245–253 7. Russ SA, Larson K, Halfon N. A national profile of childhood epilepsy and seizure disorder. Pediatrics 2012;129(2):256–264 8. Prasad AN, Sang X, Corbett BA, Burneo JG. Prevalence of childhood epilepsy in Canada. Can J Neurol Sci 2011;38(5):719–722 9. Ngugi AK, Bottomley C, Kleinschmidt I, Sander JW, Newton CR. Estimation of the burden of active and life-time epilepsy: a meta-analytic approach. Epilepsia 2010;51(5):883–890 10. Benamer HTS, Grosset DG. A systematic review of the epidemiology of epilepsy in Arab countries. Epilepsia 2009;50(10):2301–2304

15. Sillanpää M. Remission of seizures and predictors of intractability in long-term follow-up. Epilepsia 1993;34(5):930–936

17. Camfield PR, Camfield CS. Antiepileptic drug therapy: when is epilepsy truly intractable? Epilepsia 1996;37(Suppl  1): S60–S65 18. Camfield P, Camfield C. The frequency of intractable seizures after stopping AEDs in seizure-free children with epilepsy. Neurology 2005;64(6):973–975 19. Camfield CS, Camfield PR. The adult seizure and social outcomes of children with partial complex seizures. Brain 2013;136 (Pt 2):593–600 20. Geerts A, Arts WF, Stroink H, et al. Course and outcome of childhood epilepsy: a 15-year follow-up of the Dutch Study of Epilepsy in Childhood. Epilepsia 2010;51(7):1189–1197 21. Berg AT, Kelly MM. Defining intractability: comparisons among published definitions. Epilepsia 2006;47(2):431–436

3  Epidemiology of Intractable Epilepsy in Children 22. Berg AT, Shinnar S, Levy SR, Testa FM, Smith-Rapaport S, Beckerman B. Early development of intractable epilepsy in ­ children: a prospective study. Neurology 2001;56(11):1445–1452 23. Berg AT, Levy SR, Testa FM, D’Souza R. Remission of epilepsy after two drug failures in children: a prospective study. Ann Neurol 2009;65(5):510–519 24. Wirrell EC, Wong-Kisiel LC, Mandrekar J, Nickels KC. What predicts enduring intractability in children who appear ­medically intractable in the first 2 years after diagnosis? Epilepsia 2013;54(6):1056–1064 25. Wirrell E, Wong-Kisiel L, Mandrekar J, Nickels K. Predictors and course of medically intractable epilepsy in young children presenting before 36 months of age: a retrospective, population-based study. Epilepsia 2012;53(9):1563–1569 26. Sillanpää M, Schmidt D. Early seizure frequency and aetiology predict long-term medical outcome in childhood-onset epilepsy. Brain 2009;132(Pt 4):989–998 27. Sillanpää M, Schmidt D. Is incident drug-resistance of childhood-onset epilepsy reversible? A long-term follow-up ­ study. Brain 2012;135(Pt 7):2256–2262 28. Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000;342(5):314–319 29. Kwong KL, Sung WY, Wong SN, So KT. Early predictors of medical intractability in childhood epilepsy. Pediatr Neurol 2003;29(1):46–52 30. Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 2010;51(6):1069–1077 31. Steffenburg U, Hedström A, Lindroth A, Wiklund LM, ­Hagberg G, Kyllerman M. Intractable epilepsy in a population-based series of mentally retarded children. Epilepsia 1998;39(7):767–775 32. Spencer SS, Berg AT, Vickrey BG, et al; Multicenter Study of Epilepsy Surgery. Health-related quality of life over time since resective epilepsy surgery. Ann Neurol 2007;62(4):327–334 33. Birbeck GL, Hays RD, Cui X, Vickrey BG. Seizure reduction and quality of life improvements in people with epilepsy. Epilepsia 2002;43(5):535–538 34. Kwan P, Brodie MJ. Drug treatment of epilepsy: when does it fail and how to optimize its use? CNS Spectr 2004;9(2):110–119 35. Camfield PR, Camfield CS, Gordon K, Dooley JM. If a first antiepileptic drug fails to control a child’s epilepsy, what are the chances of success with the next drug? J Pediatr 1997;131(6):821–824 36. Téllez-Zenteno JF, Hernández-Ronquillo L, Buckley S, Zahagun R, Rizvi S. A validation of the new definition of drug-resistant epilepsy by the International League Against Epilepsy. Epilepsia 2014;55(6):829–834

37. Huttenlocher PR, Hapke RJ. A follow-up study of intractable seizures in childhood. Ann Neurol 1990;28(5):699–705 38. Camfield PR, Camfield CS. Intractable seizures after a lengthy remission in childhood-onset epilepsy. Epilepsia 2017;58(12): 2048–2052 39. Geelhoed M, Boerrigter AO, Camfield P, et al. The accuracy of outcome prediction models for childhood-onset epilepsy. Epilepsia 2005;46(9):1526–1532 40. Berg AT, Testa FM, Levy SR. Complete remission in nonsyndromic childhood-onset epilepsy. Ann Neurol 2011;70(4):566–573 41. Dhamija R, Moseley BD, Cascino GD, Wirrell EC. A ­populationbased study of long-term outcome of epilepsy in childhood with a focal or hemispheric lesion on neuroimaging. Epilepsia 2011;52(8):1522–1526 42. Wirrell EC. Predicting pharmacoresistance in pediatric epilepsy. Epilepsia 2013;54(Suppl 2):19–22 43. Camfield P, Camfield C. Childhood epilepsy: what is the evidence for what we think and what we do? J Child Neurol 2003;18(4):272–287 44. Riikonen R. Infantile spasms: therapy and outcome. J Child Neurol 2004;19(6):401–404 45. Kim HJ, Kim HD, Lee JS, Heo K, Kim DS, Kang HC. Long-term prognosis of patients with Lennox—Gastaut syndrome in recent decades. Epilepsy Res 2015;110:10–19 46. Fisher RS, Cross JH, French JA, et al. Operational classification of seizure types by the International League Against Epilepsy: position Paper of the ILAE Commission for classification and terminology. Epilepsia 2017;58(4):522–530 47. Spooner CG, Berkovic SF, Mitchell LA, Wrennall JA, Harvey AS. New-onset temporal lobe epilepsy in children: lesion on MRI predicts poor seizure outcome. Neurology 2006;67(12): 2147–2153 48. Berg AT, Nickels K, Wirrell EC, et al. Mortality risks in new-onset childhood epilepsy. Pediatrics 2013;132(1):124–131 49. Christensen J, Pedersen CB, Sidenius P, Olsen J, Vestergaard M. Long-term mortality in children and young adults with epilepsy—A population-based cohort study. Epilepsy Res 2015;114:81–88 50. Camfield CS, Camfield PR, Veugelers PJ. Death in c­hildren with epilepsy: a population-based study. Lancet 2002; 359(9321):1891–1895 51. Sillanpää M, Shinnar S. Long-term mortality in childhood-onset epilepsy. N Engl J Med 2010;363(26):2522–2529 52. Donner EJ, Camfield P, Brooks L, et al. Understanding death in children with epilepsy. Pediatr Neurol 2017;70:7–15 53. Tomson T, Surges R, Delamont R, Haywood S, Hesdorffer DC. Who to target in sudden unexpected death in epilepsy prevention and how? Risk factors, biomarkers, and intervention study designs. Epilepsia 2016;57(Suppl 1):4–16

27

4

  Genetics in Epilepsy Surgery Maurits W. C. B. Sanders, Floor E. Jansen, Bobby P. C. Koeleman, and Kees P. J. Braun

Summary The role of genetics in the etiology of generalized as well as focal epilepsies has become increasingly acknowledged in the last decades. Since the identification of CHRNA4 (cholinergic receptor nicotinic alpha 4 subunit gene) as the first pathogenic gene variant in epilepsy, many other disease-causing variants in ion-coding and non-ion coding genes have been identified. The introduction of next-generation sequencing (NGS) in 2008 allowed sequencing of multiple genes, the whole exome or the whole genome in a single analysis. As a result the genetic basis of numerous structural and nonstructural epilepsy syndromes becomes progressively unraveled. Within the group of focal epilepsies, the most important novel gene discoveries have been made in the mammalian target of rapamycin (mTOR) pathway. Mutations in this pathway cause focal epilepsies due to MCDs, including tuberous sclerosis complex (TSC) and FCD (in particular type II), as well as nonlesional focal epilepsy subtypes. Another broad clinical spectrum of nonlesional epilepsies with a genetic basis comprises the early onset epileptic encephalopathies, with SCN1A, KCNQ2, KCNT1, ARX, CDKL5, SLC25A22, SPTAN1 STXBP1, MECP2, and GRIN2A as important causative variants. Efforts to identify epilepsy-causing genes and elucidating its neurobiological framework are driven by the goal of improving epilepsy care. The finding of a genetic cause of epilepsy during the presurgical trajectory could either prevent unnecessary invasive diagnostics or even resective surgery (e.g. if the pathogenic variant points toward a genetic syndrome without a focal structural cause), or encourage further presurgical diagnostic exploration (e.g. if the pathogenic variant points toward a structural epilepsy that is caused by an “MRI-invisible” microstructural lesion). In addition, the identification of a causative pathogenic variant could guide prognostic and genetic counseling. To date, genetic diagnostics are not routinely performed within the presurgical evaluation trajectory, and the importance of genetic testing in patients who are considered possible candidates for epilepsy surgery needs to be further elucidated. Keywords:  gene discovery, mTOR-pathway, presurgical (genetic) testing, epilepsy surgery

„„ Introduction The role of genetics in the etiology of generalized as well as focal epilepsies has become increasingly acknowledged in the last decades. Linkage studies of familial epilepsies provided

the basis for human epilepsy gene discovery. In 1995, Sanger sequencing, first described in 1977 and the most widely used sequencing method for decades, enabled the identification of CHRNA4 as the first pathogenic gene variant in epilepsy related to the phenotype of autosomal-dominant nocturnal frontal lobe epilepsy, more recently renamed as sleep-related hypermotor epilepsy (SHE).1 Many disease-causing variants in other genes, initially consisting of genes related to channel function, have been identified since. Examples of these, in chronological order of discovery, are SCN1B, KCNQ2, KCNQ3, SLC2A1, SCN1A, GABRA1, and GABRG2.2–8 These discoveries would eventually form the genetic basis for human, previously called idiopathic ­epilepsy. Since the early 2000s, microarray techniques have been widely used to detect large (> 1 kb) deletions and duplications, so called copy number variants (CNVs), of small parts of chromosomes that can include single or multiple genes. The introduction of array comparative genomic hybridization and single nucleotide polymorphism microarrays led to an efficient systematic discovery of CNVs in patients with epilepsy. Literature has shown that pathogenic CNVs can be identified in over 5% of patients with different major classes of epilepsy, such as populations with generalized epilepsy, focal epilepsy, or epileptic encephalopathy.9–11 Furthermore, some CNVs that are not primarily considered to be the cause of the epilepsy syndrome are more commonly found in patients with epilepsy compared to controls and are therefore likely to act as risk factors for developing epilepsy. These CNVs are referred to as susceptibility CNVs. Well-known examples of these are deletions on chromosomes 15q11.2, 15q13.3, and 16p13.11 and deletions and duplications on chromosome 1q21.1 and 16p11.2.12–14 The introduction of NGS in 2008 represented an entirely new principle of sequencing technology following Sanger (first generation) sequencing. It became possible to sequence multiple genes, the whole exome or the whole genome in a single analysis. This enabled the identification of many novel candidate genes for epilepsy on an exome wide scale. The introduction of NGS accelerated discovery of novel epilepsy genes in different cohorts of patients that were often selected based on their phenotype.15–16 In particular whole exome sequencing (WES), through which the coding sequence of all genes in the human genome can be scanned for apparent disease-causing sequence variations, accelerated discovery of novel epilepsy genes in the past decade. The introduction of WES solved the genetic basis of many sporadic epilepsies in severely affected individuals, in particular the so-called epileptic encephalopathies. In these individually rare epilepsies, a genetic basis was long expected, but remained elu-

4  Genetics in Epilepsy Surgery sive due to difficulty of gene discovery. Through WES it was shown that a major proportion of these patients carried a de novo mutation, a new mutation occurring in a germ cell of one of the parents that proved to be causative for epileptic encephalopathy. This new paradigm solved many syndromic forms of epilepsy, and today, more than 500 genes are known to be causally related to epilepsy.17 These comprise genes that cause so-called “epilepsy-only” phenotypes, and genes that underlie syndromes with epilepsy as the core symptom or just as one of many manifestations. Furthermore, metabolic disorders and malformations of cortical development (MCDs), both well-known e ­tiological categories of epilepsies, may be caused by single gene ­mutations.17 Within the group of focal epilepsies, the most important novel gene discoveries have been made in the mTOR pathway. Mutations in this pathway cause focal epilepsies due to MCDs, for example, TSC as well as nonlesional focal epilepsy subtypes.16,​18 In addition, WES allowed the identification of novel somatic mutations; these are post-zygotic mutations detected in the abnormal (resected) tissue of 30 to 50% of MCD cases.16,​18–19

„„ Neurobiological Pathways in Genetic Epilepsy Gene products in epilepsy are part of different pathways. The first human epilepsy gene discovery led to the concept of different channelopathies as pathophysiological mechanisms for the development of epilepsy.15,​20 In 2002, however, LGI1 was discovered as a non-ion-coding causative gene in autosomal-dominant partial epilepsy with auditory features (ADPEAF), a rare form of lateral temporal lobe epilepsy. Since then, many non-ion-channel-coding genes for different epilepsy phenotypes have been identified, implicating a broad range of disease mechanisms. Important pathways, apart from those that involve ion channels, include synaptic transmission dysfunction, transporter defects, transcriptional dysregulation, impaired DNA repair, axon myelination, interneuron migration, interneurite formation, and cell–cell adhesion and metabolic defects (Table 4.1).21 Gene products within the same pathway can present with an indistinguishable phenotype. For example, DEPDC5 was

Table 4.1  Major epilepsy associated genes classified according to assumed neurobiological pathway (Based on McTague et al)21

Epilepsy gene

Neurobiological pathway

SLC2A1

Glucose transporter

Blood–brain barrier

SLC35A2; ST3GAI3; SLC12A5; SLC13A5; SLC25A22; QARS, AARS; PIGA; PIGQ; PIGO

Protein translation and modification

Neuron

CDKL5; FOXG1; ARX; MEF2C; PNKP; CHD2; SIK1; CDKL5; MECP2; LGI1

Transcription, DNA repair and modification, synaptogenesis

PIK3CA; PIK3R2; AKT3; MTOR; DEPDC5; NPRL2; NPRL3; PI3K/Akt/mTOR pathway: cell growth, synaptic PTEN; TSC1; TSC2 plasticity, etc. ARX; LIS1; DCX; TUBA1A; TUBB2B; GPR56; FLNA; NEDD4L; RELN, CNTNAP2

Neuronal migration, neurogenesis

PCDH19

Cell–cell adhesion

WWOX; WDR45

Apoptosis, autophagy

ALG1

Glycosylation

CHD2; IQSEC2; PNPO; PRICKLE1; HDAC4

Other

SPTAN1

Axon myelination

SCN1A; SCN2A; SCN8A; SCN1B; SCN9A

Na channel

KCNQ2; KCNQ3

K+ channel

CACNA1A; CACNA2D2

Ca+ channel

DNM1; NECAP1; TBC1D24;

Eco-endocytic cycling of vesicles

STXBP1; GOSR2; PRRT2

SNARE complex

KCNT1

K+/Na+ channel

KCNA2; KCNB1

K+ channel

GABRG2; GABRA1; GABRB3; GABRD

GABAA receptor

GRIN1; GRIN2A; GRIN2B

NMDA receptor

CHRNA4; CHRNA2; CHRNB2; PRIMA1

nACh receptor

Axon

+

DOCK7; SYNGAP1; ARHGEF9; GNAO1; PLCB1

G-protein-coupled-­transduction

IP3; DAG: PKC

Second messengers

HCN1

K+/Na+ (HCN) channel

Presynaptic membrane

Postsynaptic membrane

29

30

I  Introduction to Epilepsy in Children identified as a causal gene in the mTOR pathway causing focal epilepsy associated with cortical brain malformations.22 Sequencing of other genes in this pathway led to the identification of pathogenic variants in NPRL2 and NPRL3, which together with DEPDC5 are the components of the GATOR1 ­complex.23 Involvement of any of the genes of this complex is associated with similar types of epilepsy. To date, mutations in the GATOR1 complex genes are thought to explain at least 11% of focal epilepsies.18 On the other hand, a remarkable and well-described phenomenon associated with almost all epilepsy genes is their variable clinical expression, or phenotypic heterogeneity. This variability in phenotype can be subtle, but can also include the two extremes of disease severity. For example, pathogenic variants in the gene encoding the voltage-gated sodium channel (SCN1A) are related to epilepsy syndromes characterized by fever sensitive seizures, ranging from relatively mild phenotypes apparent in families with genetic epilepsy with febrile seizures “plus” (GEFS+), to severe epileptic encephalopathy (Dravet syndrome, or severe myoclonic epilepsy of infancy; SMEI).24 An extensive body of research on this subject has been published and to date more than 600 sequence variants of the SCN1A gene have been identified, enabling detailed genotype– phenotype correlations that explain the SCN1A phenotypic heterogeneity by focusing on the mutation type and location.25–26 Other genes with reported mutations related to this phenotype are SCN1B and GABRG2.2 Self-limiting epilepsies in newborns and infants represent a group of relatively rare syndromes that are transmitted with an autosomal-dominant inheritance pattern and are characterized by clusters of focal seizures occurring in the first year of life and eventually disappearing spontaneously. These epilepsies have been associated with pathogenic variants in the KCNQ2 and KCNQ3 gene (encoding the voltage dependent K+ channel) related to the clinical syndrome of benign familialneonatal seizures.27 However, de novo mutations in specific domains of KCNQ2 can lead to a very severe epileptic encephalopathy, with onset in the neonatal period. Phenotypic heterogeneity has even been reported within single families. For example, family members that carry the same pathogenic variants in KCNT1 may present with either autosomal-dominant sleep-related hypermotor epilepsy (ADSHE) or the severe and early onset epilepsy of infancy with migrating focal seizures. Variability of the severity of the disease is not always completely explained by type and location of the mutation, which led to speculation of disease modifiers. This is especially true for Dravet syndrome caused by SCN1A mutations. Several GEFS+ families have been described in which a single family member who suffers from Dravet syndrome carries the same mutation as family members with the more benign GEFS+ phenotype.28 Also, within sporadic Dravet patients a range of disease severity can be observed that is unexplained by type and location of the mutation. The hypothesis of a second hit in a modifier gene is supported by several studies implicating a range of genes. Finally, a recent study presented evidence that ~8% of Dravet patients are mosaic for the SCN1A mutation and therefore have milder phenotypes.28 For several known pathogenic variants, genotype–phenotype correlations remain unclear.29 In general, the mechanisms underlying phenotypic heterogeneity may be related to somatic mosaicism, variant type, other genetic/epigenetic factors, or differential expression of alternative gene transcripts.

„„ Genetic Principles in Nonstructural and Structural Epilepsies Epilepsy has been historically divided into two main categories, referring to its etiology: (1) (presumed) symptomatic and (2) “idiopathic” epilepsy. Symptomatic epilepsies are caused by an obvious structural brain abnormality, including both MCDs and acquired brain lesions, or by a metabolic, inflammatory, or degenerative disorder. Epilepsy was classified as “presumed symptomatic” (or “cryptogenic”) when a brain abnormality was suspected, but could not be detected by available diagnostic investigations. Symptomatic epilepsies can either have a g ­ enetic basis, or can be determined by external factors, or have a combined origin. The remainder of the patients were in the past classified as “idiopathic epilepsies” when there was no obvious known underlying cause, other than a presumed genetic predisposition. Over the past decades, studying genetics in epilepsy in both categories proved essential for clinical purposes as well as for understanding of the underlying mechanisms.30 Nowadays, the new International League Against Epilepsy classification requires a thorough diagnostic workup in all epilepsy patients for the following etiological categories: (1) structural, (2) genetic, (3) metabolic, (4) infectious, (5) immune, and (6) unknown cause, and the previous classification of epilepsies into symptomatic or idiopathic subtypes has been abandoned.31

Nonstructural Epilepsies The genetic basis of numerous nonstructural epilepsy syndromes becomes progressively unraveled, varying from confirmed monogenic causes (single-gene defects) with a Mendelian inheritance pattern to complex epilepsy with presumed polygenic or complex inheritance that contributes to disease susceptibility. Clinically, nonstructural epilepsy syndromes encompass a wide phenotypic spectrum of typically generalized or ­multifocal semiology, but can also include focal seizure types, often with a phenotypic overlap between different associated genes (Fig. 4.1). An important broad clinical spectrum of nonstructural epilepsies includes the early onset epileptic encephalopathies, with manifestations varying from (clusters of) focal or migrating focal seizures to epileptic spasms and tonic or tonic– clonic seizures, typically in combination with a severe and early onset developmental delay. In many of these epileptic encephalopathies, the epileptic activity itself is thought to contribute to the severity of the cognitive problems. These epileptic encephalopathies can be caused by pathogenic variants in ion channel genes such as SCN1A, KCNQ2, and KCNT1, as well as in non-ion channel coding genes such as, ARX, CDKL5, SLC25A22, and SPTAN1.32,​33,​34 On the other hand, pathogenic variants in genes have been identified in patients with epilepsy and encephalopathy in whom the severe encephalopathy continued after seizure control. Examples of these genes are STXBP1, MECP2, and GRIN2A.35 The idiopathic, or today called genetic, generalized epilepsies (IGEs or GGEs) comprise a large group of epilepsies and consist of several syndromes including childhood absence epilepsy, juvenile absence epilepsy, juvenile myoclonic epilepsy, and epilepsy with generalized tonic–clonic seizures alone. GGEs are mostly thought to be genetic epilepsies with a

4  Genetics in Epilepsy Surgery

Generalized/multifocal

Focal

Cerebral cavernoma

• KRIT1; CCM2; PDCD10

LIS

• PIK3CA • AKT3 • mTOR

• LIS1; DCX; TUBA1A

PMG

• GPR56; PIK3R2; TUBB2B HH PNH

TLE GEFS+ DS

• PIK3CA • AKT3 • mTOR

• SCN1A • PCDH19

• GRIN1 *b • GRIN2B

• GRIN2A

epilepsy-aphasia spectrum

• KCNT1 • KCNQ2 • KCNQ3

GEFS+ DS • SCN1B; GABRG2 • SCN2A; SCN9A; GABRD

GGE (JME, JAE) GGE

• CLCN2 • EFHC1 • GABRA1 • GABRB3 • CACNA1A

• PIK3R2 • PTEN • RNF135

Structural

HMEG

• ARX various *c

EFMR

SCN8A STXBP1 SYNGAP1 SPTAN1 CDKL5 MECP2 WDR45 ALG1 CHD2 IQSEC2 ARHGEF9 PNPO HDAC4

SWS

• GNAQ

• FLNA

• NF1 *a

LEAT

• BRAF

• Shh genes

• TSC1; TSC2

• • • • • • • • • • • • •

FCD

• • • • • • • • • • • • •

PIGA SLC25A22 MEF2C FOXG1 GNAO1 GOSR2 PRICKLE1 NEDD4L DNM1 HCN1 CACNA2D2 SLC2A1 TBC1D24

• NPRL2 • NPRL3 • DEPDC5

*d

FCD ADSHE TLE FFEVF

ADSHE EIMFS EIEE various

GEFS+

• • • •

CHRNA4 CHRNA2 CHRNB2 PRIMA1

ADSHE

Non-structural • LGI1 • RELN • CNTNAP2 EIEE

• PRRT2

• SLC12A5

ADPEAF

BFIE

EIMFS

Fig. 4.1  Overview of main monogenic causes of epilepsy classified into generalized/multifocal versus focal phenotypes, and structural versus nonstructural disorders. ADPEAF, autosomal dominant partial epilepsy with auditory features; ADSHE, autosomal dominant sleep-related hypermotor epilepsy; BFIE, benign familial infantile epilepsy; DS, Dravet syndrome; EFMR, epilepsy and mental retardation limited to females; EIEE, early infantile epileptic encephalopathy; EIMFS, epilepsy of infancy with migrating focal seizures; FCD, focal cortical dysplasia; FFEF, familial focal epilepsy with variable foci; FS, febrile seizures; GEFS+, generalized epilepsy with febrile seizures+; GGE, genetic generalized epilepsy; HH, hypothalamus hamartoma; HMEG, hemimegalencephaly; JAE, juvenile absence epilepsy; JME, juvenile myoclonic epilepsy; LEAT, low-grade epilepsy-associated tumor; NF, neurofibromatosis; PMG, polymicrogyria; PNH, periventricular nodular heterotopia; SWS, Sturge–Weber syndrome; TSC, tuberous sclerosis complex; TLE, temporal lobe epilepsy. *a Secondary to hippocampal sclerosis/dual pathology. *b Associated MCD has been (sporadically) described. *c Multiple types of focal as well as generalized seizures. *d Unclear if/how many non-operated MRI-negative patients have FCD.

31

32

I  Introduction to Epilepsy in Children complex, polygenic inheritance. However, in family linkage studies, pathogenic variants in single genes associated with this phenotypic spectrum have been reported. Pathogenic variants in GABRA1, CLCN2, and EFHC1 have been described in families with juvenile myoclonic epilepsy, but also in heterogeneous GGE phenotypes.4,​36 GABRG2 and CACNA1H pathogenic mutations were identified in families with childhood absence epilepsy.37 Other studies provided some evidence that GABRD, ME2, BRD2, and NEDD4L are susceptibility genes for GGE.38–41 Recent genome-wide association studies have shown that GGE is indeed a multifactorial genetic disease, in which multiple susceptibility loci across the genome confer low risk for disease.42

Structural Epilepsies Similar advances in epilepsy genetics—especially over the past two decades—have been made in identifying the genetic and molecular basis of structural epilepsies, which are clinically related to a focal or multifocal phenotype. Focal cortical dysplasias (FCDs), in particular FCD type II, are well-known underlying cortical malformations of severe pharmacoresistant epilepsies.43 Important clinical syndromes predominantly, but not exclusively related to FCDs are ADSHE, familial focal epilepsy with variable foci and ADPEAF. Although FCD is usually sporadic, family studies have identified germline or somatic single gene mutations in the mTOR pathway in many cases of FCD, in particular type II.16 This pathway is involved in various developmental central nervous system processes such as neuronal growth, migration and proliferation. Upregulation of the mTOR pathway is also implicated in the pathogenesis of other neurological conditions associated with cortical malformations and intractable seizures, including TSC and hemimegalencephaly. The mTOR complex 1 (mTORC1) represents one

of the most important regulators of cell growth within this pathway. GATOR1 (DEPDC5, NPRL2, NPRL3) as well as TSC1 and TSC2 are all negative regulators of the mTORC1 complex. O ­ ther key molecules, more upstream in this signaling pathway, are PI3K and AKT, and pathogenic variants in several genes related to these regulators (e.g., PIK3CA, PIK3R2, and AKT3) have been identified in focal epilepsy patients with various types of brain malformations (Fig. 4.2).44 The main hypothesis explaining the genetic basis of a focal structural etiology in general and the observed variation of lesions and phenotypical differences within families with the same pathogenic variant in particular is the so-called second hit theory. This theory emphasizes that lesions could result from two different mutational events: one causing a germline variant in a cortical development gene carried within the family that is by itself not sufficient to cause disease, and the other causing a somatic mutation of the other allele of that gene expressed within the FCD (or other MCD). This hypothesis may explain why family members may remain asymptomatic, if a second hit does not occur. Furthermore, differences in lesion characteristics could be explained by the assumption that, between family members with the same pathogenic variant, the second mutation can affect neurodevelopmental precursors at different time points.45 Other important MCDs with an acknowledged genetic basis include diffuse neuronal migration disorders such as lissencephaly, periventricular heterotopia, and—­ mainly bilateral—polymicrogyria.46 So far, more than 100 genes ­presumably associated with one or more types of MCD have been identified.46 Aside from MCDs, other genetic s­tructural disorders, increasingly recognized as neurodevelopmental ­ diseases with focal or multifocal epilepsy as important ­clinical feature, are neurofibromatosis type I, cerebral cavernous ­malformations, and Sturge–Weber syndrome.47

Fig. 4.2  Overview of important structural epilepsies associated with a genetic cause. FCD, focal cortical dysplasia; PMG, polymicrogyria; PNH, periventricular nodular heterotopia.

4  Genetics in Epilepsy Surgery

Challenges of Genetics in Epilepsy Surgery All efforts to identify epilepsy-causing genes and elucidating its neurobiological framework are driven by the goal of improving epilepsy care. The identification of a causative pathogenic variant could guide treatment choices, as well as prognostic and genetic counseling. Examples of so-called precision ­medicine for genetic epilepsies are the avoidance of sodium channel blockers in Dravet syndrome (SCN1A), and ketogenic diet as the major treatment for GLUT-1 deficiency (SLC2A1).48 From the perspective of epilepsy surgery, the identification of epilepsy genes is being increasingly translated into precision medicine in specific cases. In general, epilepsy surgery is presumed to be the optimal treatment for refractory epilepsy patients with a focal, lesional epilepsy, implicating a well-defined epileptogenic zone that can be localized and delineated in order to enable complete resection or disconnection, provided it is located outside eloquent cortex. In contrast, patients with a primary genetic cause of epilepsy, implicating a more diffuse pathophysiological mechanism that cannot be translated in to a well-defined removable epileptogenic zone, are presumed to be less eligible for resective surgery as well as for invasive diagnostics. Although there are several prognostic factors for postsurgical seizure outcome, selection of suitable candidates is still not optimal and exact indications and contraindications are not completely known. In many patients, especially those without a visible structural cerebral abnormality (MRI-negative patients), extensive preoperative diagnostic workup is required, including

invasive intracranial electroencephalography monitoring in an increasing number of patients. Currently, on average only 65% of patients achieve seizure freedom after surgery, and operated MRI-negative patients have an even lower chance of reaching seizure freedom.49 Furthermore, the absence of a specific histopathological abnormality (~8% of all operated patients) is a major predictor of poor outcome.50 The crucial differentiation between patients with operable and nonoperable epilepsy (i.e., between a presumed lesional and nonlesional cause of seizures) requires new and reliable biomarkers. Recent literature showed that some of the pathogenic variants underlying epilepsy may be related to the effectiveness of epilepsy surgery in focal epilepsy patients. A systematic review, which included a total of 82 patients, showed that surgery was successful in terms of complete control of seizures in only 14% of patients with pathogenic variants in genes related to channel function and synaptic transmission (e.g., SCN1A, SCN1B, and CNTNAP2), despite focal semiology of at least some of the patient’s seizures (Table 4.2 and Table 4.3).51 In contrast, epilepsy surgery completely controlled seizures in 58% of patients with epilepsy due to germline mutations in the mTOR pathway (e.g., DEPDC5, NPRL2, and NPRL3). The success rate in this subgroup was 71% for germline and somatic mutations combined.51 The disappointing overall seizure outcome of surgery in patients with pathogenic variants in genes related to channel function and synaptic transmission suggests a relative con-

Table 4.2  Success rates of epilepsy surgery for patients with different genetic causes (germline mutations) of epilepsy

MRI-lesional seizure-free/total

MRI-nonlesional seizure-free/total

Total group seizure-free/total

SCN1A

FCD: 0/2 HS: 0/2 Encephalomalacia: 0/1 Subcortical area of abnormal signal: 0/1

0/2

0/8

SCN1B

HS: 1/1

1/1

2/2

CNTNAP2

HS: 0/2

0/1

0/3

STXBP1

—

0/1

0/1

Overall

1/9

1/5

2/14 (14%)

DEPDC5

FCD: 3/6

2/3

5/9

PTEN

Hemimegalencephaly: 1/1

—

1/1

NPRL2

—

0/1

0/1

NPRL3

FCD: 1/1

—

1/1

Overall

5/8

2/4

7/12 (58%)

Microdeletions

HS: 9/10

0/2

9/12

Neurofibromatosis type 1

FCD: 2/2 HS: 4/6 Polymicrogyria 0/1 Tumour 5/11

1/1

12/21

Fragile-X syndrome

HS: 2/2

—

2/2

Mitochondrial mutations

HS: 1/3

—

1/3

Overall

23/35

1/3

24/38 (63%)

29/52 (56%)

4/12 (33%)

33/64 (52%)

Genetic cause Gene mutations involved with channelopathies and disorders of synaptic transmission

mTOR pathway gene mutations

Other genetic causes of epilepsy

Total

33

34

I  Introduction to Epilepsy in Children Table 4.3  Success rates of epilepsy surgery for patients with different genetic causes (somatic mutations) of epilepsy

MRI-lesional seizure-free/total

MRI-nonlesional seizure-free/total

Total group seizure-free/total

PIK3CA

Hemimegalencephaly: 5/5 FCD: 1/1

—

6/6

AKT3

Hemimegalencephaly: 1/3 FCD: 1/1

—

2/4

mTOR

Hemimegalencephaly: 1/1 FCD: 6/7

—

7/8

15/18 (83%)

—

15/18 (83%)

Genetic cause mTOR pathway gene mutations

Total

Abbreviations: FCD, focal-cortical dysplasia; HS, hippocampal sclerosis. Source: Reproduced with permission from Stevelink et al 2018.51

traindication for epilepsy surgery (and invasive diagnostics), ­particularly in MRI-negative patients. An exception could be if epilepsy surgery is used as palliative treatment to target s­ pecific intractable focal seizures related to, for example, ­hippocampal sclerosis in SCN1A-related epilepsy, without the primary aim of achieving complete seizure freedom for all seizure types. On the other hand, the finding of a germline or mosaic mTOR gene pathogenic variant could justify continuation of the ­presurgical diagnostic process, even if a patient is MRI-negative. Such a finding could increase the chance of an underlying MCD being responsible for the patients’ focal epilepsy. Although the low number of surgical cases for most genetic causes hampers firm conclusions, these findings encourage the evaluation of pathogenic variants in novel and known epilepsy genes as potential biomarkers to improve presurgical decision-making

and ultimately to enhance postsurgical seizure outcome in patients with refractory epilepsy. To date, genetic diagnostics are not routinely performed w ­ ithin the presurgical evaluation trajectory. In patients who undergo genetic testing, the genetic diagnosis is often e ­ stablished only after the patient was already accepted or rejected for ­surgery. This could predispose patients to either undergo u ­ nnecessary invasive diagnostics or even resective surgery (e.g., if the pathogenic variant points toward a genetic syndrome without an identifiable lesion), or to be unjustly rejected for surgery (e.g., if the pathogenic variant points toward a structural epilepsy that is caused by an “MRI-invisible” microstructural lesion). Larger and prospective studies are needed to further elucidate the importance of genetic testing in patients who are considered possible candidates for epilepsy surgery.

References 1. Steinlein OK, Mulley JC, Propping P, et al. A missense mutation in the neuronal nicotinic acetylcholine receptor alpha 4 subunit is associated with autosomal dominant nocturnal frontal lobe epilepsy. Nat Genet 1995;11(2):201–203 2. Baulac S, Huberfeld G, Gourfinkel-An I, et al. First genetic evidence of GABA(A) receptor dysfunction in epilepsy: a mutation in the gamma2-subunit gene. Nat Genet 2001;28(1):46–48 3. Charlier C, Singh NA, Ryan SG, et al. A pore mutation in a novel KQT-like potassium channel gene in an idiopathic epilepsy family. Nat Genet 1998;18(1):53–55 4. Cossette P, Liu L, Brisebois K, et al. Mutation of GABRA1 in an autosomal dominant form of juvenile myoclonic epilepsy. Nat Genet 2002;31(2):184–189 5. Escayg A, MacDonald BT, Meisler MH, et al. Mutations of SCN1A, encoding a neuronal sodium channel, in two families with GEFS+2. Nat Genet 2000;24(4):343–345 6. Seidner G, Alvarez MG, Yeh JI, et al. GLUT-1 deficiency syndrome caused by haploinsufficiency of the blood-brain barrier hexose carrier. Nat Genet 1998;18(2):188–191 7. Singh NA, Charlier C, Stauffer D, et al. A novel potassium channel gene, KCNQ2, is mutated in an inherited epilepsy of newborns. Nat Genet 1998;18(1):25–29 8. Wallace RH, Wang DW, Singh R, et al. Febrile seizures and generalized epilepsy associated with a mutation in the Na+-channel beta1 subunit gene SCN1B. Nat Genet 1998;19(4):366–370 9. Mefford HC, Muhle H, Ostertag P, et al. Genome-wide copy number variation in epilepsy: novel susceptibility loci in idiopathic generalized and focal epilepsies. PLoS Genet 2010;6(5):e1000962

10. Olson H, Shen Y, Avallone J, et al. Copy number variation plays an important role in clinical epilepsy. Ann Neurol 2014;75(6):943–958 11. Vlaskamp DRM, Callenbach PMC, Rump P, et al. Copy number variation in a hospital-based cohort of children with epilepsy. Epilepsia Open 2017;2(2):244–254 12. de Kovel CG, Trucks H, Helbig I, et al. Recurrent microdeletions at 15q11.2 and 16p13.11 predispose to idiopathic generalized epilepsies. Brain 2010;133(Pt 1):23–32 13. Mefford HC. CNVs 2014;2:162–167

in

Epilepsy.

Curr

Genet

Med

Rep

14. Mullen SA, Carvill GL, Bellows S, et al. Copy number variants are frequent in genetic generalized epilepsy with intellectual disability. Neurology 2013;81(17):1507–1514 15. Helbig I, Tayoun AA. Understanding genotypes and ­phenotypes in epileptic encephalopathies. Mol Syndromol 2016;7(4): 172–181 16. Møller RS, Weckhuysen S, Chipaux M, et al. Germline and somatic mutations in the MTOR gene in focal cortical dysplasia and epilepsy. Neurol Genet 2016;2(6):e118 17. Wang J, Lin ZJ, Liu L, et al. Epilepsy-associated genes. Seizure 2017;44:11–20 18. Weckhuysen S, Marsan E, Lambrecq V, et al. Involvement of GATOR complex genes in familial focal epilepsies and focal cortical dysplasia. Epilepsia 2016;57(6):994–1003 19. Guerrini R, Dobyns WB. Malformations of cortical development: clinical features and genetic causes. Lancet Neurol 2014;13(7):710–726

4  Genetics in Epilepsy Surgery 20. Kalachikov S, Evgrafov O, Ross B, et al. Mutations in LGI1 cause autosomal-dominant partial epilepsy with auditory features. Nat Genet 2002;30(3):335–341

35. Stamberger H, Nikanorova M, Willemsen MH, et al. STXBP1 encephalopathy: a neurodevelopmental disorder including epilepsy. Neurology 2016;86(10):954–962

21. McTague A, Howell KB, Cross JH, Kurian MA, Scheffer IE. The genetic landscape of the epileptic encephalopathies of infancy and childhood. Lancet Neurol 2016;15(3):304–316

36. D’Agostino D, Bertelli M, Gallo S, et al. Mutations and polymorphisms of the CLCN2 gene in idiopathic epilepsy. Neurology 2004;63(8):1500–1502

22. Dibbens LM, de Vries B, Donatello S, et al. Mutations in DEPDC5 cause familial focal epilepsy with variable foci. Nat Genet 2013;45(5):546–551

37. Chen Y, Lu J, Pan H, et al. Association between genetic variation of CACNA1H and childhood absence epilepsy. Ann Neurol 2003;54(2):239–243

23. Ricos MG, Hodgson BL, Pippucci T, et al; Epilepsy Electroclinical Study Group. Mutations in the mammalian target of rapamycin pathway regulators NPRL2 and NPRL3 cause focal epilepsy. Ann Neurol 2016;79(1):120–131

38. Dibbens LM, Ekberg J, Taylor I, et al. NEDD4–2 as a potential candidate susceptibility gene for epileptic photosensitivity. Genes Brain Behav 2007;6(8):750–755

24. Harkin LA, McMahon JM, Iona X, et al; Infantile Epileptic Encephalopathy Referral Consortium. The spectrum of SCN1A-­ related infantile epileptic encephalopathies. Brain 2007;130 (Pt 3):843–852 25. Claes LR, Deprez L, Suls A, et al. The SCN1A variant database: a novel research and diagnostic tool. Hum Mutat 2009;30(10):E904–E920 26. Zuberi SM, Brunklaus A, Birch R, Reavey E, Duncan J, Forbes GH. Genotype-phenotype associations in SCN1A-related epilepsies. Neurology 2011;76(7):594–600 27. Singh NA, Westenskow P, Charlier C, et al; BFNC Physician Consortium. KCNQ2 and KCNQ3 potassium channel genes in benign familial neonatal convulsions: expansion of the functional and mutation spectrum. Brain 2003;126(Pt 12): 2726–2737 28. de Lange IM, Koudijs MJ, van’t Slot R, et al. Mosaicism of de novo pathogenic SCN1A variants in epilepsy is a frequent phenomenon that correlates with variable phenotypes. Epilepsia 2018;59(3):690–703 29. Myers KA, Scheffer IE. GRIN2A-related speech disorders and ­epilepsy. In: Adam MP, Ardinger HH, Pagon RA, Wallace SE, Bean LJH, Stephens K, et al. editors. Gene Reviews (R). Seattle, WA: ­University of Washington; 1993. 30. Helbig I, Scheffer IE, Mulley JC, Berkovic SF. Navigating the channels and beyond: unravelling the genetics of the epilepsies. Lancet Neurol 2008;7(3):231–245 31. Scheffer IE, Berkovic S, Capovilla G, et al. ILAE classification of the epilepsies: position paper of the ILAE Commission for classification and terminology. Epilepsia 2017;58(4): 512–521 32. Kato M, Saitoh S, Kamei A, et al. A longer polyalanine expansion mutation in the ARX gene causes early infantile epileptic encephalopathy with suppression-burst pattern (Ohtahara syndrome). Am J Hum Genet 2007;81(2):361–366 33. Elia M, Falco M, Ferri R, et al. CDKL5 mutations in boys with severe encephalopathy and early-onset intractable epilepsy. Neurology 2008;71(13):997–999 34. Milh M, Villeneuve N, Chouchane M, et al. Epileptic and nonepileptic features in patients with early onset epileptic encephalopathy and STXBP1 mutations. Epilepsia 2011;52 (10):1828–1834

39. Dibbens LM, Feng HJ, Richards MC, et al. GABRD encoding a protein for extra- or peri-synaptic GABAA receptors is a susceptibility locus for generalized epilepsies. Hum Mol Genet 2004; 13(13):1315–1319 40. Greenberg DA, Cayanis E, Strug L, et al. Malic enzyme 2 may underlie susceptibility to adolescent-onset idiopathic generalized epilepsy. Am J Hum Genet 2005;76(1):139–146 41. Pal DK, Evgrafov OV, Tabares P, Zhang F, Durner M, Greenberg DA. BRD2 (RING3) is a probable major susceptibility gene for common juvenile myoclonic epilepsy. Am J Hum Genet 2003;73(2):261–270 42. Steffens M, Leu C, Ruppert AK, et al; EPICURE Consortium. EMINet Consortium. Genome-wide association analysis of genetic generalized epilepsies implicates susceptibility loci at 1q43, 2p16.1, 2q22.3 and 17q21.32. Hum Mol Genet 2012;21(24):5359–5372 43. Baldassari S, Licchetta L, Tinuper P, Bisulli F, Pippucci T. GATOR1 complex: the common genetic actor in focal epilepsies. J Med Genet 2016;53(8):503–510 44. Baulac S, Ishida S, Marsan E, et al. Familial focal epilepsy with focal cortical dysplasia due to DEPDC5 mutations. Ann Neurol 2015;77(4):675–683 45. Leventer RJ, Jansen FE, Mandelstam SA, et al. Is focal cortical ­dysplasia sporadic? Family evidence for genetic susceptibility. Epilepsia 2014;55(3):e22–e26 46. Parrini E, Conti V, Dobyns WB, Guerrini R. Genetic basis of brain malformations. Mol Syndromol 2016;7(4):220–233 47. Shirley MD, Tang H, Gallione CJ, et al. Sturge–Weber syndrome and port-wine stains caused by somatic mutation in GNAQ. N Engl J Med 2013;368(21):1971–1979 48. Weber YG, Biskup S, Helbig KL, Von Spiczak S, Lerche H. The role of genetic testing in epilepsy diagnosis and management. Expert Rev Mol Diagn 2017;17(8):739–750 49. Téllez-Zenteno JF, Hernández Ronquillo L, Moien-Afshari F, Wiebe S. Surgical outcomes in lesional and non-lesional ­epilepsy: a systematic review and meta-analysis. Epilepsy Res 2010;89(2–3):310–318 50. Wang ZI, Alexopoulos AV, Jones SE, Jaisani Z, Najm IM, Prayson RA. The pathology of magnetic-resonance-imaging-negative epilepsy. Mod Pathol 2013;26(8):1051–1058 51. Stevelink R, Sanders MW, Tuinman MP, et al. Epilepsy surgery for patients with genetic refractory epilepsy: a systematic review. Epileptic Disord 2018;20(2):99–115

35

5

  Surgical Neuropathology of Pediatric Epilepsy Thomas W. Smith

Summary This chapter presents a general survey of the pathological conditions that may be encountered in children who undergo lesional surgery for epilepsy. These lesions are grouped into six major categories: malformations of cortical development, neoplasms, hamartomas, hippocampal sclerosis, vascular malformations, and other acquired lesions, including vascular, traumatic, infectious, and autoimmune conditions. The major emphasis will be on the description of the macroscopic and histologic appearance of these lesions, including immunohistochemical and molecular/genetic findings where appropriate. Keywords:  malformations of cortical development, glioneuronal tumor, hamartoma, hippocampal sclerosis, vascular malformation, autoimmune encephalitis

„„ Introduction An important goal of surgery for childhood epilepsy is to identify the nature of the underlying pathological process accounting for the seizure disorder. Although imaging technology, in a particular MRI, has significantly improved identification and to some degree the general classification of these lesions, histological examination remains necessary to establish the exact nature of the pathological process and to provide ancillary information concerning adequacy of surgical resection, biologic behavior, and prognosis.

„„ Malformations of Cortical Development Malformations of Cortical Development Due to Abnormal Neuronal and Glial Proliferation or Differentiation These lesions (together with glial and glioneuronal tumors, GNTs) are among the most frequent causes of surgically treatable epilepsy in the pediatric age group. They include the focal cortical dysplasias (FCDs) and related conditions—cortical tubers (tuberous sclerosis) and hemimegalencephaly.

Focal Cortical Dysplasia In this section, we will use the current three-tiered International League Against Epilepsy (ILAE) classification of FCD.1 The milder forms of FCD include FCD types Ia-c (abnormalities of radial and/ or tangential cortical lamination occurring as isolated lesions) and FCD types IIIa-d (cortical lamination abnormalities associated with other epileptogenic lesions).1,​2 These lesions are difficult to detect radiologically and their exact relationship to the cause of the seizure disorder is uncertain. Moreover, the histological identification and classification of a cortical lamination abnormality can be challenging due to misorientation of specimens and difficulty in interpreting very subtle changes or variations of normal structure. Other histological findings often grouped within this category include aberrant clustering of neurons and/or ­oligodendroglia, isolated “ectopic” neurons in white matter, oligodendroglial hyperplasia, and some glial-neuronal hamartomas. FCD type II is the most frequent and reliably identifiable FCD lesion both histologically and on neuroimaging.1–3 Macroscopically, the affected cortex is usually thickened with an indistinct boundary between cortex and white matter. On microscopic examination the normal pattern of cortical lamination is disrupted, which can be readily appreciated on hematoxylin-eosin or cresyl violet-stained sections. Gliosis and abnormal myelination may be present. The histological appearance of FCD type II is defined by two specific cell types: dysmorphic neurons and balloon cells.2,​3 If the lesion contains only dysmorphic neurons but not balloon cells, then it is classified as FCD type IIa. If both dysmorphic neurons and balloon cells are present, then the lesion is classified as FCD type IIb. Dysmorphic neurons (Fig. 5.1a) are characterized by their markedly increased size (neuronal cytomegaly), abnormal orientation or “polarity” of cellular processes, tortuous dendrites, and abnormal cytoplasmic filaments that are immunoreactive for phosphorylated neurofilament protein and other neuronal proteins. These cells should not be confused with normal Betz neurons or misinterpreted as neoplastic cells, as may occur in GNTs such as ganglioglioma. Balloon cells (Fig. 5.1b) are large round to oval cells with eccentric glassy pink cytoplasm and eccentric nuclei, and may include multinucleate or giant forms. They superficially resemble large reactive astrocytes, but are inconsistently immunoreactive for glial fibrillary acidic protein (GFAP). Some may express only neuronal markers or both glial and neuronal markers, which reflects their abnormal maturation or differentiation. Both dysmorphic neurons and balloon cells can be found in all cortical layers and in the subcortical white matter.

5  Surgical Neuropathology of Pediatric Epilepsy

Fig. 5.1  Focal cortical dysplasia (FCD) type II. (a) Dysmorphic neurons. Markedly enlarged neurons with abnormal cytoarchitecture including fibrillary cytoplasm. Dysmorphic neurons are present in FCD types IIa and IIb. (b) Balloon cells. Abnormal large cells with eccentric nuclei and glassy pink cytoplasm; these cells often share morphologic and immunohistochemical features in common with both neurons and astrocytes. Balloon cells are present only in FCD type IIb.

Tuberous Sclerosis Complex Surgical resection of an epileptogenic cortical tuber may be performed in patients with tuberous sclerosis complex (TSC) and intractable seizures. Since cortical tubers represent a form of cortical dysplasia, they share many histological features in common with FCD II including disordered cortical lamination and the presence of dysmorphic neurons and balloon cells.2,​3 Dysmorphic neurons in a cortical tuber may not be as frequent or show as extreme cytomegaly as in FCD type II. Diffuse gliosis, atypical-appearing astrocytes and calcifications may also be more frequent in tubers. In many cases, however, a histological distinction between FCD and TSC may not be possible and requires correlation with other clinical and radiographic features of TSC or testing for TSC1/TSC2 mutations.

Hemimegalencephaly Hemimegalencephaly (HME) is an extreme form of cortical dysplasia that involves large portions of one hemisphere usually resulting in asymmetrical enlargement of the brain. HME usually occurs as a sporadic lesion, however a systemic variant of this condition has been associated with partial or total hemigigantism and several neurocutaneous syndromes.3 The extent of cortical involvement in HME can be variable and may not always affect all regions of a hemisphere. The abnormal cortex is usually excessively thick and shows poor demarcation between gray and white matter. Abnormally broad gyri or polymicrogyria-type cortex may be present in some cases. The histological features of HME resemble those of severe FCD type II and include disorganized cortical lamination, enlarged dysmorphic neurons and balloon cells. Since the histology of the dysplastic cortex in HME can show considerable overlap with both FCD and cortical tubers, it may be difficult to distinguish between these conditions in surgical specimens in the absence of supportive clinical and radiologic information.

Pathogenesis of Cortical Dysplastic Lesions The remarkable similarity of the cortical histology and immunophenotype in FCD type II, TSC, and HME is consistent

with the presence of a common abnormality affecting the periventricular germinal epithelium that occurs in early life (between gestational weeks 8–20) and leads to the aberrant migration, maturation, and differentiation of neuroglial progenitor cells. All three conditions have been shown to have either germline or somatic mutations directly or indirectly involving the phosphatidylinositol 3-kinase/Akt/mammalian target of rapamycin (PI3K/Akt/mTOR) pathway, which plays a critical role in neuronal growth and migration.2–6 TSC is characterized by germline mutations in TSC1 (encoding hamartin) or TSC2 (encoding tuberin). Both hamartin and tuberin are regulatory proteins that negatively modulate mTOR. In contrast, FCD II and HME have postzygotic somatic mosaic mutations in the mTOR pathway, which include MTOR, PI3KCA, AKT3, and DEPDC5. The dysmorphic neurons and balloon cells in all three conditions have been shown to express activated (phosphorylated) components of the mTOR pathway, such as pS6.2 The timing of the expression of these mutations during the cell cycle may account for the differences in phenotypic expression: late cycle mutations would be more likely cause smaller lesions characteristic of FCD II or TSC, whereas early cycle mutations cause the more extensive lesions characteristic of HME.4 The underlying cause of the somatic mTOR pathway mutations in FCD II and HME is unknown. A viral etiology (human papillomavirus) has been proposed but not confirmed.4 In addition to the molecular changes involving the mTOR pathway, the dysmorphic neurons, and balloon cells in both FCD and TSC have been found to aberrantly express ­doublecortin-like (DCL), a microtubule-associated protein which is critically involved in neuronal division and radial migration. This suggests a probable role for doublecortin (DCX)-related genes in the pathogenesis of these conditions, although specific mutations have not yet been identified.7 The pathogenesis of FCD types I and III is not well understood, but is not believed to involve the mTOR signaling pathway.

Malformations of Cortical Development Due to Abnormal Neuronal Migration Lissencephaly Lissencephaly is commonly grouped into two categories: classical lissencephaly (LIS type I) and cobblestone lissencephaly (LIS type II).2 Classical lissencephaly (LIS1) is an abnormality of cortical development characterized by absent or deficient gyration.2,​8 The spectrum of LIS1 may include absent or near absent gyration (agyria) or broad, simplified gyration (pachygyria). LIS1 usually affects both hemispheres, although hemi-lissencephaly may also occur. Although LIS1 is usually associated with severe intractable epilepsy, surgical removal of a specific seizure focus may be indicated in some cases. The correct identification of the cortical lesion as LIS1 is usually provided by neuroimaging and can be helpful for the pathologist who may receive only a small portion of the lesion for examination. The microscopic appearance of LIS1 is readily identifiable as a thick, disorganized cortex having abnormal horizontal lamination with four rather than the usual six layers. Beginning from the cortical surface, the four layers can be further described as: (1) a poorly defined molecular with increased cellularity; (2) a superficial highly cellular zone with diffusely scattered large pyramidal neurons; (3) a relatively neuron-sparse zone with numerous myelinated fibers;

37

38

I  Introduction to Epilepsy in Children and (4) a thicker deep cortical zone with disorganized neurons often arranged in a columnar pattern. Classical lissencephaly is a genetic disorder, most often related to mutations in doublecortin (DCX) on the X chromosome or LIS1 on chromosome 17.2,​8–10 DCX mutations cause classical lissencephaly in hemizygous males and subcortical band heterotopia (SBH, see the next section) in heterozygous females. Both the DCX and LIS1 genes are involved in normal neuronal migration and maturation. Other gene mutations may be associated with LIS1, although the corresponding histological phenotypes are not as well characterized. Cobblestone lissencephaly (LIS2) differs histologically from classical lissencephaly, and is characterized by: (1) absence of a recognizable pattern of lamination, and (2) over-migration of neurons and glia into the arachnoid space resulting in the formation of an extracortical neuroglial layer which accounts for the agyric “cobblestone” appearance of the brain surface.2,​8 The cortical abnormality in LIS2 has been attributed to a defect of the outermost pial–glial layer of the brain, resulting in abnormal settlement of the cortical plate. Cobblestone lissencephaly is pathognomic of a continuum of autosomal recessive diseases associated with cerebral, ocular and muscular deficits attributed to abnormal glycosylation of alpha-dystroglycan, associated with mutations in at least six genes (POMT1, POMT2, POMGNT1, LARGE, FKTN, and FKRP).11

Subcortical Band Heterotopia SBH is characterized by discrete bands of heterotopic gray matter (“double cortex”) located in the white matter between the lateral ventricles and normal-appearing cerebral cortex.2,​8 The bands are typically bilateral and symmetrical, and are slightly more prominent anteriorly. The brain usually shows no other malformations. Microscopically the heterotopic bands consist of a superficial zone of well-differentiated but disorganized neurons, an intermediate zone of small neurons arranged in columns, and a deeper zone where the heterotopia may have a nodular configuration. The clinical symptoms of SBH include intellectual disability and epilepsy, the severity of which roughly correlates with heterotopia band thickness. SBH is thought to represent a defect of neuronal migration during development resulting in the formation of the heterotopic bands. SBH is most often caused by germline mutations in the DCX gene and is inherited in an X-linked dominant pattern. Females will have SBH, whereas males will have classical lissencephaly. Some cases may be caused by a somatic mutation involving the LIS1 gene on chromosome 17.8–10

Periventricular Nodular Heterotopia Periventricular nodular heterotopia (PNH) are the most common form of brain heterotopia and consist of nodular masses of gray matter that lie adjacent to or protrude into the lateral ventricles.2 They may be single or multiple, and separated or contiguous. The trigones and occipital horns of the lateral ventricles are most commonly involved. Microscopically, PNH consist of mature neurons that have no apparent organization or at most rudimentary lamination. The gray nodules may be separated by or include myelinated fibers. The major clinical manifestation of PNH is severe epilepsy, occurring in 80 to 90% of patients. In some patients, surgical treatment may be beneficial when an epileptogenic focus can be identified.

Although occasional PNH may occur sporadically, most patients with numerous PNH are associated with mutations in the FLNA gene which encodes filamin A, an actin-binding protein involved in cell structure, adhesion, and motility.2,​8–10 Thus, a defect in this protein could disrupt the normal pattern of neuronal migration during brain development and also lead to abnormal adhesion of neurons adjacent to ventricular walls. Some cases of PNH have also been associated with mutations in ARFGEF2 gene or abnormalities in chromosome 5, which are also involved in neuronal migration.8–9

Malformations of Cortical Development Due to Abnormal Cortical Organization Polymicrogyria Polymicrogyria (PMG) is a malformation characterized by excessive numbers of small irregular gyri separated by shallow sulci resulting in an irregular cortical surface.2,​8 The extent of PMG can be quite variable, ranging from involvement of a single gyrus to more diffuse involvement which may be unilateral or bilateral, symmetrical or asymmetrical. The perisylvian cortex is most often involved. PMG usually occurs as an isolated malformation, although a variety of other brain malformations may be seen in some cases. There are two major histological subtypes of PMG: yy Unilayered PMG characterized by a continuous molecular layer lacking a convolutional pattern. The cortical neurons have a radial distribution and lack a laminar o ­ rganization. This subtype is consistent with an early disruption of normal neuronal migration with subsequent cortical disorganization. yy Four-layered PMG consisting of a superficial acellular molecular layer with inward folding forming “­microsulci,” an outer neuronal layer, a nerve fiber layer, and an inner neuronal layer. Although its exact pathogenesis is unknown, this subtype has been attributed to vascular perfusion failure occurring between the 20th and 24th weeks of gestation resulting laminar necrosis and a late disorder of cortical migration and organization. Both histological subtypes may occur in the same cortical region. The etiology and pathogenesis of PMG is complex and multifactorial; both acquired and genetic causes have been implicated.2,​8,​12 Acquired etiologies have included hypoxia or hypoperfusion, and congenital infections such as toxoplasmosis, rubeola, and CMV. PMG may be caused by specific genetic mutations, although with few exceptions (e.g., GPR56 in bilateral frontoparietal PMG; PAX6 in unilateral PMG) most have not yet been identified. PMG may also occur as part of a multiple congenital anomaly syndrome associated with a chromosomal abnormality. The highly variable expression of PMG and its possible association with other brain malformations account for the wide clinical presentation of this disorder. The majority of patients with PMG will have a seizure disorder. Although PMG is often a widespread lesion, surgical resection may be appropriate in situations where an epileptogenic zone can be identified.

Schizencephaly Schizencephaly (SCZ) is an uncommon cortical malformation characterized by the presence of a cerebrospinal fluid-filled

5  Surgical Neuropathology of Pediatric Epilepsy cleft that extends across the entire cerebral hemisphere from the pial surface to the ependymal lining of the lateral ventricle.2,​8 The walls of the cleft are always lined by abnormal gray matter that has the appearance of PMG. The clefts may be unilateral or bilateral, and closed (SCZ type I) or open (SCZ type II). Approximately two-thirds of the clefts are frontal or parietal, and approximately one-third are temporal or occipital. The histological appearance of the abnormal cortex in the cleft walls is indistinguishable from typical PMG (see above). SCZ is considered to result from failure of development of the cerebral mantle in the zones of cleavage of the primary cerebral fissures; thus SCZ must be distinguished from clefts due to a prior destructive lesion (usually infarction) that are referred to as porencephaly. SCZ may occur in conjunction with other cerebral malformations, most often agenesis of the septum pellucidum and optic nerve hypoplasia. The etiology of SCZ is not well understood; both environmental (e.g., intrauterine CMV infection, toxins, and embryonic vascular insult) and genetic risk factors (e.g., mutations in EMX2, SIX3, SHH, and COL4A1) have been implicated.2,​8,​13 The clinical presentation of SCZ is variable depending on the severity of involvement, and may include developmental delay, language and motor deficits, and seizures, which are commonly present in this disorder. Epilepsy surgery may be beneficial in cases where a specific seizure focus can be identified (usually abnormal cortex near the cleft). Recognition of a surgically resected lesion as part of a SCZ malformation will usually depend on correlation with neuroimaging.

„„ Neoplastic Lesions Brain tumors are among the most common structural causes of long-term epilepsy in children. Many of these tumors can be grouped under the general category of “long-term epilepsy-­ associated tumors” (LEATs).14,​15 LEATs generally differ from other brain tumors by their young age of symptom onset, slow growth, and localization to the neocortex, often the temporal lobe. Many LEATs are composed of both neuronal and glial e ­ lements, hence the designation “glioneuronal.” They are histo­logically low-grade tumors, usually grade I in the WHO classification system. LEATs may also coexist with histological changes resembling FCD type IIIb in the cortex adjacent to the tumor.

Glioneuronal Tumors Dysembryoplastic Neuroepithelial Tumor (WHO Grade I) Dysembryoplastic neuroepithelial tumor (DNT) is a low-grade GNT that is almost always associated with a seizure disorder.14,​16 It is histologically complex tumor that may contain astrocytic, oligodendroglial and neurocytic elements, has a characteristic nodular (or multinodular) architecture, and has a predominantly intracortical location with a predilection for the temporal lobe. The histological hallmark of DNT is the “specific glioneuronal element,” which consists of columns of oligodendrocyte-like or neurocytic cells and axon bundles separated by microcystic spaces containing a myxoid material in which larger normal-­ appearing neurons appear to float (“floating neurons”) (Fig. 5.2).

Fig. 5.2  Dysembyroplastic neuroepithelial tumor. (a) Specific glioneuronal element consisting of columns of oligodendrocyte-like cells separated by loose microcystic areas. (b) “Floating neuron” (arrow) in microcystic space.

Abnormal neurons typical of ganglion cell tumors (GCTs) are usually not present in DNTs. Dysplastic cytoarchitecture is often present in the cortex adjacent to the tumor nodules. Degenerative changes such as calcification, cystic change and white matter rarefaction may be present. The vast majority of DNTs are slow growing and histologically indolent, corresponding to grade I in the WHO classification; they have scant if any mitotic activity and lack other histological features suggestive of anaplastic change. Surgical excision is considered curative; most DNTs do not recur, even after partial excision. True malignant transformation of a DNT is extremely rare. The histological diagnosis of DNT can be challenging, especially when the specific glioneuronal element is not apparent. In such cases an appropriate diagnosis of DNT can often be made after correlation with other clinical and radiographic features. DNTs should be distinguished from other GNTs (e.g., GCTs) and glial tumors such as oligodendroglioma, which may have a similar histological appearance. Testing for isocitrate dehydrogenase 1/2 (IDH1/2) mutations and 1p/19q codeletion may help differentiate DNT from other glial tumors in difficult cases.

Ganglion Cell Tumors (WHO Grade I or III) GCTs are slow-growing, usually low grade GNTs, and are frequently associated with a long-standing seizure disorder.14 GCTs most often occur in the temporal lobes but can arise in many ­other regions of the brain as well as in the spinal cord. GCTs are usually circumscribed, homogeneous masses but can also present as a larger cystic lesion with a mural nodule (similar to p ­ ilocytic astrocytoma). GCTs can be further subclassified depending on whether neurons constitute the only tumor component (gangliocytoma)17 or both neurons and neoplastic glial cells are present (ganglioglioma).18 In both GCT subtypes, the ganglion cell component consists of mature-appearing neurons that show abnormal clustering, polarity and/or cytoarchitecture including marked size variability and bi- or multinucleate forms (Fig. 5.3). The glial component in gangliogliomas may exhibit histological features similar to pilocytic astrocytoma (see below) including a dense fibrillary cytoarchitecture, granular eosinophilic bodies, and Rosenthal fibers. On occasion the glial c­ omponent

39

40

I  Introduction to Epilepsy in Children

Fig. 5.3  Ganglioglioma. Neurons in the tumor are increased in number and show abnormal clustering and size variation. The astrocytic component is characterized by a densely fibrillary matrix interspersed among the neoplastic neurons. Inset (lower right): high magnification showing abnormal cytoarchitecture of neurons including variation in size and shape.

can completely overshadow the presence of the neuronal component, which may require confirmation by immunohistochemical staining for neuronal markers such as synaptophysin. Tumor cells immunoreactive for the stem cell marker CD34 are frequently observed in GCTs and may be helpful in diagnostically challenging cases. Since the usual biologic behavior of most GCTs is that of a slow-growing, benign tumor, they are classified as WHO grade I tumors. In rare gangliogliomas, the glial component may show histological changes of anaplasia, including mitoses, elevated Ki-67 proliferative index, and in some cases microvascular proliferation and necrosis; these tumors are designated as WHO grade III.19 Twenty to 60% of all GCTs have BRAF V600E point mutations which are thought to be activating the RAS/MAPK signaling pathway in these tumors; these can be evaluated by direct molecular mutation analysis or by immunostaining for the BRAF V600E mutant protein.18 GCTs typically do not have mutations in IDH1/2, TP53, PTEN, CDK4, or EGFR, or show codeletion of 1p/19q on FISH; thus these tests may be useful in differentiating histologically challenging GCTs from other glial tumors.

Glial Tumors Oligodendroglioma (WHO Grade II or III) Oligodendrogliomas are diffusely infiltrating WHO grade II gliomas that are primarily located in the cerebral hemispheres.14,​20 They are often associated with a history of chronic seizures, which may reflect their relatively slow growth. Most oligodendrogliomas occur in adults, and are relatively rare in children under the age of 15. The classic histological appearance of oligodendroglioma consists of a homogeneous arrangement of tumor cells with round to oval nuclei surrounded by perinuclear halos (“fried egg” appearance) which is a consistent artifact seen on paraffin-embedded sections (Fig. 5.4a). The tumor is frequently associated with a delicate network of thin-walled b ­ ranching

Fig. 5.4  (a) Oligodendroglioma: highly cellular tumor composed of cells with round to oval nuclei and prominent perinuclear halos. (b) Diffuse astrocytoma: moderately cellular tumor composed of astrocytic cells with pleomorphic hyperchromatic nuclei, pink cytoplasm and fibrillary processes.

capillaries and microcalcifications. This classic histological appearance is highly correlated with the presence of an IDH1 or IDH2 mutation and codeletion of 1p/19q.20 A small subset of otherwise histologically typical oligodendrogliomas, mainly occurring in children and adolescents, may lack IDH mutations or 1p/19q codeletion on molecular testing.20 This subset has not yet been completely characterized molecularly; however, some have been associated with BRAF fusion genes (similar to pilocytic astrocytoma). Since DNTs and other tumors such as pilocytic astrocytoma and neurocytoma can have histological features ­ that overlap with oligodendroglioma, these entities should be excluded before making a diagnosis of oligodendroglioma in a childhood tumor that is negative for an IDH mutation and 1p/19 codeletion. Anaplastic oligodendrogliomas (WHO grade III) are characterized by increased cell density, mitotic activity, microvascular proliferation, and/or necrosis.21 They are very rare in children and are somewhat less likely to present as a seizure disorder compared to their grade II counterpart. Pediatric anaplastic oligodendrogliomas are also less likely to have IDH mutations or 1p/19q codeletion, although a distinct molecular profile has not yet been established.

Diffuse Astrocytoma (WHO Grades II–IV) This category includes all diffusely infiltrating glial tumors whose cell type can be reliably identified as astrocytic (Fig. 5.4b).14,​22 In the WHO classification, they are designated as grade II (­diffuse astrocytoma), grade III (anaplastic astrocytoma), or grade IV (glioblastoma); the histological grade is based on assessment of four parameters: cytological atypia, mitotic activity, ­microvascular proliferation, and necrosis.22 Diffuse astrocytomas are more likely to arise in adults compared to the pediatric age group, although examples of all grades may occur in children. Seizures may be a presenting or accompanying feature of diffuse astrocytomas occurring in the cerebrum, however compared to most GNTs, they are less likely to manifest as a long-standing seizure disorder. The classification of diffuse astrocytoma has been recently updated to include molecular data, in particular IDH mutation status.23 IDH mutations are frequently seen in WHO

5  Surgical Neuropathology of Pediatric Epilepsy grade II astrocytomas, as well as in many grade III tumors; however, most glioblastomas are IDH wildtype, with the exception of “secondary” glioblastomas that arise through progressive anaplasia of a lower grade (usually IDH mutant) astrocytoma.23 Pediatric diffuse astrocytomas histologically resemble adult diffuse astrocytomas, and are assigned a WHO grade using the same histological criteria. There are, however, some major differences: (1) although the majority occur in the cerebral hemispheres, unlike adults, a greater proportion involve midline structures such as thalamus and brainstem; and (2) they usually have a different genetic profile: many do not express the IDH, TP53, or ATRX mutations that are more common in adult astrocytomas, and they are more likely have alterations in a variety of other genes such as MYB and BRAF.23 The higher grade pediatric astrocytomas located in the cerebral hemispheres have been reported to have alterations in a variety of different genes, including NTRKT, TP53, ATRX, SETD2, CDKN2A, and PDGFRA, with IDH, TERT, and EGFR alterations being relatively uncommon.24 H3 variant K27M mutations are a characteristic feature of many diffuse midline gliomas and have been associated with a worse prognosis compared to wildtype cases.25

Pilocytic Astrocytoma (WHO Grade I) Pilocytic astrocytoma is a low-grade glial tumor that can occur almost anywhere in the neuraxis.14,​26 Pilocytic astrocytomas are infrequently located in the cerebral cortex, but because of their slow-growing nature, they may be associated with a chronic seizure disorder. Pilocytic astrocytomas are typically well-­circumscribed masses that may be solid or cystic. The typical histological appearance is that of an astrocytic tumor with a biphasic pattern which may include both compact areas composed of densely fibrillary astrocytes with Rosenthal fibers, and looser areas with microcysts and occasional eosinophilic granular bodies (Fig. 5.5). Although cytologic atypia, rare mitoses, and microvascular proliferation may be focally present in some pilocytic astrocytomas, these features do not usually c­ onstitute evidence of anaplastic change in this tumor. Rare pilocytic astrocytomas may show higher than expected cell density and high mitotic activity; however, the biologic behavior of these “atypical” pilocytic astrocytomas

Fig. 5.5  Pilocytic astrocytoma. (a) Biphasic pattern with compact and loose areas. (b) Compact fibrillary zone with irregular pinkish-red Rosenthal fibers.

is not the same as the higher grade diffuse astrocytoma. True anaplastic transformation is extremely rare. The gene expression profiles of pilocytic astrocytoma differ from those of diffuse astrocytoma. Almost all pilocytic astrocytomas are characterized by the presence of genetic mutations coding for proteins in the MAPK pathway. The most frequent alteration in pilocytic astrocytoma is tandem duplication of 7q34 involving the BRAF gene resulting in oncogenic BRAF fusion proteins. Pilocytic astrocytomas do not show IDH or TP53 mutations or 1p/19q codeletion.

Pleomorphic Xanthoastrocytoma (WHO Grade II or III) Pleomorphic xanthoastrocytoma (PXA) is a relatively rare ­glial tumor that most commonly affects children and young adults.14,​27 Patients with PXAs often present with a long history of seizures. Almost all PXAs are supratentorial and most often located in the temporal lobe. They are superficial tumors, involving the leptomeninges and cerebrum, and are frequently cystic, and may have a mural nodule. Histologically PXAs consistent of large pleomorphic or bizarre-appearing cells are frequently m ­ ultinucleated, as well as spindle-shaped cells and lipidized cells (Fig. 5.6). Reticulin stains reveal a dense pericellular reticulin network. Almost all PXAs contain granular eosinophilic bodies. Mitoses are typically rare or absent and necrosis is rare. Immunohistochemistry consistently confirms the presence of glial differentiation with most tumor cells expressing GFAP and S100 ­protein; however, many PXAs also have cells expressing neuronal markers such as neurofilament, synaptophysin, and MAP2. This biphenotypic feature of PXA has suggested a possible ­origin of the tumor from multipotent neuroectodermal precursor cells, possibly located in the subpial region. Most PXAs generally have a favorable prognosis, especially when compared to diffuse astrocytoma, and are classified as WHO grade II tumors. Anaplastic PXAs (WHO grade III) have a significantly worse survival, and have a mitotic rate greater than 5 per 10 high power fields and often show necrosis.28 PXAs commonly express the BRAF V600E mutation, and this feature in conjunction with absence of a IDH mutation will help support the diagnosis.27

Fig. 5.6  Pleomorphic xanthoastrocytoma. (a) Pleomorphic tumor cells in a fascicular arrangement. (b) High magnification showing tumor cells that vary in size and shape, and contain multiple nuclei and/or vacuoles.

41

42

I  Introduction to Epilepsy in Children

„„ Hamartomas Hamartomas are tumor-like malformations that are composed of tissue elements normally present in a particular site but usually arranged in a disorganized manner. On imaging studies, these lesions may be difficult to distinguish from true neoplasms, hence the need for histological evaluation. The three major CNS hamartomas frequently associated with a seizure disorder are described below.

Glioneuronal and Glial Hamartomas Glioneuronal and glial hamartomas have been described in various cortical locations, particularly in the temporal and frontal lobes. They are mainly microscopic lesions and thus may be difficult to identify on imaging studies. Histologically they are composed of collections of mature glial cells and neurons that can be distinguished from normal cortex by their haphazard orientation and tendency toward clustering.29 These lesions may also be seen adjacent to FCD or some GNTs, in particular DNTs. Glial hamartomas are circumscribed clusters of glial cells with medium to large atypical nuclei; they are characteristically associated with neurofibromatosis 2 (NF2) and are primarily located in the molecular or deep layers the cerebral cortex.30

Hypothalamic Hamartomas Hypothalamic hamartomas are nodular nonneoplastic masses that usually arise from and are contiguous with the base of the hypothalamus.31 Some examples may lie free in the suprasellar space or interpeduncular cistern. Although many are incidental findings, some cases have been associated with gelastic seizures or endocrinologic abnormalities, in particular precocious puberty. Microscopically these lesions closely resemble normal hypothalamus and consist of mature neurons scattered haphazardly or clustered in a neuropil that may contain reactive astrocytes. Some lesions may show immunoreactivity for hypothalamic releasing hormones (e.g., gonadotropin-releasing hormone).

Fig. 5.7  Meningioangiomatosis. Cerebral cortex with nonneoplastic proliferation of spindle-shaped cells of meningothelial or fibroblast origin surrounding small vessels. Mature neurons are aberrantly distributed between the vessels.

Meningioangiomatosis Meningioangiomatosis is a plaque-like lesion composed of a nonneoplastic proliferation of meningothelial or fibroblast-like cells that involves the subarachnoid space and extends into the cerebral cortex around small vessels (Fig. 5.7).32 Sporadic cases usually present as a single cortical lesion; however, multifocal lesions and noncortical lesions may occur in NF2.30 The lesion mainly occurs in children and young adults, and because of its usual cortical location, patients can present with a chronic seizure disorder. The pathogenesis of this lesion is uncertain; a possible meningothelial origin is suggested by the presence of an associated meningioma in some cases (with or without associated NF2) and variable staining of the cells for epithelial membrane antigen (a marker for meningothelial/arachnoid cells).33

„„ Hippocampal (Mesial Temporal) Sclerosis Hippocampal sclerosis, also known as mesial temporal or Ammon’s horn sclerosis, is a well-established structural cause of chronic seizures.34–36 Although hippocampal sclerosis is most often identified in adults, the lesion has been increasingly r­ ecognized in children. Surgical treatment for patients with hippocampal sclerosis usually consists of resection of the affected hippocampus, amygdala, and portions of the temporal lobe, thus these structures will require histological evaluation, not only to confirm the diagnosis of hippocampal sclerosis but also to determine the presence of any dual pathology, such as cortical dysplasia, neoplastic or hamartomatous lesions, or vascular malformations. The predominant and characteristic microscopic features of hippocampal sclerosis are present in the hippocampus.34–36 This part is often removed and labeled as a separate specimen, which must be submitted entirely for adequate assessment of the histopathology. The hallmark of hippocampal sclerosis is the presence of marked loss of pyramidal neurons and gliosis affecting virtually all hippocampal subzones, with cornu ammonis 1 (CA1), and CA4 usually being the most severely involved (Fig. 5.8a). There is usually a relatively abrupt t­ransition

Fig. 5.8  Hippocampal (mesial temporal) sclerosis. (a) CA1 sector of hippocampus (pale center band) showing marked reduction in width and nearly total loss of pyramidal neurons. (b) Dentate gyrus with dispersion and looser packing of granule cell neurons.

5  Surgical Neuropathology of Pediatric Epilepsy from the severely affected CA1 zone to a normal-appearing subiculum. The CA1 sector will often be markedly reduced in thickness, and may also have prominent corpora amylacea. In some cases, neuronal loss and gliosis may primarily affect either the CA1 or CA4 sectors only, with relatively mild or no involvement of the other sectors. Most cases of hippocampal sclerosis will also show abnormalities of the dentate gyrus, typically dispersion or looser packing of the granule cell neurons as well as actual loss of these cells (Fig. 5.8b). Neuronal loss and gliosis can be assessed using standard hematoxylin-eosin staining; however, more challenging cases may require additional immunostaining for neuronal and glial markers. The recently proposed ILAE’s hippocampal sclerosis classification scheme,35 which is based on a semiquantitative assessment of neuronal loss in the hippocampal subsectors and dentate gyrus, may be helpful in the assessment of some cases. Mossy fiber sprouting is another characteristic feature of hippocampal sclerosis; however, its identification requires special histological methods (e.g., Timm silver staining or immunostaining for dynorphin A) which may not be routinely available. The histological appearance of the other structures removed at surgery—amygdala and temporal lobe—is more variable, and may range from only gliosis (often most conspicuous in the subpial region [Chaslin gliosis]) to focal neuronal loss. It also is important to identify discrete lesions (FCD, tumors, vascular lesions, etc.) that may be present in addition to hippocampal sclerosis in these temporal lobectomy specimens. These changes must be distinguished from focal lesions related to electroencephalography monitoring (e.g., depth electrode placement). The etiology and pathogenesis of hippocampal sclerosis is unclear, and it is still unresolved whether the lesion is a direct cause or a result of the seizure disorder; it is likely that both contribute to its pathogenesis.36 Most cases occur sporadically, and no consistent genetic etiology has been identified, with the possible exception of mutations in SCN1A found in some patients with hippocampal sclerosis (the SCN1A gene encodes a brain-expressed sodium channel).37 Patients with hippocampal sclerosis often have a history of febrile seizures in early childhood, but this is not an invariable finding; however it is known that severe prolonged seizures can lead to excessive release of the excitatory neurotransmitters glutamate and aspartate, which can result in neuronal cell death, especially in the ­hippocampus. Conversely the presence of a preexisting “developmental” abnormality is suggested by the dual occurrence of hippocampal sclerosis and other focal structural changes, such as FCD, certain GNTs such as DNT and GCT, and even more subtle malformation-like abnormalities such as the presence of excessive numbers of ectopic neurons in temporal white matter.

s­ eizures in these patients is most likely related to the presence of extensive perilesional gliosis, microhemorrhages, or hemosiderin deposition, in combination with local ischemic changes related to abnormal vascular shunting. Although these lesions are often described as “congenital,” they should be regarded as dynamic lesions that may continue to expand over time. The histology of these lesions is generally easy to recognize: AVMs are typically composed of variably-sized but usually fairly large abnormal-appearing vessels having the histological appearance of arteries, veins, or a “hybrid” of both types.39 There is no intervening capillary bed. The abnormal vessels are located in the overlying subarachnoid space and within the brain, where the vessels are typically separated by parenchyma that may show extensive gliosis and hemosiderin deposition, as well as necrotic areas suggestive of ischemic injury. Cavernomas are composed of contiguous thin-walled vascular channels of varying size that lack features of either arteries or veins, such as elastic tissue and smooth muscle.39 They are lined by a single layer of endothelium. Calcification and thrombosis may be present. Unlike AVMs they do not have a meningeal component and the vascular channels are usually not separated by intervening brain tissue. The brain surrounding a cavernoma typically shows extensive gliosis and hemosiderin deposition.

Sturge–Weber Syndrome Sturge–Weber syndrome (SWS) is a neurocutaneous disorder, caused by somatic mosaic mutations in the GNAQ gene, that is characterized by the presence of angiomas that involve the leptomeninges and skin of the face (port wine stain), typically in the distribution of the ophthalmic and maxillary branches of the trigeminal nerve.40 The meningeal angioma is most often unilateral, although bilateral involvement has been described; the parietal and occipital lobes are most often involved. The histopathology of the brain in SWS consists of two components: meningeal angiomas (Fig. 5.9a) and atrophy, gliosis, and extensive calcifications in the underlying cerebral cortex (Fig. 5.9b). The meningeal angioma has a unique appearance and differs from other brain vascular malformations by the presence of numerous thin-walled vessels resembling capillaries and small veins. The brain c­ hanges

„„ Vascular Malformations Arteriovenous and Cavernous Vascular Malformations Both arteriovenous malformations (AVMs) and cavernous angiomas (cavernomas) may occur in the brain in children and can be associated with a seizure disorder.3,​38 The etiology of the

Fig. 5.9  Sturge–Weber syndrome. (a) Leptomeningeal angioma consisting of numerous thin-walled vessels resembling capillaries and small veins. (b) Cerebral cortex adjacent to angioma showing calcifications and gliosis.

43

44

I  Introduction to Epilepsy in Children are considered to be secondary lesions, most likely related to hypoxic-ischemic injury due to venous stasis. A seizure disorder is frequently seen in patients with SWS, which may require surgical resection of the brain lesion in t­ reatment-refractory cases.

„„ Acquired Lesions Vascular and Traumatic Lesions Medically refractory seizures may develop in children who have experienced a prior vascular or traumatic insult during the perinatal period or early childhood. Surgical resection of an epileptogenic lesion may be beneficial in some of these patients.41 These lesions include ischemic strokes (e.g., porencephalic cysts), hypoxic-ischemic lesions (e.g., ulegyria), cortical contusions, and intracerebral hemorrhages. Because of their chronicity, these lesions histologically usually show only nonspecific gliosis and hemosiderin deposition if hemorrhage was initially present. Exact classification of the lesion will therefore depend on correlation of clinical, neuroimaging, and pathological findings. One specific vascular abnormality that may be present in children with a seizure disorder is ulegyria. This is a p ­ erinatal hypoxia-ischemic lesion that preferentially affects the cortex in the depths of sulci with greater preservation of the gyral crowns, resulting in “mushroom-shaped” gyri.42 Ulegyria may occur unilaterally or bilaterally, and most often involves the parietal and occipital lobes in the territory of the posterior cerebral artery or middle/posterior cerebral artery border zone region. The gross morphology of ulegyria is usually apparent on neuroimaging, and histopathological examination will show shrinkage of the cortex with marked neuronal loss and gliosis that is most severe in the depthsof a sulcus.

Inflammatory Lesions

require surgical hemispherectomy. The histological a ­ ppearance of Rasmussen’s encephalitis will vary depending on the duration of the condition.3,​44 Acute or subacute cases will show widespread microglial activation, microglial nodule formation with or without neuronophagia, variable perivascular cuffing primarily by T lymphocytes (CD8 > CD4), and gliosis (Fig. 5.10). Cases of longer duration show neuronal loss, which may be patchy and focal or sometimes more extensive (pan-­laminar), and marked gliosis, often with a “spongiotic” character. Inflammation is diminished. Patchy inflammation and myelin loss may be apparent in the subcortical white matter. The exact cause of Rasmussen’s encephalitis is unknown but it is generally thought to have a probable autoimmune ­etiology.44,​45 An early report suggesting an association between Rasmussen’s encephalitis and the presence of antibodies to the GluR3 subunit of the AMPA receptor has not been confirmed. An alternative suggestion is that Rasmussen’s encephalitis may be mediated by cytotoxic T-cell mediated injury to neurons and astrocytes that express the MHC class I antigen, possibly related to a prior viral infection.

Autoimmune Encephalitis Noninfectious encephalitis can be caused by or associated with specific antibodies directed against neuronal cell surface or intracytoplasmic proteins.44,​45 The most common autoimmune encephalitis described in children is anti-NMDA receptor encephalitis.45 The clinical manifestations of this ­disorder include abnormalities in cognition, memory and behavior as well as seizures. Surgical biopsy or resection of abnormal tissue in patients with autoimmune encephalitis is rarely indicated, and would be likely done only in diagnostically challenging cases. For this reason, the histopathological features of autoimmune encephalitis have not been w ­ ell-­described; however, they would likely resemble the encephalitic process present in Rasmussen’s encephalitis.

Infections A variety of infections occurring in children may be associated with a chronic seizure disorder, some of which may be due to an epileptogenic lesion amenable to surgical resection. Lesions in this category include chronic bacterial abscesses, tuberculomas, parasitic infections such as cysticercosis, and rarely chronic fungal infections.43 Some patients recovering from viral encephalitis (e.g., herpes virus type 1) may develop chronic epilepsy, although surgery is rarely necessary for seizure control in these patients. Descriptions of the histopathology of these lesions are readily available in a number of excellent references.

Noninfectious or Autoimmune Inflammatory Lesions Rasmussen’s Encephalitis Rasmussen’s encephalitis is probably the most common noninfectious encephalitis occurring in childhood. It typically presents with progressive refractory partial seizures, cognitive deterioration, and focal neurologic deficits that may eventually lead to atrophy of one hemisphere. Control of the seizure disorder may

Fig. 5.10  Rasmussen’s encephalitis. Cerebral cortex showing perivascular lymphocytic inflammation, and activated (elongated) microglial cells and lymphocytes in neuropil adjacent to neurons. This histological appearance is nonspecific and could be seen in other forms of autoimmune encephalitis, as well as in viral encephalitis.

5  Surgical Neuropathology of Pediatric Epilepsy

References 1. Blümcke I, Thom M, Aronica E, et al. The clinicopathologic spectrum of focal cortical dysplasias: a consensus classification proposed by an ad hoc Task Force of the ILAE Diagnostic Methods Commission. Epilepsia 2011;52(1):158–174 2. Aronica E, Becker AJ, Spreafico R. Malformations of cortical development. Brain Pathol 2012;22(3):380–401 3. Thom M, Sisodiya S, Najm I. Neuropathology of Epilepsy. In: Love S, Louis DN, Ellison DW, eds. Greenfield’s Neuropathology. 8th ed. London: Hodder Arnold; 2008:833–887 4. Blümcke I, Sarnat HB. Somatic mutations rather than viral infection classify focal cortical dysplasia type II as mTORopathy. Curr Opin Neurol 2016;29(3):388–395 5. Siedlecka M, Grajkowska W, Galus R, DembowskaBagińska B, Jóźwiak J. Focal cortical dysplasia: molecular disturbances and clinicopathological classification (Review). Int J Mol Med 2016;38(5):1327–1337 6. Iffland PH II, Crino PB. Focal cortical dysplasia: gene mutations, cell signaling, and therapeutic implications. Annu Rev Pathol 2017;12:547–571 7. Srikandarajah N, Martinian L, Sisodiya SM, et al. Doublecortin expression in focal cortical dysplasia in epilepsy. Epilepsia 2009;50(12):2619–2628 8. Leventer RJ, Guerrini R, Dobyns WB. Malformations of cortical development and epilepsy. Dialogues Clin Neurosci 2008;10(1):47–62 9. Liu JS. Molecular genetics of neuronal migration disorders. Curr Neurol Neurosci Rep 2011;11(2):171–178 10. Parrini E, Conti V, Dobyns WB, Guerrini R. Genetic basis of brain malformations. Mol Syndromol 2016;7(4):220–233 11. Devisme L, Bouchet C, Gonzalès M, et al. Cobblestone lissencephaly: neuropathological subtypes and correlations with genes of dystroglycanopathies. Brain 2012;135(Pt 2):469–482 12. Jansen A, Andermann E. Genetics of the polymicrogyria syndromes. J Med Genet 2005;42(5):369–378 13. Yoneda Y, Haginoya K, Kato M, et al. Phenotypic spectrum of COL4A1 mutations: porencephaly to schizencephaly. Ann Neurol 2013;73(1):48–57 14. Thom M, Blümcke I, Aronica E. Long-term epilepsy-associated tumors. Brain Pathol 2012;22(3):350–379 15. Blümcke I, Aronica E, Becker A, et al. Low-grade epilepsy-associated neuroepithelial tumours—the 2016 WHO classification. Nat Rev Neurol 2016;12(12):732–740 16. Pietsch T, Hawkins C, Varlet P, Blümcke I, Hirose T. Dysembryoplastic neuroepithelial tumor. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:132–135 17. Capper D, Becker AJ, Giannini C, et al. Gangliocytoma. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:136–137 18. Becker AJ, Wiestler OD, Figarella-Branger D, Blümcke I, Capper D. Ganglioglioma. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:138–141 19. Becker AJ, Wiestler OD, Figarella-Branger D, Blümcke I, Capper D. Anaplastic ganglioglioma. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:141 20. Reifenberger G, Collins VP, Hartmann C, et al. Oligodendroglioma, IDH-mutant and 1p/19q-codeleted. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:60–69

21. Reifenberger G, Collins VP, Hartmann C, et al. Oligodendroglioma, IDH-mutant and 1p/19q-codeleted. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:70–74 22. Louis DN, von Deimling A, Cavenee WK. Diffuse astrocytic and oligodendroglial tumours—Introduction. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:16–17 23. von Deimling A, Huse JT, Yan H, et al. Diffuse astrocytoma, IDHmutant. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:18–23 24. Louis DN, Suvà ML, Burger PC, et al. Glioblastoma, IDH-wildtype. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:28–45 25. Hawkins C, Ellison DW, Sturm D. Diffuse midline glioma, H3 K27M-mutant. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon International Agency for Research on Cancer; 2016:57–59 26. Collins VP, Tihan T, Vandenberg SR, et al. Pilocytic astrocytoma. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:80–88 27. Giannini C, Paulus W, Louis DN, et al. Pleomorphic xanthoastrocytoma. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: IARC; 2016:94–97 28. Giannini C, Paulus W, Louis DN, et al. Anaplastic pleomorphic xanthoastrocytoma. In: Louis DN, Ohgaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. 4th ed. Lyon: International Agency for Research on Cancer; 2016:98–99 29. Volk EER, Prayson RA. Hamartomas in the setting of chronic epilepsy: a clinicopathologic study of 13 cases. Hum Pathol 1997;28(2):227–232 30. Louis DN, Ramesh V, Gusella JF. Neuropathology and molecular genetics of neurofibromatosis 2 and related tumors. Brain Pathol 1995;5(2):163–172 31. Coons SW, Rekate HL, Prenger EC, et al. The histopathology of hypothalamic hamartomas: study of 57 cases. J Neuropathol Exp Neurol 2007;66(2):131–141 32. Wiebe S, Munoz DG, Smith S, Lee DH. Meningioangiomatosis. A comprehensive analysis of clinical and laboratory features. Brain 1999;122(Pt 4):709–726 33. Perry A, Kurtkaya-Yapicier O, Scheithauer BW, et al. Insights into meningioangiomatosis with and without meningioma: a clinicopathologic and genetic series of 24 cases with review of the literature. Brain Pathol 2005;15(1):55–65 34. Blümcke I, Coras R, Miyata H, Özkara C. Defining cliniconeuropathological subtypes of mesial temporal lobe epilepsy with hippocampal sclerosis. Brain Pathol 2012;22(3):402–411 35. Blümcke I, Thom M, Aronica E, et al. International consensus classification of hippocampal sclerosis in temporal lobe epilepsy: a Task Force report from the ILAE commission on diagnostic methods. Epilepsia 2013;54(7):1315–1329 36. Thom M. Review: Hippocampal sclerosis in epilepsy: a neuropathology review. Neuropathol Appl Neurobiol 2014;40(5):520–543 37. Kasperaviciute D, Catarino CB, Matarin M, et al; UK Brain Expression Consortium. Epilepsy, hippocampal sclerosis and febrile seizures linked by common genetic variation around SCN1A. Brain 2013;136(Pt 10):3140–3150

45

46

I  Introduction to Epilepsy in Children 38. Burch EA, Orbach DB. Pediatric central nervous system vascular malformations. Pediatr Radiol 2015;45(Suppl 3):S463–S472 39. Ferrer I, Kaste M, Kalimo H. Vascular diseases. In: Love S, Louis DN, Ellison DW, eds. Greenfield’s Neuropathology. 8th ed. London: Hodder Arnold; 2008:121–240 40. Stafstrom CE, Staedtke V, Comi AM. Epilepsy mechanisms in neurocutaneous disorders: tuberous sclerosis complex, neurofibromatosis type 1, and Sturge–Weber syndrome. Front Neurol 2017;8:87 41. Ghatan S, McGoldrick P, Palmese C, et al. Surgical management of medically refractory epilepsy due to early childhood stroke. J Neurosurg Pediatr 2014;14(1):58–67

42. Nikas I, Dermentzoglou V, Theofanopoulou M, Theodoropoulos V. Parasagittal lesions and ulegyria in hypoxic-ischemic enceph­ alopathy: neuroimaging findings and review of the pathogenesis. J Child Neurol 2008;23(1):51–58 43. Vezzani A, Fujinami RS, White HS, et al. Infections, inflammation and epilepsy. Acta Neuropathol 2016;131(2):211–234 44. Bauer J, Vezzani A, Bien CG. Epileptic encephalitis: the role of the innate and adaptive immune system. Brain Pathol 2012;22(3):412–421 45. Armangue T, Petit-Pedrol M, Dalmau J. Autoimmune encephalitis in children. J Child Neurol 2012;27(11):1460–1469

6

  Epilepsy and Brain Plasticity Shilpa D. Kadam and Michael V. Johnston

Summary The human brain continues to undergo significant developmental maturation, circuit development, and synaptic pruning during periods spanning the perinatal period to early childhood. It is therefore not surprising that seizures during this critical period can significantly affect and modulate the ongoing physiological brain plasticity. The brain not only has a tremendous capacity of plasticity transiently when adapting to perinatal insults, but also chronically to function optimally in spite of congenital or acquired lesions. The immature brain is highly susceptible to seizures due to inherent developmental features unique to the developing brain that are also critical in helping the immature brain to connect and develop in the postnatal period. However, all seizures are not equal. Etiologies along with severity of seizures have risen up to be some of the main underlying causes that determine long-term effects on brain plasticity and the ensuing comorbidities. Early life insults that result in cell-death and structural lesions like focal sclerosis and white matter tract alterations are now highlighted as predictors of ongoing epileptogenesis that underlie the disease progression and can result in the development of additional epileptogenic foci at distant sites. Early surgical intervention to remove primary structural foci has shown promise in preventing further epileptogenesis. Failures related to seizure recurrence are more common with late surgical interventions or when secondary foci have already developed distant to the original site of injury or seizure generation. Keywords:  kindling, mossy fiber sprouting, intractable epilepsy, inflammation, neurogenesis, epileptogenesis

„„ Introduction: Seizure Susceptibility and Transient Plasticity in the Developing Brain Plasticity in the developing brain has been categorized under many systematic subsections1 as related to their adaptive or maladaptive nature. These time-sensitive responses portend heightened plasticity during critical and sensitive periods, allowing for windows of opportunity for interventions. The developing brain has a tremendous capacity for plasticity as it undergoes developmental maturation, both structural and functional and, in response to outside stimuli both

helpful and injurious. Adaptive and temporally progressive genetic, molecular, and cellular mechanisms underlie this plasticity as well as the critical “window periods” of optimal malleability.2,​3 The mechanisms underlying seizures in the immature brain and the long-term consequences of epileptic neuronal activity are better understood through the revised classification system for the epilepsy subtypes4,​5 and the seizure burden, and the excitotoxic cellular injury associated with them.6 The immature brain is more susceptible to seizures than the adult brain in many ways7 and seizures themselves have been linked to many neuroplastic changes in hippocampal circuits, neurogenesis, and dendritic growth.8,​9

„„ Etiology of Seizure Onset and Long-term Plasticity Human data on long-term structural plasticity in the brain are predominantly based on postmortem histopathology. Another source of brain histopathology and electrophysiological recordings comes from tissue specimens resected during surgery from patients who have epilepsy that is pharmacologically unresponsive (“intractable”) to antiseizure drugs. This is usually when the surgical option of removing the area thought to trigger seizures is sought. This situation often occurs in temporal lobe epilepsy (TLE)10,​11,​12 and cortical dysplasias,13 and the areas resected may include not only the hippocampus but also adjacent cortical regions and the amygdala. Hippocampal histopathology has revealed reorganization of the neural circuitry of the dentate gyrus that favors the hyperexcitability of dentate granule cells (DGCs). These changes include the loss of populations of subgranular hilar interneurons, hilar mossy cells, and granule cells. Additionally, there is sprouting of granule cell mossy fiber recurrent collaterals into the dentate inner molecular layer and the sprouting of neuropeptide Y, somatostatin, and substance P containing axons throughout the molecular layer.14,​15 Findings from these resected and postmortem tissues have given us tremendous insights into possible plasticity mechanisms underlying epileptogenesis and associated neurological comorbidities. They also help the research community design appropriate animal models that both replicate the human tissue findings and reproduce the clinical features of the respective epilepsy.16,​17 Follow-up MRIs in children with history of neonatal encephalopathies are also proving to be predictive

48

I  Introduction to Epilepsy in Children of long-term comorbidities, especially language outcomes and verbal IQ.18,​19 Research on how a single seizure versus multiple recurrent seizures may result in these long-term plasticity changes that sometimes result in the emergence of drug-resistant epilepsy that requires surgical intervention has yielded interesting insights.20 Given the variability of clinical presentations, it remains challenging to directly address these types of questions in human studies. Here, immature rodent models of epilepsy have offered important advantages, including control of preexisting factors, and of seizure number, duration, severity and location. When evaluated in rodent pups, prolonged seizures alone did not cause neuronal cell death in the immature brain. However, when the same seizures were superimposed on mild to moderate hypoxic–ischemic injury, neuropathologic injury increased significantly,21 particularly in the hippocampus highlighting etiology and seizure burden as critical factors in long-term outcomes. Animal models, however, also show differential molecular and functional consequences based on the method by which seizures are induced in the model to begin with, which can complicate translational value if care is not taken.5 Clinically it is now clear that repeated seizures are just as likely to produce neuronal damage, which was earlier thought to be only likely in cases of status epilepticus.22,​23 These long-term progressive structural and functional alterations in response to seizures are evident from longitudinal imaging studies and long-term neuropsychological studies in epilepsy patients, and are also replicated in models of chronic progressive epilepsy.24,​25

„„ Seizures Beget Seizures? “Seizures beget seizures” is a long discussed and debated concept that indicates that each seizure experienced increases the risk for a subsequent seizure, but is only partly supported by clinical observations.26,​27 Whether this increase in probability is due to the progression of an underlying pathology or whether each spontaneous seizure itself induces brain changes that lower the threshold for a subsequent seizure are some of the questions being investigated using a multitude of model systems. It is important, especially when it comes to clinical relevance, to understand that epileptogenesis, pharmacoresistance, and seizures begetting seizures may not all be the same thing.28 In neuroscience research, kindling is understood as the process wherein repeatedly induced seizures result in a graded increase in induced seizure durations with increasing behavioral involvement until a plateau is reached on a graded scale (e.g., Racine scale). Kindling is one of the most widely used models of seizures and epilepsy wherein seizures can be induced by focal electrical stimulation in the brain.29 TLE generally is thought of as progressive over time.30 Specifically, structural damage is more pronounced in individuals with a longer duration of epilepsy, and damage progresses over time as documented by neuroimaging. Progressive epilepsy has been documented in the natural history of certain animal models that show TLE-like hippocampal sclerosis and spontaneous seizures.25,​31 Progression in TLE and the possibility of secondary epileptogenesis in humans is evident with documented development of an independent contralateral epileptogenic focus.32 This phenomenon is associated with a less favorable surgical outcome.

„„ Neurogenesis in Epilepsy Neural stem cells that persist in the subgranular zone of the hippocampal dentate gyrus generate DGCs and the subventricular zone of the forebrain lateral ventricles throughout life. While the majority of cells are generated early in life, new DGCs arise at a lower rate throughout adulthood and into senescence in humans.33 Adult-born DGCs integrate into the preexisting hippocampal circuitry as they mature and, over time, acquire electrophysiological characteristics similar to their neighbors. DGC neurogenesis seems to be critical in certain forms of hippocampus-dependent learning and memory and in the modulation of emotional behavior or anxiety. Prolonged seizures can acutely increase adult DGC neurogenesis, and the functional implications of altered neurogenesis in mesial TLE are a topic of heavy investigation. Interestingly, seizures early in life alter DGC birth in different and even opposing fashion to adult seizure models.34 The postnatal age at which seizures occur in rodent models seem to play an important role. There is also evidence to show that altered neurogenesis does contribute to several well-characterized cellular abnormalities documented in human mesial TLE like mossy fiber sprouting, DGC layer dispersion, and ectopic DGCs. Epilepsy with evidence of hippocampal pathology such as Ammon’s horn sclerosis in infants and young children is associated with increased numbers of stem and dividing cells compared with age matched controls.35,​36 However, severe seizures during early childhood were associated with anatomic signs of decreased postnatal granule cell neurogenesis (e.g., PSA-NCAM IR [polysialylatedneural cell adhesion molecule immunoreactivity]—a candidate marker for neuronal plasticity) and aberrant mossy fiber axon connections (neo-Timm’s: for visualizing zinc-containing neuronal axons that help the detection of newly sprouted axons and axon terminals within the central nervous system) without evidence of seizure-induced cell death indicating the increased number of dividing stem cells found in the other studies may be maturing into nonneuronal cell types instead of becoming DGCs.37 Whether, these long-term changes are adaptive or epileptogenic over time remains part of the focus in the research field.

„„ Cell Death and Axonal Sprouting Axonal sprouting and synaptic reorganization following celldeath in the process of acquired epileptogenesis has been a significant part of the epilepsy research field for a long time.38 A specific example is Timm’s stain in the inner molecular layer of the dentate gyrus (i.e., also called mossy fiber sprouting representing recurrent excitation) from patients with, and animal models of, spontaneous seizures following neonatal ischemic insults (Fig. 6.1). The concept of axonal sprouting has encompassed among many other hypotheses: (1) recurrent excitation39 where pyramidal neurons in the dentate gyrus and CA3 of the hippocampus sprout axons to innervate neighboring pyramidal cells; and (2) loss of GABAergic neurons during a primary insult resulting in a compensatory mechanism where surviving GABAergic neurons sprout axons to either excite other GABAergic neurons and increase overall inhibitory tone in the region of the cell loss, or sprout collaterals and increase synapses with pyramidal neurons to make up for any lost inhibitory control. The counter play, or fine balance between the compensatory

6  Epilepsy and Brain Plasticity

Fig. 6.1  Representative coronal cresyl-violet and Timm-stained sections focusing on the ipsi- and contralateral dorsal hippocampi from control and epileptic Sprague Dawley rats following perinatal ischemia. Top and bottom panels show ipsilateral cresyl-violet (a, g), Timm-stained sections (b, h) and magnified views of the inner molecular layer of the dentate gyrus (c, i) from a control and hypoxia–ischemia (HI)-treated epileptic rat, respectively. Middle panel (d–f) shows the corresponding contralateral sections from the same HI-treated epileptic rat as shown in (c). The dorsal hippocampus from a control rat (a) shows fascia dentata (FD) and the Ammon’s horn (CA3, CA2, and CA1). The Timm-stained section shows dark brown precipitate in the region of mossy fiber innervation (b). Grade 0 Timm-stained product is seen in dentate inner molecular layer (c). The contralateral dorsal hippocampus (d) shows no apparent cell loss of CA3 neurons. Timm stain shows a more robust distribution compared to control (e) and minimal (grade 1) stain product is seen in dentate inner molecular layer (f). Massive loss of lateral CA3 neurons causing a distinct demarcation between the CA3 and CA2 (relatively spared) junction is seen in the ipsilateral region of the epileptic rat (g); note the relative preservation of the FD medial CA3 and hilar neurons. Robust innervation of the CA3 region till the CA3/CA2 junction by mossy fibers masks presence of the few surviving neurons in the region by cresyl violet counterstain (h). The dentate inner molecular layer shows grade 3 stain product (i); note atrophy of ipsilateral dorsal hippocampus in ipsilateral compared to the contralateral and control sections. Scale bars = 250 μm in (h) (also applies to a, b, d, e, g); 50 μm in (i) (also applies to c, f). (Reproduced with permission from Kadam, Dudek 2007.17)

inhibition masking the increased recurrent excitation following axonal sprouting over time has been hypothesized to underlie the episodic nature of seizures in epilepsy wherein recurrent excitation overrides the inhibitory surround that allows waves of synchronized firing in an epileptic brain. Epilepsy following an initial pediatric or neonatal insult usually involves a latent period before the occurrence of the first spontaneous seizure in acquired forms of epilepsies. One of the reasons that synaptic reorganization, particularly axon sprouting and the formation of new recurrent excitatory

circuits, is extensively investigated is because this mechanism would be expected to provide a slow and continuous process that could contribute to the latent periods seen in many patients and animal models of temporal lobe and other forms of acquired epilepsy. A lot is still unclear with these hypotheses. For example, numerous types of interneurons are present in the hippocampus and cortex, and the question is whether all or only some of them undergo reorganization. Do interneurons inhibit other interneurons and thereby decrease of GABAergic inhibition? Are excitatory and inhibitory receptors at these new

49

50

I  Introduction to Epilepsy in Children synapses altered permanently? Studies conducted using surgical specimens resected from individuals with medically refractory TLE indicate that both excitatory and inhibitory amino acid receptors are altered in the hippocampus of patients with TLE40 and associated with a significant reduction in benzodiazepine receptor binding in CA1 and CA4. Similarly, prolonged and focal complex febrile convulsions have been shown to produce acute hippocampal injury that evolves into hippocampal atrophy, however some of these hippocampal abnormalities predate the first clinical seizure,41,​42 thus complicating our understanding of cause and effect in such situations.

„„ Inflammation and Epilepsy Brain inflammation, which is identified as a common substrate contributing to epileptogenesis, specifically in drug-resistant epilepsies of different etiologies, is a focus of ongoing research. Recurrent seizures in drug-resistant epilepsies themselves can be a cause of the persistence of long-term inflammation.43,​44 Following this view, new research has shown that the pharmacological blockade of specific inflammatory molecules and pathways can significantly reduce seizures in experimental models of seizures and epilepsy.45 Similarly, the role of inflammation in seizure development following traumatic brain injury using interleukin 1 (IL-1) beta as a potential biomarker and the use of targeted IL-1 beta therapies to help reduce post-trauma epilepsy have generated interest.46 From a clinical standpoint, the role of inflammation in the pathophysiology of human epilepsy is still hypothetical, although this possibility is supported by abundant evidence.47 Recent studies have highlighted the potential of the IL-1RA/IL-6 ratio as a biomarker of acute hippocampal injury following febrile status epilepticus.48 Advances such as these could help identify patients at risk for ultimately developing mesial TLE very early on. Increased blood–brain barrier (BBB) permeability and changes in endothelial transporter function induced by both seizures and inflammation have significant clinical implications. BBB leakage allows the entry of compounds with immunogenic or inflammatory potentials, as predicted, for example, in interferon-induced seizures,49 but it may also facilitate the entry of compounds with therapeutic potential with limited or no access to the central nervous system.

„„ Lesional Epilepsy and Mirror Focus Development The concept of a mirror focus has been considered in the clinical literature for decades, and numerous examples of clinical observations have suggested its occurrence in humans.15,​50–52 Long-term, continuous radio-telemetric recordings in a model of progressive epilepsy following perinatal hypoxia–ischemia25 showed examples of contralateral seizure onsets from nonlesioned cortex (compared to ipsilateral lesioned due to unilateral induced perinatal ischemia). Thus, although major structural differences between the ipsi- and contralateral cortex are present in the model over the 6-month period of continuous EEG monitoring in the study, overtime, the contralateral homotypic cortex appeared to reorganize in such a manner that its electrophysiological

properties at the level of electrically evoked local field poetical appeared to evolve and look very similar to the ipsilateral injured cortex. Similar findings recently reported with human TLE32 help validate the translational value of models of acquired epilepsy.

„„ Possible Clinical Implications Clinical studies investigating the efficacy of resective surgery as a potential cure in children with lesional extratemporal epilepsies have shown promise.53,​54 However, these resective surgeries may remain underutilized as a potential cure, even with evidence that seizure freedom is more likely when surgery is done within 7 years of seizure history.55 Longer seizure history (> 7 years) appears to result in a lower success rate. The possible emergence of independent epileptogenic foci at distant sites may make epilepsy, which could have been treatable with early resection of the initial epileptogenic zone, far less amenable to surgical treatment. In addition, psychiatric comorbidities associated with epilepsy also seem to be a predictor of long-term seizure outcomes following TLE surgery,56 and highlight the prognostic value of presurgical psychiatric evaluations. Additionally, functional plasticity in the contralateral hemisphere can be “unmasked” by resective surgery. An example of such unmasking was documented in a Rasmussen’s patient whose language area had moved over from the left to the right hemisphere due to early childhood seizures on the left side, but this new area was not usable until the epileptic focus was surgically removed with a left hemispherotomy (Fig. 6.2). On functional MRI, the right activation was seen mainly in regions that could not be detected preoperatively, but mirrored those previously found in the left hemisphere, suggesting reorganization in a preexisting bilateral network.57

„„ Conclusion About 30% of children with epilepsy continue to have seizures even while taking medicine. The percent population with intractable seizures has remained stable in spite of the slew of new antiseizure drugs that have entered clinical practice over the last decade. There are risks with any surgery; however, it is also important to know that there are risks of having uncontrolled seizures. As mentioned above, continued seizures can change how a brain works. It is important to stop seizures before they affect a child’s development. As discussed above, the developing brain of a child is much more “plastic” than an adult brain. This also means that it is possible for functions like movement and language to be controlled by a new part of the brain if the original area is damaged. The adult brain loses this ability beyond the “critical developmental window period.” For this reason, when indicated, it may be important to have surgery as early as possible. Significant improvements in surgical methods and reliability of determining seizure focus have considerably improved seizure control success rates in these refractory cases. Currently, three major categories of pediatric epilepsy surgery are in practice: resective surgery, corpus callosotomy, and implantation of a vagus nerve stimulator. Patients with obvious structural lesions on MRI scanning may be appropriate epilepsy surgery candidates,

6  Epilepsy and Brain Plasticity

Fig. 6.2  Plasticity in contralateral hemisphere unmasked by late surgical intervention. (a) Presurgery functional MRI. Statistical parametric map of word generation compared with rest, overlaid on the statistical parametric mapping (SPM) template. White lines indicate the extent of slice acquisition. Height threshold, p < 0.01 (uncorrected, at a voxel-level), with corrected extent threshold of p < 0.227 per cluster. Left-sided activated network involves inferior frontal gyrus, supplementary sensory-motor area, and supramarginal gyrus (p < 0.07 in each cluster; Z > 3.28). Note that the activation in the right supramarginal gyrus does not reach cluster-level significance (p = 0.227; Z = 4.26). (b) Postsurgery functional MRI after left hemisphere disconnection in same patient. Note left hemisphere atrophy and brain resection. Height threshold, p < 0.001 (uncorrected at a voxel level), with corrected extent threshold of p < 0.05 per cluster. Statistical parametric map of word generation compared with rest. Rightlateralized activated network involves mostly frontal regions (inferior and middle frontal gyri, precentral gyrus, supplementary motor area) and a few superior and inferior temporal regions, as well as the supramarginal gyrus. Activation can be also observed on the left in the supramarginal gyrus, and bilaterally in occipital cortex and frontal poles. (Reproduced with permission from Hertz-Pannier et al 2002.57)

especially if their seizures are focally generated. As described above there are consequences to continued seizures, especially in the developing brain of a child and young infant. In these cases, progression can be quantified not only in progressive epilepsy which may be masked as a result of helpful plasticity

in the form of GABAergic axonal sprouting that may help contain the excitatory circuits, but also in the progression of functional anomalies like but not limited to school academic failure, poor self-esteem, anxiety, depression, and higher risk of suicide in a child diagnosed with epilepsy.

References 1. Ismail FY, Fatemi A, Johnston MV. Cerebral plasticity: windows of opportunity in the developing brain. Eur J Paediatr Neurol 2017;21(1):23–48 2. Johnston MV. Plasticity in the developing brain: implications for rehabilitation. Dev Disabil Res Rev 2009;15(2):94–101 3. Johnston MV, Ishida A, Ishida WN, Matsushita HB, Nishimura A, Tsuji M. Plasticity and injury in the developing brain. Brain Dev 2009;31(1):1–10 4. Fisher RS, Cross JH, French JA, et al. Operational classification of seizure types by the International League Against Epilepsy: position Paper of the ILAE Commission for classification and terminology. Epilepsia 2017;58(4):522–530 5. Kang SK, Kadam SD. Neonatal Seizures: impact on neurodevelopmental outcomes. Front Pediatr 2015;3:101 6. Pitkänen A, Sutula TP. Is epilepsy a progressive disorder? Prospects for new therapeutic approaches in temporal-lobe epilepsy. Lancet Neurol 2002;1(3):173–181 7. Rakhade SN, Jensen FE. Epileptogenesis in the immature brain: emerging mechanisms. Nat Rev Neurol 2009;5(7):380–391 8. Scharfman HE. Epilepsy as an example of neural plasticity. Neuroscientist 2002;8(2):154–173 9. Naegele J. Epilepsy and the plastic mind. Epilepsy Curr 2009;9(6):166–169

10. de Lanerolle NC, Lee TS, Spencer DD. Histopathology of Human Epilepsy. In Noebels JL, Avoli M, Rogawski MA, et al., eds. Jasper's Basic Mechanisms of the Epilepsies. 4th edition. Maryland, US: National Center for Biotechnology Information; 2012. 11. Margerison JH, Corsellis JA. Epilepsy and the temporal lobes. A clinical, electroencephalographic and neuropathological study of the brain in epilepsy, with particular reference to the temporal lobes. Brain 1966;89(3):499–530 12. Zaveri HP, Duckrow RB, de Lanerolle NC, Spencer SS. Distinguishing subtypes of temporal lobe epilepsy with background hippocampal activity. Epilepsia 2001;42(6):725–730 13. Andres M, Andre VM, Nguyen S, et al. Human cortical dysplasia and epilepsy: an ontogenetic hypothesis based on volumetric MRI and NeuN neuronal density and size measurements. Cereb Cortex 2005;15(2):194–210 14. Pitkänen A, Sutula TP. Is epilepsy a progressive disorder? Prospects for new therapeutic approaches in temporal-lobe epilepsy. Lancet Neurol 2002;1(3):173–185 15. Teyler TJ, Morgan SL, Russell RN, Woodside BL. Synaptic plasticity and secondary epileptogenesis. Int Rev Neurobiol 2001;45:253–267 16. Majores M, Schoch S, Lie A, Becker AJ. Molecular neuropathology of temporal lobe epilepsy: complementary approaches

51

52

I  Introduction to Epilepsy in Children in animal models and 2007;48(Suppl 2):4–12

human

disease

tissue.

Epilepsia

17. Kadam SD, Dudek FE. Neuropathogical features of a rat model for perinatal hypoxic-ischemic encephalopathy with associated epilepsy. J Comp Neurol 2007;505(6):716–737 18. Steinman KJ, Gorno-Tempini ML, Glidden DV, et al. Neonatal watershed brain injury on magnetic resonance imaging correlates with verbal IQ at 4 years. Pediatrics 2009;123(3):1025–1030 19. Shapiro KA, Kim H, Mandelli ML, et al. Early changes in brain structure correlate with language outcomes in children with neonatal encephalopathy. Neuroimage Clin 2017;15:572–580 20. Baram TZ. Long-term neuroplasticity and functional consequences of single versus recurrent early-life seizures. Ann Neurol 2003;54(6):701–705 21. Wirrell EC, Armstrong EA, Osman LD, Yager JY. Prolonged seizures exacerbate perinatal hypoxic-ischemic brain damage. Pediatr Res 2001;50(4):445–454 22. Sutula TP, Pitkänen A. More evidence for seizure-induced neuron loss: is hippocampal sclerosis both cause and effect of epilepsy? Neurology 2001;57(2):169–170 23. Sutula TP, Hagen J, Pitkänen A. Do epileptic seizures damage the brain? Curr Opin Neurol 2003;16(2):189–195 24. Sutula TP. Mechanisms of epilepsy progression: current theories and perspectives from neuroplasticity in adulthood and development. Epilepsy Res 2004;60(2–3):161–171 25. Kadam SD, White AM, Staley KJ, Dudek FE. Continuous electroencephalographic monitoring with radio-telemetry in a rat model of perinatal hypoxia-ischemia reveals progressive post-stroke epilepsy. J Neurosci 2010;30(1):404–415 26. Morimoto K, Fahnestock M, Racine RJ. Kindling and status epilepticus models of epilepsy: rewiring the brain. Prog Neurobiol 2004;73(1):1–60 27. Berg AT, Shinnar S. Do seizures beget seizures? An assessment of the clinical evidence in humans. J Clin Neurophysiol 1997;14(2):102–110 28. Sills GJ. Seizures beget seizures: a lack of experimental evidence and clinical relevance fails to dampen enthusiasm. Epilepsy Curr 2007;7(4):103–104 29. Bertram E. The relevance of kindling for human epilepsy. Epilepsia 2007;48(Suppl 2):65–74 30. Coan AC, Cendes F. Epilepsy as progressive disorders: what is the evidence that can guide our clinical decisions and how can neuroimaging help? Epilepsy Behav 2013;26(3):313–321 31. Williams PA, White AM, Clark S, et al. Development of spontaneous recurrent seizures after kainate-induced status epilepticus. J Neurosci 2009;29(7):2103–2112 32. Gollwitzer S, Scott CA, Farrell F, et al. The long-term course of temporal lobe epilepsy: From unilateral to bilateral interictal epileptiform discharges in repeated video-EEG monitorings. Epilepsy Behav 2017;68:17–21 33. Eriksson PS, Perfilieva E, Björk-Eriksson T, et al. Neurogenesis in the adult human hippocampus. Nat Med 1998;4(11):1313–1317 34. Porter BE. Neurogenesis and epilepsy in the developing brain. Epilepsia 2008;49(Suppl 5):50–54 35. Blümcke I, Schewe JC, Normann S, et al. Increase of nestinimmunoreactive neural precursor cells in the dentate gyrus of pediatric patients with early-onset temporal lobe epilepsy. Hippocampus 2001;11(3):311–321 36. Takei H, Wilfong A, Yoshor D, Armstrong DL, Bhattacharjee MB. Evidence of increased cell proliferation in the hippocampus in children with Ammon’s horn sclerosis. Pathol Int 2007;57(2):76–81

37. Mathern GW, Leiphart JL, De Vera A, et al. Seizures decrease postnatal neurogenesis and granule cell development in the human fascia dentata. Epilepsia 2002;43(Suppl 5):68–73 38. Dudek FE. Axon sprouting and synaptic reorganization of GABAergic interneurons: a focused look at a general question. Epilepsy Curr 2010;10(5):126–128 39. Dudek FE, Sutula TP. Epileptogenesis in the dentate gyrus: a critical perspective. Prog Brain Res 2007;163:755–773 40. McDonald JW, Garofalo EA, Hood T, et al. Altered excitatory and inhibitory amino acid receptor binding in hippocampus of patients with temporal lobe epilepsy. Ann Neurol 1991;29(5):529–541 41. Lewis DV, Barboriak DP, MacFall JR, Provenzale JM, Mitchell TV, VanLandingham KE. Do prolonged febrile seizures produce medial temporal sclerosis? Hypotheses, MRI evidence and unanswered questions. Prog Brain Res 2002;135:263–278 42. VanLandingham KE, Heinz ER, Cavazos JE, Lewis DV. Magnetic resonance imaging evidence of hippocampal injury after prolonged focal febrile convulsions. Ann Neurol 1998;43(4):413–426 43. Vezzani A, Auvin S, Ravizza T, Aronica E. Glia-neuronal interactions in ictogenesis and epileptogenesis: role of inflammatory mediators. Jasper's Basic Mechanisms of the Epilepsies. 4th ed. 2012 44. Vezzani A, Aronica E, Mazarati A, Pittman QJ. Epilepsy and brain inflammation. Exp Neurol 2013;244:11–21 45. Choi J, Koh S. Role of brain inflammation in epileptogenesis. Yonsei Med J 2008;49(1):1–18 46. Diamond ML, Ritter AC, Failla MD, et al. IL-1β associations with posttraumatic epilepsy development: a genetics and biomarker cohort study. Epilepsia 2014;55(7):1109–1119 47. Vezzani A, Granata T. Brain inflammation in epilepsy: experimental and clinical evidence. Epilepsia 2005;46(11):1724–1743 48. Gallentine WB, Shinnar S, Hesdorffer DC, et al; FEBSTAT Investigator Team. Plasma cytokines associated with febrile status epilepticus in children: A potential biomarker for acute hippocampal injury. Epilepsia 2017;58(6):1102–1111 49. Pavlovsky L, Seiffert E, Heinemann U, Korn A, Golan H, Friedman A. Persistent BBB disruption may underlie alpha interferon-induced seizures. J Neurol 2005;252(1):42–46 50. Gilmore R, Morris H III, Van Ness PC, Gilmore-Pollak W, Estes M. Mirror focus: function of seizure frequency and influence on outcome after surgery. Epilepsia 1994;35(2):258–263 51. Kim J, Shin HK, Hwang KJ, et al. Mirror focus in a patient with intractable occipital lobe epilepsy. J Epilepsy Res 2014;4(1):34–37 52. Morrell F, deToledo-Morrell L. From mirror focus to secondary epileptogenesis in man: an historical review. Adv Neurol 1999;81:11–23 53. Englot DJ, Breshears JD, Sun PP, Chang EF, Auguste KI. Seizure outcomes after resective surgery for extra-temporal lobe epilepsy in pediatric patients. J Neurosurg Pediatr 2013;12(2):126–133 54. Englot DJ, Chang EF. Rates and predictors of seizure freedom in resective epilepsy surgery: an update. Neurosurg Rev 2014;37(3):389–404, discussion 404–405 55. Cascino GD. Temporal lobe epilepsy is a progressive neurologic disorder: Time means neurons! Neurology 2009;72(20): 1718–1719 56. Koch-Stoecker SC, Bien CG, Schulz R, May TW. Psychiatric lifetime diagnoses are associated with a reduced chance of seizure freedom after temporal lobe surgery. Epilepsia 2017;58(6): 983–993 57. Hertz-Pannier L, Chiron C, Jambaqué I, et al. Late plasticity for language in a child’s non-dominant hemisphere: a pre- and postsurgery fMRI study. Brain 2002;125(Pt 2):361–372

7

 E  ffects of Seizures and Their Comorbidities on the Developing Brain Susan Lee Fong and Carl E. Stafstrom

Summary A seizure is a paroxysmal change in neurological function or behavior secondary to excessive, hypersynchronous discharges of neurons. These pathological electrical discharges manifest as clinically observable seizures, which are undoubtedly the most striking presentation of epilepsy. It is not surprising, then, that recurrent seizures would prompt concern about potentially ­harmful consequences of seizures on the developing brain. Epilepsy is also known to have many associated cognitive and psychiatric comorbidities, which can have considerable impact on quality of life. When considering the effects of seizures on the developing brain, several questions arise. Are the comorbid conditions a consequence of brain injury secondary to seizures? What are the neurobiological mechanisms underlying the many impairments faced by patients with epilepsy? Do current treatment modalities prevent seizure-related brain damage? In this chapter, we address these questions, review the data on behavioral changes associated with epilepsy, and explore the relationship between seizures and the developing brain. The overall goal, in line with the theme of this volume, is to discuss the role of epilepsy surgery in alleviating seizure-associated effects on the developing brain. Keywords:  epilepsy, early-life seizure, cognition, psychiatric, comorbidity

„„ Comorbidities in Epilepsy Although epilepsy is defined by recurrent seizures, there is growing appreciation of the profound impact of the many comorbid conditions that are associated with epilepsy.1 Neurological, behavioral, and psychiatric comorbidities—such as learning disabilities, cognitive deficits, depression, anxiety, attention deficit, hyperactivity, social dysfunction, and autism—are associated with significant morbidity and have a profound impact on quality of life for patients and their families. Compared to the general population, there is an increased incidence of these comorbid conditions in individuals with epilepsy (Table 7.1), and by some accounts,2,​3 the burden of the comorbidities can outweigh that of the seizures themselves.

Cognitive Deficits Among the myriad comorbidities associated with epilepsy, cognitive impairment is one of the most common and concerning.

Cognitive impairment occurs more frequently in children with epilepsy,4,​5 with approximately 25% of patients demonstrating evidence of cognitive dysfunction.6 A broad range of skills are impacted in children with epilepsy, including arithmetic, reading, spelling, writing, processing speed, memory, and executive function.7,​8 Even patients with “uncomplicated” epilepsy, who have normal neurological examinations, normal neuroimaging, and normal overall intelligence, often require special education services,4,​9 suggesting the presence of subtle cognitive disturbances. Although the association between epilepsy and cognitive impairment is well established, the mechanisms underlying this association are less clear. On cursory evaluation, it may appear that core clinical features of seizures, such as seizure type, age of onset, seizure frequency and duration, and epilepsy syndromes, can explain the heterogeneity with regard to cognitive outcomes in children with epilepsy. Indeed, the natural history of certain pediatric epilepsy syndromes is such that clinicians may be able to provide some prognostic information about expected neurodevelopmental outcome. For example, epileptic encephalopathies, including Ohtahara, West, Dravet, Lennox–Gastaut syndrome, and Landau–Kleffner syndrome, are by definition characterized by drug-resistant epilepsy and progressive ­cognitive i­mpairment,10,​ 11 and there is some evidence that a decrease in seizures in the epileptic encephalopathies is associated with improved cognitive outcome.12,​13 In contrast, benign focal epilepsies of childhood, such as benign childhood e ­ pilepsy with centrotemporal spikes, have no associated structural lesions, easily controlled focal seizures, and Table 7.1  Prevalence of comorbidities associated with epilepsy

Comorbid diagnosis

Prevalence

Any psychopathology

up to 66%

Attention deficit hyperactivity disorder

30–40%

Anxiety

10–30%

Depression

~20%

Suicidal ideation

20%

Psychosis or psychotic features

2–8%

Intellectual disability

25%

Autism

~20%

Psychosocial impairment

~40%

Source: Reproduced with permission from Camfield and Camfield 20175 and Filippini M, Gobbi G. Behavioural problems in childhood epilepsy. In: Lagae L, ed. Cognition and Behaviour in Childhood Epilepsy. London: Mac Keith Press; 2017:17–42.

54

I  Introduction to Epilepsy in Children relatively subtle, if any, cognitive deficits.14,​15 Data from a large community-based prospective cohort study showed that epileptic encephalopathy is an independent predictor of poor cognitive performance and that this subgroup of patients has significantly worse cognitive functioning than all other nonepileptic encephalopathies combined.6 Other factors associated with cognitive function included age of seizure onset, remission, and current use of antiseizure drugs (ASDs). Clinical features of seizures and epileptic syndromes are, however, only part of the puzzle when it comes to understanding the many factors that influence cognitive outcomes in epilepsy. In studies excluding patients with epileptic encephalopathies, there is a poor correlation between cognitive abilities and clinical seizure characteristics or epilepsy syndrome.16 In fact, a variety of other seizure-independent factors, such as home environment (e.g., caregiver anxiety), familial characteristics (e.g., parent intelligence), and the extent of structural brain lesions, have been associated with cognitive phenotype.8,​16 Careful examination of the timing, degree, and longitudinal progression of cognitive deficits with respect to the course of disease in children with epilepsy has also provided insight. A prospective study of children with recent-onset epilepsy showed that baseline cognitive deficits are already present at or near the time of diagnosis.17 Moreover, longitudinal evaluation over a 5- to 6-year period showed that differences in cognitive functioning between children with and without epilepsy persist unchanged over time, neither worsening nor recovering, despite progression of disease and initiation of treatment. The presence of cognitive differences at such an early stage of disease highlights the possibility that both seizures and cognitive deficits arise as a result of a common underlying neurobiological mechanism. For example, both acquired insults (e.g., ­hypoxic ischemic injury, trauma, mesial temporal sclerosis secondary to febrile seizures) and genetic abnormalities (e.g., fragile X syndrome, Rett syndrome, and tuberous sclerosis complex) can lead to the development of epilepsy as well as to significant cognitive dysfunction.

adaptive behavior.20 On a biochemical level, seizures may increase susceptibility to depression by disrupting neurotransmitter systems, especially serotonin and norepinephrine.21 Moreover, the vast majority of patients with epilepsy are taking pharmacologic treatments, some of which alter neurotransmitter concentrations in the brain21 and carry adverse psychotropic effects.18 Conversely, patients with depression and anxiety have an increased risk of developing epilepsy.18 In fact, when the timing of depressive episodes is examined relative to the diagnosis of epilepsy, data show an increase in the rate of depressive episodes preceding the onset of seizures.21 Recent neuroimaging studies of patients with depression have identified both anatomical and functional c­hanges in the hippocampus and other limbic structures involved in epilepsy. Importantly, stress-induced brain inflammation is emerging as a critical factor linking epilepsy with its comorbidities, also raising the possibility that anti-inflammatory treatment might ameliorate both seizures and seizure-related comorbidities.22 As with cognitive deficits, neuropsychiatric comorbidities can precede seizure onset, suggesting that a shared etiology may increase susceptibility to both epilepsy and these comorbid conditions. Seizures themselves are unlikely to be the sole cause of the developmental and neuropsychiatric comorbidities in epilepsy. Rather, the literature suggests a complex relationship, likely involving genetic and environmental influences and highlighting the importance of a better understanding of illness mechanisms. The relationship between seizures and its comorbidities is regulated by multiple determinants, some not yet identified. There is strong evidence that mood disturbances, depression, and psychological distress are reliable predictors of quality of life in patients with epilepsy, independent of seizure frequency and control.2 Moreover, neuropsychiatric and developmental disorders early in life have long-lasting consequences that can impact social success into adulthood. Understanding the pathophysiological link between epilepsy and its associated conditions could have a major impact on the lives of people living with epilepsy.

Psychiatric Comorbidities

„„ Effects of Early-Life Seizures

In addition to cognitive impairment, behavioral and psychiatric comorbidities in patients with epilepsy have been reported extensively (Table 7.1). Patients with epilepsy have higher frequencies of internalizing disorders such as depression, bipolar disorder,18 and anxiety,19 as well as externalizing disorders such as attention deficit hyperactivity disorder.9 Depression is one of the most common of the psychiatric comorbidities in both children and adults, with reports of depressive symptoms and suicidal ideation as high as five-fold higher in patients with epilepsy compared to the general population, or to patients with other chronic diseases.18 The relationship between epilepsy and psychiatric comorbidities has been described as bidirectional.18 On the one hand, patients with epilepsy are at greater risk of developing depression, which may be due in part to social and functional impairments secondary to the seizures. Indeed, parents of children with e ­ pilepsy often express concern about their child’s ability to partake in typical childhood activities, and social factors are likely to contribute to overall psychiatric well-being. This observation is consistent with reports that patients with epilepsy have poor long-term social outcomes in terms of finishing school, ­getting married, and maintaining employment,9 and impaired

Many advances in understanding the neurobiological effects of ­seizures come from laboratory studies using animal models, where investigators are better able to control relevant variables such as age of onset, number and duration of seizures, as well as underlying etiology. Such studies have demonstrated both structural and functional changes in the brain as a consequence of seizures. Of particular interest are findings showing that outcomes vary in an age-dependent manner;23 mature and developing brains differ substantially in their response to seizures. For example, seizure-induced damage is far more extensive in the mature brain, where significant cellular structural alterations, including neuronal death, can be seen.24 In several epilepsy models, adult animals develop cognitive and behavioral abnormalities after seizures, in addition to epilepsy. In contrast, younger animals suffer relatively less observable structural damage, even when seizures are more severe. Despite a normal gross appearance of the brain following seizures early in life, there are morphological and physiological changes, at times subtle, that are observable in neurons and their networks.25 Early-life seizures have been shown to cause synaptic reorganization, alterations in the expression of neurotransmitter receptors, reduced neurogenesis, abnormalities in neuronal pathways, and hyperexcitable

7  Effects of Seizures and Their Comorbidities on the Developing Brain circuits.25 Indeed, interictal discharges in rat pups (even without seizures) were associated with fewer hippocampal neurons and memory impairments when tested as adults.26 Morphological and functional alterations at the cellular level early in life may interfere with normal synaptic and network development, thereby contributing to increased susceptibility to seizures as well as lasting cognitive and behavioral changes. Data from both experimental animal models and clinical studies demonstrate that recurrent early-life seizures are associated with long-term neurocognitive impairments. Children with ­early-life seizures are at greatest risk for cognitive deficits, particularly when seizures are frequent and resistant to treatment.27 Neuroimaging studies allow examination of structural changes in the brain and alterations in the organization of functional networks in children with epilepsy. Such investigations have demonstrated alterations in brain volume,28 and suboptimal structural reorganization such that children with early-life seizures have greater network segregation and reduced integration across networks.29 Although these alterations are already evident at the time of epilepsy diagnosis, it is still unclear when, how, or why specific networks are affected. Aberrant epileptic activity occurring during critical periods of brain maturation may interfere with multiple neurodevelopmental processes, thereby causing alterations in network organization and functional deficits.30

„„ The Impact of Treatment Despite incomplete understanding of the relationship between seizures and its many comorbidities, it is clear that the developing brain is vulnerable to cognitive and behavioral deficits yet also amenable to therapeutic interventions. Pharmacological interventions are the mainstay of treatment of clinical seizures. ASDs work primarily by modulating neuronal activity and synaptic function via a variety of molecular mechanisms. While ASDs may suppress seizure occurrence, the desire for optimal seizure control must be balanced against iatrogenic side effects such as sedation. Due to their sedating action, ASDs contribute to neurodevelopmental deficits in children with epilepsy, most consistently described with the use of barbiturates and benzodiazepines.31 However, some studies have failed to prove an association between seizure remission, ASD use, and cognitive functioning.7 The independent contribution of pharmacotherapy to cognitive disability has been difficult to establish, as drug use strongly correlates with many core seizure characteristics. If medications fail to control seizures, then resective epilepsy surgery can be another viable treatment modality for­

appropriate patients.32,​33 With regard to postoperative cognitive outcomes, the direct effect of epilepsy surgery on cognition is still debated.34,​35 When considering the effects of surgery on eventual cognitive and behavioral status, multiple variables need to be considered, all of which relate to dynamic changes occurring in the developing brain. These factors include the age of onset of epilepsy (earlier or more severe epilepsy is associated with less favorable outcome), the duration of the epilepsy prior to surgery (longer epilepsy portends poorer outcome, due to more seizures and longer time on ASDs), the epilepsy syndrome, the type of surgery (most data is available for hemispherectomy and temporal lobe resection), and the location of the resection. Prior developmental level also plays a role, as children with more severe deficits prior to surgery may have relatively more to gain once the seizure focus is resected. All of these issues are discussed in depth in a recent review.36 Overall, studies have reliably shown improved seizure control, adaptive functioning (e.g., educational level and employment status), psychiatric status, and quality of life, even in cases in which seizure control is not complete.37,​38 If epilepsy surgery is appropriate based on seizure characteristics, families can be counseled that cognitive and behavioral outcome may improve as well. Though these questions are amenable to a reductionist approach using animal models, experimental studies on cognitive outcome after epilepsy surgery are essentially nonexistent.4 Dietary therapies and environmental enrichment also provide beneficial effects on cognition in children with epilepsy and in epilepsy models.39,​40

„„ Conclusion Despite the frequency with which cognitive and psychiatric impairments occur in association with epilepsy, gaps still exist in our understanding of this relationship. Seizures are the manifestation of pathological brain activity, and at the same time, seizures can cause structural and functional abnormalities in neural networks. A multitude of factors, including genetic predisposition, environmental factors, functional insult during neural development, and underlying brain abnormalities, to name a few, are likely to contribute to the occurrence and consequences of epilepsy. The complex clinical phenotype seen in patients with epilepsy is likely a result of the combined effect of a confluence of these risk factors, as well as core seizure characteristics (Fig. 7.1). There are several potential links between seizures and their comorbidities: (1) comorbidities may be direct or indirect consequence of seizures; (2) comorbidities may be due to an entirely separate pathology and have a phenotype that is accentuated by

Fig. 7.1  Interaction between brain development, seizures, and seizure comorbidities. Genetics, environmental factors, and acquired brain insults modulate ongoing brain development. Recurrent seizures (yellow lightning bolts) can alter brain function depending on the stage of development and occurrence during several “critical periods,” facilitating the occurrence of comorbidities. In turn, comorbidities may manifest differently at different developmental stages. Asterisks indicate potential sites of therapeutic intervention, including surgery.

55

56

I  Introduction to Epilepsy in Children the onset of seizures; (3) both seizures and comorbid conditions may share an underlying cause; or (4) the comorbid conditions may be one of many predisposing risk factors for developing

seizures. Considerable progress has been made in understanding the effect of seizures on the developing brain, but many important questions remain for future research.

References 1. Nickels KC, Zaccariello MJ, Hamiwka LD, Wirrell EC. Cognitive and neurodevelopmental comorbidities in paediatric epilepsy. Nat Rev Neurol 2016;12(8):465–476

21. Catena-Dell’Osso M, Caserta A, Baroni S, Nisita C, Marazziti D. The relationship between epilepsy and depression: an update. Curr Med Chem 2013;20(23):2861–2867

2. Gilliam F, Hecimovic H, Sheline Y. Psychiatric comorbidity, health, and function in epilepsy. Epilepsy Behav 2003;4(Suppl 4):S26–S30

22. Mazarati AM, Lewis ML, Pittman QJ. Neurobehavioral comorbidities of epilepsy: Role of inflammation. Epilepsia 2017;58(Suppl 3):48–56

3. Verrotti A, Carrozzino D, Milioni M, Minna M, Fulcheri M. Epilepsy and its main psychiatric comorbidities in adults and children. J Neurol Sci 2014;343(1–2):23–29

23. Sayin U, Hutchinson E, Meyerand ME, Sutula T. Age-dependent long-term structural and functional effects of early-life seizures: evidence for a hippocampal critical period influencing plasticity in adulthood. Neuroscience 2015;288:120–134

4. Brooks-Kayal AR, Bath KG, Berg AT, et al. Issues related to symptomatic and disease-modifying treatments affecting cognitive and neuropsychiatric comorbidities of epilepsy. Epilepsia 2013;54(Suppl 4):44–60 5. Camfield P, Camfield C. The interaction between childhood-onset epilepsy and cognition: long-term follow up. In: Lagae L, ed. Cognition and Behaviour in Childhood Epilepsy. London: Mac Keith Press; 2017:43–53 6. Berg AT, Langfitt JT, Testa FM, et al. Global cognitive function in children with epilepsy: a community-based study. Epilepsia 2008;49(4):608–614 7. Berg AT, Langfitt JT, Testa FM, et al. Residual cognitive effects of uncomplicated idiopathic and cryptogenic epilepsy. Epilepsy Behav 2008;13(4):614–619 8. Dunn DW, Johnson CS, Perkins SM, et al. Academic problems in children with seizures: relationships with neuropsychological functioning and family variables during the 3 years after onset. Epilepsy Behav 2010;19(3):455–461 9. Berg AT, Caplan R, Hesdorffer DC. Psychiatric and neurodevelopmental disorders in childhood-onset epilepsy. Epilepsy Behav 2011;20(3):550–555 10. Howell KB, Harvey AS, Archer JS. Epileptic encephalopathy: use and misuse of a clinically and conceptually important concept. Epilepsia 2016;57(3):343–347 11. Stafstrom CE, Kossoff EM. Epileptic encephalopathy in infants and children. Epilepsy Curr 2016;16(4):273–279 12. Chapman KE, Specchio N, Shinnar S, Holmes GL. Seizing control of epileptic activity can improve outcome. Epilepsia 2015;56(10):1482–1485 13. Jehi L, Wyllie E, Devinsky O. Epileptic encephalopathies: optimizing seizure control and developmental outcome. Epilepsia 2015;56(10):1486–1489 14. Vannest J, Tenney JR, Gelineau-Morel R, Maloney T, Glauser TA. Cognitive and behavioral outcomes in benign childhood epilepsy with centrotemporal spikes. Epilepsy Behav 2015;45:85–91 15. Garcia-Ramos C, Jackson DC, Lin JJ, et al. Cognition and brain development in children with benign epilepsy with centrotemporal spikes. Epilepsia 2015;56(10):1615–1622 16. Hermann BP, Zhao Q, Jackson DC, et al. Cognitive phenotypes in childhood idiopathic epilepsies. Epilepsy Behav 2016;61:269–274

24. Stafstrom CE. Assessing the behavioral and cognitive effects of seizures on the developing brain. Prog Brain Res 2002;135:377–390 25. Holmes GL. Effect of seizures on the developing brain and cognition. Semin Pediatr Neurol 2016;23(2):120–126 26. Khan OI, Zhao Q, Miller F, Holmes GL. Interictal spikes in developing rats cause long-standing cognitive deficits. Neurobiol Dis 2010;39(3):362–371 27. Berg AT, Zelko FA, Levy SR, Testa FM. Age at onset of epilepsy, pharmacoresistance, and cognitive outcomes: a prospective cohort study. Neurology 2012;79(13):1384–1391 28. Zelko FA, Pardoe HR, Blackstone SR, Jackson GD, Berg AT. Regional brain volumes and cognition in childhood epilepsy: does size really matter? Epilepsy Res 2014;108(4):692–700 29. Bonilha L, Tabesh A, Dabbs K, et al. Neurodevelopmental alterations of large-scale structural networks in children with newonset epilepsy. Hum Brain Mapp 2014;35(8):3661–3672 30. Kleen JK, Holmes GL. Disentangling the role of seizures and EEG abnormalities in the pathophysiology of cognitive dysfunction. In: Lagae L, ed. Cognition and Behaviour in Childhood Epilepsy. London: Mac Keith Press; 2017:65–78 31. Kim EH, Ko TS. Cognitive impairment in childhood onset epilepsy: up-to-date information about its causes. Korean J Pediatr 2016;59(4):155–164 32. Engel J Jr, McDermott MP, Wiebe S, et al; Early Randomized Surgical Epilepsy Trial (ERSET) Study Group. Early surgical therapy for drug-resistant temporal lobe epilepsy: a randomized trial. JAMA 2012;307(9):922–930 33. Ryvlin P, Cross JH, Rheims S. Epilepsy surgery in children and adults. Lancet Neurol 2014;13(11):1114–1126 34. D’Argenzio L, Colonnelli MC, Harrison S, et al. Cognitive outcome after extratemporal epilepsy surgery in childhood. Epilepsia 2011;52(11):1966–1972 35. Gallagher A, Jambaqué I, Lassonde M. Cognitive outcome of surgery. Handb Clin Neurol 2013;111:797–802 36. Van Schooneveld MMJ, Braun KPJ, Cross JH. Cognitive and behavioural outcomes after epilepsy surgery in children. In: Lagae L, ed. Cognition and Behaviour in Childhood Epilepsy. London: Mac Keith Press; 2017:172–186

17. Rathouz PJ, Zhao Q, Jones JE, et al. Cognitive development in children with new onset epilepsy. Dev Med Child Neurol 2014;56(7):635–641

37. Alonso NB, Mazetto L, de Araújo Filho GM, Vidal-Dourado M, Yacubian EM, Centeno RS. Psychosocial factors associated with in postsurgical prognosis of temporal lobe epilepsy related to hippocampal sclerosis. Epilepsy Behav 2015;53:66–72

18. Kanner AM. The treatment of depressive disorders in epilepsy: what all neurologists should know. Epilepsia 2013;54(Suppl 1): 3–12

38. Hamid H, Blackmon K, Cong X, et al. Mood, anxiety, and incomplete seizure control affect quality of life after epilepsy surgery. Neurology 2014;82(10):887–894

19. Pham T, Sauro KM, Patten SB, et al. The prevalence of anxiety and associated factors in persons with epilepsy. Epilepsia 2017;58(8):e107–e110 –[published online ahead of print June 9, 2017]

39. Kotloski RJ, Sutula TP. Environmental enrichment: evidence for an unexpected therapeutic influence. Exp Neurol 2015;264:121–126

20. Berg AT, Smith SN, Frobish D, et al. Longitudinal assessment of adaptive behavior in infants and young children with newly diagnosed epilepsy: influences of etiology, syndrome, and seizure control. Pediatrics 2004;114(3):645–650

40. IJff DM, Postulart D, Lambrechts DAJE, et al. Cognitive and behavioral impact of the ketogenic diet in children and adolescents with refractory epilepsy: a randomized controlled trial. Epilepsy Behav 2016;60:153–157

8

  Ethical Considerations in Pediatric Epilepsy Surgery George M. Ibrahim and Mark Bernstein

Summary The path from onset of seizures to epilepsy surgery is difficult to navigate for many patients and their families. From referral, the process onward to presurgical evaluation, operative ­ decision-making, and postoperative management, decisions must be guided by clinical acumen, evidence, and e ­ thical principles. The current chapter presents an introduction to ­ ethical principles that guide the conduct of epilepsy surgery in children. Topics include access to surgery, operative decision-­ making, and informed consent for difficult cases, including Rolandic epilepsy. Emerging topics that are discussed also include the ethical conduct of clinical epilepsy surgery research and translation of surgical innovations as well as epilepsy ­surgery in severely developmentally delayed children. These topics aim to provide an important perspective for clinicians involved in pediatric epilepsy surgery. Keywords:  assent, capacity, consent, childhood epilepsy, ethics, epilepsy surgery, informed consent, surgical innovation, therapeutic misconception

„„ Introduction The path starting from onset of seizures to epilepsy surgery is protracted, and ridden with challenges and difficulties for affected children and their families. Children undergoing presurgical evaluation for treatment of epilepsy comprise a vulnerable patient population due to the medical and psychosocial burden of epilepsy, as well as the hope of seizure freedom afforded by surgical interventions. Ethical challenges arise at each junction along the path toward epilepsy surgery. From access to epilepsy surgery, presurgical evaluations, operative decision-making, and postoperative assessment, clinicians must recognize the various ethical dilemmas that may be encountered. Additional unique considerations are also present for instance when caring for severely developmentally delayed children, who often present with comorbid epilepsy, and in the low-resource setting. The current chapter will review selected ethical topics in the conduct of pediatric epilepsy surgery arising as patients and their families navigate the complex landscape on the path toward surgical treatment. While much of the medical bioethics literature may be directly applied to inform the conduct of epilepsy surgery, it is critical to appreciate that unique ethical considerations also arise. Epilepsy is an incompletely understood disease which interacts

with and indeed modifies an individual’s agency, autonomy, and, under certain circumstances, identity. Neurological conditions, including epilepsy, are often afforded a unique standing on the basis of the importance placed on diseases affecting the brain, the seat of agency, and consciousness. While much can be extrapolated from the basic tenets of medical bioethics— beneficence, nonmaleficence, autonomy and k ­ justice—the burgeoning field of “neuroethics” expands on these in order to attempt to understand and examine these interactions between neurocognitive function, disease, and ethics.

„„ Ethical Issues Surrounding Access to Epilepsy Surgery Illustrative Case A 12-year-old boy with cerebral palsy, severe developmental delay, shunted hydrocephalus, and nonlocalization-related epilepsy continues to have dozens of seizures daily. A vagus nerve stimulator may decrease seizure frequency, but the hospital budget can only accommodate several of these devices per year.

There is greater acceptance of epilepsy surgery among physicians due to increased knowledge of the detrimental effects of seizures on the developing brain.1,​2 Longer duration of uncontrolled epilepsy has been associated with impaired development of cognitively salient neural networks3 and delayed surgical treatments are associated with a lesser likelihood of success. To facilitate the provision of epilepsy surgery, referral guidelines have been published.4 Despite growing awareness of the benefit of epilepsy surgery, the majority of children lack adequate access to epilepsy surgery, posing an ethical challenge to the tenets of beneficence and justice. An international survey of pediatric epilepsy centers showed that the mean duration of epilepsy prior to surgery was 5.7 years, with significantly longer times for older children.5 The issue of equity in access to pediatric epilepsy surgery is not only one of the ongoing advocacy but also an ethical challenge vis-à-vis the tenet of distributive justice.6 A bioethical framework has been described to inform public, institutional, and public policies toward the provision of pediatric epilepsy.6 The basic expectations of this framework are access, protection of the vulnerable, transparency, equity, and societal benefit. A particularly vulnerable patient population with limited access to epilepsy surgery is children with severe developmental

58

I  Introduction to Epilepsy in Children delay and epilepsy. Often, seizure freedom is not anticipated, but palliative surgical procedures may be considered in order to improve their quality of life. These children often suffer from an ethical quagmire that has been described as “double jeopardy.”7,​8 The children’s quality of life is first disadvantaged by their severe illness and subsequently, because of the severity of their illness, lower priority is often given to treatments that can bestow incremental improvements in their daily lives or the care that can be provided to them. With increased awareness among clinicians and the lay public alike of the benefits of epilepsy surgery, more inclusive policies regarding epilepsy surgery in childhood are indicated to facilitate the ethical provision of care. In cases where limited resources are available to treat patients, one bioethical framework that may be applied to prioritize treatment decisions (i.e., priority setting) is the accountability for reasonableness (A4R) framework.9,​10,​11 The four primary expectations of A4R are relevance, publicity, challengeability, and oversight. The personal, institutional, or public health priorities for the surgical treatment of children with epilepsy must be guided by evidence and supported by all stakeholders, effectively communicated, open to challenge through a fair appeals process and with mechanisms in place to continually improve the process. By adopting such an explicit mechanism for the prioritization of epilepsy surgery, held against the standard of reasonableness, more ethical patterns of conduct begin to arise.11

„„ Ethical Issues in the Presurgical Evaluation of Children with Intractable Epilepsy Once a child is referred for presurgical evaluation, there are important ethical considerations that must be taken into consideration. While numerous ethical challenges may arise from the process itself, perhaps the most critical aspect of the presurgical evaluation is the presentation of the findings to patients and parents within the informed consent discussion. The current segment outlines minimal requirements for the ethical conduct of informed consent within the clinical and research settings.

Informed Consent for Epilepsy Surgery Informed consent serves to protect the patient in the doctor– patient relationship, which is nearly universally characterized by greater vulnerability on the part of the patient.12 The minimal requirements of informed consent are: (1) patient capacity; (2) full disclosure; and (3) voluntariness, the lack of undue influence.13 While the definitions of each of these entities varies widely within the literature and often overlap with legal definitions, general principles are universally applicable. Most definitions appeal to the premise of reasonableness, that is, what a reasonable person might expect during the informed consent discussion. A reasonable capable patient must both “understand” and “appreciate” the risk of the procedure.14 Full disclosure, which is often not truly possible,15 may be viewed as what risks the physician believes a reasonable person may

attach significance to when determining whether or not to undergo therapy.16 Finally, the process must respect patient autonomy by being void of coercion. In reality, the informed consent process is bidirectional and dynamic and is modified by treatment factors and patient factors. Treatment factors, for instance how invasive the procedure is and patient factors include how sick the patient is and how much they require this procedure. The discussion surrounding a child who remains in intractable status epilepticus and requires urgent epilepsy surgery is often quite different from those pertaining to healthy children with infrequent or nondebilitating seizures. A “relational” view of informed consent holds that the treatment needs to be related to the patient’s subjective experience with the illness, which modifies the informed consent process.17,​18

Informed Consent in Children The informed consent discussion for procedures performed on children and adolescents requires foresight, an understanding of the family dynamics, and an appreciation of the intellectual abilities of the child. During development, children are egocentric and lack the tools to dissociate subjective experiences from generalized beliefs. Certainly, young children cannot consent to a procedure given the immaturity of their thought process. As a result, in order to protect the welfare of the child, specific jurisdictions have age-defined legal thresholds for childhood decision-making. An increasing wealth of literature suggests that it is a child’s experience and maturity, rather than an arbitrary age threshold that may determine their ability to make independent decisions.19 For instance, a study of children undergoing major surgery demonstrated that adults respected the wishes of children as young as 7 years old.20 As such, numerous authors have advocated the involvement of children throughout the ­decision-making process.19 An important realization is that consent from children cannot be genuine if a clinician accepts it only if it is consistent with what he/she feels is in the child’s best interest.19 Therefore, for children who are genuinely unable to provide consent, assent, a lesser standard of acquiescence to treatment may be obtained.21 While informed consent is largely motivated by the patient’s inalienable right to autonomy, most young children have not— and indeed could not—declare their wishes. As a result, the ethical argument for interventions in young children appeals to their “best interest,” that is, protection of the child’s welfare. This presents a challenge in epilepsy surgery. While several studies have suggested that epilepsy interferes with childhood development,3,​22 the provision of epilepsy surgery in children remains highly heterogeneous, and disagreements arise regarding the role of epilepsy surgery in improving behavioral and cognitive quotients. In cases of life-threatening illnesses, such as acute obstructive hydrocephalus or posterior fossa tumors, the welfare of the child dictates the need for intervention. The decision to proceed with epilepsy surgery is often value-laden and unique to the child’s (and the family’s) experience with epilepsy as well as the risks associated with the intervention. The risk-to-benefit profile of surgical treatment must therefore be presented within the context of the child’s experience with the illness.

8  Ethical Considerations in Pediatric Epilepsy Surgery

Informed Consent for Research and the Therapeutic Misconception It must be recognized that patients undergoing ­experimental therapy comprise a unique patient population with specific ­ethical considerations. The gamut of ethical concerns related to these patients is beyond the scope of the current chapter. An important consideration, however, is the “therapeutic misconception.” This is manifested when patients fail to recognize the competing obligations of the physician as both clinical investigator and primary caregiver and that the intervention being offered has no proven benefit or superiority as of yet.14,​23,​24 Vulnerable patient populations who may hold negative views regarding their heath are at particular risk of therapeutic misconception.25 There are various steps that clinicians can undertake in order to mitigate the manifestation of therapeutic misconception. First, an explicit acknowledgment that the purpose of research participation is to benefit future patients, rather than the current individual is necessary. One report suggests that participants should understand the following five dimensions of research in order to mitigate therapeutic misconception: (1) scientific purpose, to benefit future patients; (2) study procedures, that are not necessary for patient care; (3) uncertainty, which is greater than standard treatments; (4) adherence to protocol, which is more strict than standard treatments; and (5) clinician as investigators, the dual roles of the treating physician.26

„„ Ethical Issues in Operative Decision-making The tenet of epilepsy surgery is that patients accept the risk of intervention for a given likelihood of improvement in seizure control. In the context of adult temporal lobe epilepsy, level I evidence has shown a significant benefit of surgical intervention over continued medical management.27 Children often present with more heterogeneous epilepsy syndromes involving widespread networks. Despite this, it is generally well recognized that the benefit of surgery far exceeds that of ongoing medical treatments.

Surgery in Eloquent Cortex Illustrative Case A 3-year-old girl presents with medically intractable epilepsy characterized by hypermotor seizures. Positron emission tomography reveals an area of hypometabolism involving the primary motor cortex. She undergoes invasive monitoring which confirms colocalization of the seizure onset within the leg motor cortex. The multidisciplinary team discusses the results with the family.

Surgery in eloquent cortex poses a formidable challenge to clinicians from an evaluation, technical, and ethical perspective. Advances in the evaluation of these patients such as functional imaging and progress in techniques including awake craniotomy have facilitated the provision of safer interventions. Such advances are however useful in circumstances where the pathology is near eloquent cortex. As the illustrative case demonstrates, in some cases, the pathology itself is within eloquent

networks. While novel technologies, such as the responsive ­neurostimulation may be beneficial for a subgroup of patients, resective strategies are more likely to bestow seizure freedom.28 Certainly some centers may perceive the risks of resective surgery within eloquent cortex to be too great.6,​29 Although this would mitigate iatrogenic injury to eloquent cortex, it would also deprive a vulnerable patient population of a therapy that could dramatically improve their quality of life. Rather, when approaching patients for surgical treatment involving eloquent cortex, it is effective to establish priorities that contribute to their quality of life, a so-called hierarchy of need satisfaction.30 A life of disabling seizures may be, for some patients, a worse condition than an iatrogenic disability. Prioritizing seizure freedom over neurological function is dependent on this hierarchy and is moderated by how likely the intervention will result in seizure freedom and how significant the deficit is expected to be. For the former, concordant modalities localizing to a specific cortical region or lesional epilepsy syndromes offer greater likelihood of seizure freedom and may be more favorably viewed within the hierarchy of needs satisfaction. For the latter, facial weakness may be more acceptable to patients than deficits expected in the upper or lower extremities. Certainly, surgeons uphold the tenet of “first do no harm.” The definition of harm for patients undergoing resective surgery in eloquent cortex is, however, contextually significant and must incorporate the harm involved in not treating the epilepsy. A principle of “permissible harm” has been coined, whereby an intervention with a significant foreseeable harm is justified if it is an effect or aspect of the greater good.31

Clinical Translation of Surgical Innovations Illustrative Case A 10-year-old boy undergoes invasive monitoring for intractable ­epilepsy. The epileptogenic cortex expresses pathological high frequency oscillations (pHFOs). The area of seizure propagation is larger than the region expressing pHFOs > 200 Hz. The team meets to discuss the resection plan.

As in the remainder of the medical realm, epilepsy surgery has benefitted from advances in surgical techniques and progress in the neurosciences which facilitate greater understanding of the aberrant circuitry leading to seizures. Several of these include interstitial thermal ablation of epileptic foci or HFO-guided resections. While randomized controlled trials are the benchmark to legitimize experimental therapies, in reality most surgical innovations do not undergo such rigorous evaluations, and they may be impractical or impossible for epilepsy surgery.32 The extent to which a surgical innovation requires regulation is directly related to the extent to which it deviates from established therapies.33 This may pose a challenge in epilepsy surgery where the practice is heterogeneous and only minimal guidelines are available to guide the clinician in determining a “standard of care.” Furthermore, it has been reported that academic surgeons often cannot differentiate minor surgical modifications from extensive alterations, particularly when such changes are made incrementally.34 As the field of epilepsy surgery continues to pave new frontiers, so too must the ethical

59

60

I  Introduction to Epilepsy in Children underpinnings of new interventions, and their translation to the operating room be increasingly scrutinized.32

„„ Conclusion The conduct of epilepsy surgery must be governed by evidence, tempered by pragmatism and guided by ethical principles.

Children with epilepsy and their families must navigate a complex labyrinth of appointments, investigations, and decisions on the path toward epilepsy surgery. As such, clinicians have a fiduciary obligation to provide sound clinical judgment and ethical care to affected families. The fundamental principles pertaining to the ethical conduct of epilepsy surgery described in this chapter should inform the evaluation and treatment of children with epilepsy.

References 1. Holmes GL, Ben-Ari Y. Seizures in the developing brain: perhaps not so benign after all. Neuron 1998;21(6):1231–1234

19. Alderson P, Sutcliffe K, Curtis K. Children’s competence to consent to medical treatment. Hastings Cent Rep 2006;36(6):25–34

2. Meador KJ. Cognitive outcomes and predictive factors in epilepsy. Neurology 2002;58(8, Suppl 5):S21–S26

20. Alderson P. Consent to Surgery. Buckingham: Open University Press; 1993

3. Ibrahim GM, Morgan BR, Lee W, et al. Impaired development of intrinsic connectivity networks in children with medically intractable localization-related epilepsy. Hum Brain Mapp 2014;35(11):5686–5700

21. Weithorn LA, Campbell SB. The competency of children and adolescents to make informed treatment decisions. Child Dev 1982;53(6):1589–1598

4. Cross JH, Jayakar P, Nordli D, et al; International League against Epilepsy, Subcommission for Paediatric Epilepsy Surgery. Commissions of Neurosurgery and Paediatrics. Proposed criteria for referral and evaluation of children for epilepsy surgery: recommendations of the Subcommission for Pediatric Epilepsy Surgery. Epilepsia 2006;47(6):952–959 5. Harvey AS, Cross JH, Shinnar S, Mathern GW, Taskforce IPESS; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epilepsia 2008;49(1):146–155 6. Ibrahim GM, Barry BW, Fallah A, et al. Inequities in access to pediatric epilepsy surgery: a bioethical framework. Neurosurg Focus 2012;32(3):E2 7. Harris J. Life: quality, value and justice. Health Policy 1988; 10(3):259–266 8. Harris J. QALYfying 1987;13(3):117–123

the

value

of

life.

J

Med

Ethics

9. Daniels N, Sabin JE. Accountability for reasonableness: an update. BMJ 2008;337:a1850 10. Hasman A, Holm S. Accountability for reasonableness: opening the black box of process. Health Care Anal 2005;13(4):261–273 11. Ibrahim GM, Tymianski M, Bernstein M. Priority setting in neurosurgery as exemplified by an everyday challenge. Can J Neurol Sci 2013;40(3):378–383 12. Surbone A. Telling the truth to patients with cancer: what is the truth? Lancet Oncol 2006;7(11):944–950 13. Etchells E, Sharpe G, Walsh P, Williams JR, Singer PA. Bioethics for clinicians: 1. Consent. CMAJ 1996;155(2):177–180 14. Appelbaum PS, Grisso T. Assessing patients’ capacities to consent to treatment. N Engl J Med 1988;319(25):1635–1638 15. Bernstein M. Fully informed consent is impossible in surgical clinical trials. Can J Surg 2005;48(4):271–272 16. Canterbury v Spence C. 464 F (2nd). 1972:772 17. Donchin A. Understanding autonomy relationally: toward a reconfiguration of bioethical principles. J Med Philos 2001;26 (4):365–386 18. Feminist Health Care Ethics Research Network, Sherwin S, et al. A relational approach to autonomy in health-care. In: The Politics of Women’s Health: Exploring Agency and Autonomy. Philadelphia, PA: Temple University Press; 1988:19–44

22. Ibrahim GM, Cassel D, Morgan BR, et al. Resilience of developing brain networks to interictal epileptiform discharges is associated with cognitive outcome. Brain 2014;137(Pt 10):2690–2702 23. Appelbaum PS, Lidz CW, Grisso T. Therapeutic misconception in clinical research: frequency and risk factors. IRB 2004;26 (2):1–8 24. Lidz CW, Appelbaum PS, Grisso T, Renaud M. Therapeutic misconception and the appreciation of risks in clinical trials. Soc Sci Med 2004;58(9):1689–1697 25. Goebel S, von Harscher M, Mehdorn HM. Comorbid mental disorders and psychosocial distress in patients with brain tumours and their spouses in the early treatment phase. Support Care Cancer 2011;19(11):1797–1805 26. Henderson GE, Churchill LR, Davis AM, et al. Clinical trials and medical care: defining the therapeutic misconception. PLoS Med 2007;4(11):e324 27. Wiebe S, Blume WT, Girvin JP, Eliasziw M; Effectiveness and Efficiency of Surgery for Temporal Lobe Epilepsy Study Group. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med 2001;345(5):311–318 28. Morrell MJ; RNS System in Epilepsy Study Group. Responsive cortical stimulation for the treatment of medically intractable partial epilepsy. Neurology 2011;77(13):1295–1304 29. Erba G, Moja L, Beghi E, Messina P, Pupillo E. Barriers toward ­pilepsy surgery. A survey among practicing neurologists. Epilepsia 2012;53(1):35–43 30. Tomasini F. Exploring ethical justification for self-demand amputation. Ethics Med 2006;22(2):99–115 31. Kamm FM. Intricate Ethics: Rights, Responsibilities, and Permissible Harm. New York, NY: Oxford University Press; 2007 32. Ibrahim GM, Fallah A, Snead OC III, Drake JM, Rutka JT, Bernstein M. The use of high frequency oscillations to guide neocortical resections in children with medically-intractable epilepsy: how do we ethically apply surgical innovations to patient care? Seizure 2012;21(10):743–747 33. Bernstein M, Bampoe J. Surgical innovation or surgical evolution: an ethical and practical guide to handling novel neurosurgical procedures. J Neurosurg 2004;100(1):2–7 34. Reitsma AM, Moreno JD. Ethical regulations for innovative surgery: the last frontier? J Am Coll Surg 2002;194(6): 792–801

9

  Infantile and Childhood-Onset Catastrophic Epilepsy Syndromes Hirokazu Oguni

Summary The term “catastrophic childhood epilepsies” was first introduced by North American pioneers who began performing epilepsy surgery on children with refractory epilepsies and running catastrophic clinical courses. The incidence of these epilepsies is the highest during infancy and early childhood and they have diverse etiologies. These patients have pharmacoresistant disabling seizures and electroclinically manifest with epileptic encephalopathy. Of these, early myoclonic encephalopathy (EME), Ohtahara syndrome (OS), and infantile spasms (IS) are well-known catastrophic epilepsy syndromes in infancy. New epileptic syndromes with an onset during early infancy and running catastrophic clinical courses have recently been identified and include KCNQ2 encephalopathy, epilepsy of infancy with migrating focal seizures, and Dravet syndrome (DS), which is mostly caused by mutations in cerebral ion channel genes. These syndromes may be diagnosed by genetic analyses early in the clinical course before electroclinical features sufficiently mature for a syndromic diagnosis. In early childhood, a number of catastrophic epilepsy syndromes exist including ­epilepsy with myoclonic–atonic seizures (EMAS), Lennox–Gastaut ­syndrome (LGS), epilepsy with continuous spike-wave during slow sleep (CSWS), Landau–Kleffner syndrome (LKS), and ­Rasmussen’s syndrome (RS). Long-term video-electroencephalography (EEG) monitoring, neuroimaging studies, metabolic screening tests, and genetic analyses, according to electroclinical information, are performed in an attempt to diagnose these catastrophic epilepsy syndromes in the early stages. Although treatment strategies for most of these epilepsies are currently limited, an etiology-specific treatment (such as vitamin B6 and folinic acid), adrenocorticotropic hormone (ACTH) or high-dose steroids, ketogenic diet therapy, and surgery, if it is indicated, need to be considered in addition to antiepileptic drugs (AEDs) for altering catastrophic clinical courses. Keywords:  catastrophic epilepsies, epileptic syndrome, ­epileptic encephalopathy, structural and metabolic etiology, pharmacoresistant, neuroimaging, long-term video-EEG monitoring

„„ Introduction The term “catastrophic childhood epilepsies” was first introduced by North American pioneers who began performing epilepsy surgery on children whose seizures started early in life and were so frequent and intense that their psychomotor development and daily life were significantly impaired.1,​2,​3 These

­pilepsies are the most prevalent during infancy and early e childhood and are often caused by serious underlying disorders. The etiology of epilepsy in the first 3 years of life has recently been elucidated in more than 50% of cases.4,​5,​6 A structural–metabolic etiology was the most frequent, accounting for 35 to 54%, followed by a genetic etiology, accounting for 7 to 22%, with the remaining cases being categorized as unknown. However, the incidence of an unknown etiology has been decreasing due to advances in neuroimaging devices, metabolic assays, and genetic analyses. The incidence of epilepsy with the structural– metabolic etiology, potentially causing catastrophic epilepsies, was previously reported to be the highest in the first 6 months of life.4 Therefore, congenital metabolic disorders are important causes, that is, vitamin B6 deficiency, GABA metabolic disorders, nonketotic hyperglycinemia or glucose transporter 1 deficiency syndrome, peroxisomal disorders, and lysosomal diseases.7,​8 Since some of these disorders are treatable, an early diagnosis and treatment are inevitable.9 Advances in neuroimaging modalities, particularly high-­ resolution MRI, has contributed to a better understanding of the malformations of cortical development (MCDs).10 In the treatment of catastrophic infantile and childhood epilepsies caused by MCDs, clinical phenotypes, MRI structural characteristics, neuropathology, and underlying genetic mechanisms are all comprehensively considered with seizure semiology and EEG findings.11 Regarding the genetic etiology, the introduction of next-generation sequencing has accelerated the identification of new epilepsy-causing genes among catastrophic epilepsies beginning in infancy.12 Therefore, the search for the underlying etiology is as important as reaching an electroclinical diagnosis of epilepsy or ­epilepsy syndrome because the outcomes of catastrophic epilepsies and their therapeutic options largely depend on the etiology.

„„ Catastrophic Epilepsies with an Onset During Infancy and Childhood A number of International League Against Epilepsy (ILAE)-­ recognized epilepsy syndromes have been reported to potentially cause catastrophic epilepsies. The incidence rate of these syndromes varies, with IS and LGS accounting for nearly 50% ­ of cases in a hospital-based survey (Fig. 9.1).3 This chapter will overview the diagnosis and treatment of infantile and childhoodonset catastrophic epilepsy syndromes according to the onset age of epilepsy.

62

I  Introduction to Epilepsy in Children

Fig. 9.1  Syndromic classification of catastrophic epilepsies aged less than 6 years. The findings of a multi-institutional study that recruited 314 children with catastrophic epilepsies aged less than 6 years were shown. The most frequent epileptic syndrome was infantile spasms, accounting for 36%, followed by unclassified epilepsy at 21%, neocortical focal epilepsies at 19%, Lennox–Gastaut syndrome at 12% and others. EMAS, epilepsy with myoclonic–atonic seizures; OS, Ohtahara syndrome; EIMFS, epilepsy of infancy with migrating focal seizures. (Reproduced with permission from Oguni et al 2013.3)

Neonatal Period (~1 Month of Age) Most neonatal convulsions are acute seizures caused by perinatal central nervous system (CNS) insults due to hypoxic–ischemic injuries, hypoglycemia, or CNS infections, and do not always progress to chronic epilepsy. Although the incidence of neonatal-onset epilepsy is markedly lower than that of acute seizures, neonates with congenital metabolic disorders or MCDs are more likely to develop chronic epileptic seizures.13 Among ILAE-recognized epilepsy syndromes, early EME, OS, and epilepsy of infancy with migrating focal seizures (EIMFS) initially manifest in this age period.14,​15,​16 Thus, the new concept of “neonatal-onset epileptic encephalopathies,” covering these three epileptic syndromes and recently identified KCNQ2 encephalopathy caused by a de novo KCNQ2 mutation, has been proposed because of significant overlap in their clinical phenotypes, EEG findings, and responsible genes.17,​18,​19

KCNQ2 Encephalopathy KCNQ2 mutations cause not only benign familial neonatal convulsions, but also severe epileptic encephalopathy starting exclusively in the first week of life.17,​18,​19 This is characterized by moderate to severe developmental delays, tonic seizures with postural and autonomic signs,17 and suppression-burst EEG patterns, which shares electroclinical features with OS. However, responses to carbamazepine (CBZ) or other sodium channel blockers were shown to be good and the evolution to IS was rare.17 In contrast to good seizure control, developmental outcomes are generally poor.

Early Myoclonic Encephalopathy The onset age of early EME is always within the first month of life. Its predominant electroclinical features are a fragmentary or erratic myoclonus and suppression-burst EEG pattern.15,​20 Brain MRI findings are not specific. EME is most commonly caused by congenital metabolic disorders including nonke-

totic hyperglycinemia. Mutations of SLC25A22 and GABAAG2 were recently identified in some patients with EME.21,​22 Longterm outcomes are extremely poor because 50% of patients die before 2 years of age.20

Infant Period (Aged 1–24 Months) In a recent survey, 40% or less of epilepsies with an onset younger than 1 to 2 years of age were a specific form of epileptic encephalopathy,4,​6 among which IS was the most common form. These epilepsies are extremely pharmacoresistant, causing poor neurological and intellectual outcomes. The following ILAE-recognized epilepsy syndromes are discussed.

Ohtahara Syndrome The onset age of OS is between the neonatal period and up to 3 months of age. OS is characterized by epileptic spasms (ES) in clusters and a suppression-burst EEG pattern in awake and sleep EEG.20,​23 The structural and metabolic etiology was previously considered to be an important cause of OS. However, the genetic etiology has recently received increasing attention and pathogenic de novo mutations of, for example, STXBP1, SCN2A, and KCNQ2 which were found to be responsible for more than one third of OS.18,​24,​25 Regarding its treatment, OS is not only resistant to AEDs, but also ACTH and high-dose steroid therapy.20 There have been a few exceptional cases of OS caused by hemimegalencephaly that responded to hemispherotomy.26 OS frequently evolves to IS after 3 months of age, at which point ACTH is expected to exert some effect. Long-term seizure and intellectual outcomes are poor.

Epilepsy of Infancy with Migrating Focal Seizures EIMFS is characterized by clusters of focal motor seizures or focal to bilateral tonic–clonic seizures (FBTCS) beginning as early as the first week of life to 6 months of age.27 Each seizure lasts for 1 to 4 minutes and occurs successively, with frequencies sometimes escalating to one hundred times per day (Video 9.1).27,​28 Described as migrating focal seizures, EEG seizure foci shift from one seizure to another. Psychomotor development

9  Infantile and Childhood-Onset Catastrophic Epilepsy Syndromes is arrested and markedly delayed with axial hypotonia. Brain MRI findings may show delayed myelination with white matter hyperintensity.28 Long-term EEG monitoring contributes to an electroclinical diagnosis, which may disclose either shifting areas of an ictal onset between hemispheres or overlapping seizures with different areas of an ictal onset in differing hemispheres.28 A genetic diagnosis is useful because more than 50% of cases carry the mutation of KCNT1, and, less frequently, those of SCN1A, PLCB1, SCN2A, and SCN8A.28,​29 Quinidine therapy, which acts on potassium channels, has been tried with or without success to improve seizures in cases of KCNT1 mutations.30,​31 In other triVideo 9.1 Epilepsy of infancy with migrating focal seizures due to KCNT1 mutation. (This video is provided courtesy of Hirokazu Oguni.) ht t ps://www.thieme.de/de/q.ht m?p=opn/ tp/255910102/9781626238176_c009_v001&t=video

als, potassium bromide (KBr), stiripentol (STP) with clonazepam (CZP), and levetiracetam (LEV) were found to be effective.32,​33

Infantile Spasms The triad of IS consists of ES in clusters, hypsarrhythmia in interictal EEG, and the arrest or deterioration of psychomotor development.34 It is most prevalent among the known epileptic syndromes during infancy, accounting for 20 to 40% of all infantile epilepsies.4,​6 The onset age is between 3 and 11 months and rarely more than 2 years of age. ES are characterized by brief contractions of the flexor and extensor muscles of the trunk, neck, and extremities, repeating every 8 to 15 seconds and lasting for several minutes. The interval and intensity of ES become longer and milder, respectively, with ES repeating in one cluster. They occur most frequently on awakening.35,​36 Some patients with IS constantly show asymmetry or asynchrony in ES semiology (Fig. 9.2a,b).37 Other patients may have focal seizures preceding ES clusters (focal

Fig. 9.2  Symptomatic focal epilepsy due to focal cortical dysplasia. A girl of age 10 years and 7 months had developed epileptic spasms (ES) in clusters at 2 weeks of age. An ictal video-polygraph at 2 months of age showed asymmetrical ES corresponding to diffuse irregular polyspike-wave discharges lateralized in the left hemisphere (a). ES were controlled by zonisamide. She achieved good psychomotor development until 1 year and 4 months of age, when she developed recurrent focal seizures with a postural sign arising from the right frontal region (b). An axial T2-weighted MRI image demonstrated no lesions at 2 months of age, but revealed a slightly high signal intensity lesion in the left frontal lobe at 1 year and 0 months of age, which became apparent at 1 year 6 months of age (c), surrounded by a (white dotted lines). She successfully underwent left frontal lobectomy at 1 year and 8 months of age. The pathological specimen revealed cortical dysplasia type IIB.

63

64

I  Introduction to Epilepsy in Children seizures evolving to IS).38 In these cases, we need to carefully evaluate the localized structural etiology by serial MRI imaging, which at times permits the detection of subcortical white matter abnormalities after 1 to 2 years of age (Fig. 9.2c).39  A hypsarrhythmic EEG pattern is another unique characteristic, displaying very high voltage, irregular, asynchronous slow waves overriding multifocal independent epileptiform discharges.40 However, the constant asymmetry of hypsarrhythmia over both hemispheres may indicate a unilateral structural etiology and require a detailed neuroimaging study (Fig. 9.2).41 Regarding its etiology, approximately 80% of cases are classified as symptomatic IS caused by heterogeneous disorders, including perinatal hypoxic–ischemic encephalopathy, periventricular leukomalacia, cerebral hemorrhage, chromosomal abnormalities, and congenital multiple anomaly syndromes. However, it was unknown in more than 50% of cases.42 Genetic testing has increasingly identified the pathogenic de novo mutations of genes (including ARX, STK9 /CDKL5, CASK, ALG13, STXBP1, PNPO, and ADSL) in symptomatic IS with an unknown etiology.12,​43 A neuroimaging study is important not only for investigating the etiology but also for candidates for epilepsy surgery. Focal cortical dysplasia and hemimegalencephaly are important structural etiologies of IS.39,​44,​45 Regarding the treatment of IS, ACTH, vigabatrin, or their ­combination is currently the first choice in developed ­countries.36,​46,​47,​48 Other AEDs are only of limited value, although high-dose vitamin B6, valproic acid (VPA), topiramate (TPM), zonisamide, and benzodiazepines are often tried.36,​49 ­Ketogenic diet therapy using a milk formulation is worthwhile even if ACTH fails to control IS.50 As for surgical treatments, if it is indicated, total or subtotal hemispherectomy has achieved ­ ­better surgical outcomes than limited focal resection in most surgical case series.44,​51,​52 Regarding short-term seizure outcomes, 50 to 80% of cases were free of IS by ACTH or other medications.46,​47,​48 However, 30 to 40% of the remaining cases later evolved to LGS.14 In the long-term prognosis, nearly 50% of all cases with IS eventually developed some form of epileptic seizure and more than 80% of cases remained mentally handicapped.36

Dravet Syndrome In DS, seizures begin between 2 and 10 months of age in the form of focal, unilateral, or generalized clonic or tonic–clonic seizures in otherwise developmentally normal infants.53,​54 Focal or unilateral seizures may alternate the side of onset in one seizure after another.55 They are extremely sensitive to an elevated body temperature regardless of the etiology (infection and hot bathing) and easily culminate into status epilepticus.53 Interictal EEG findings appear to be normal until 1 year of age.54,​55 Between 1 and 4 years of age, focal seizures with motor (head deviation, stiffness or jerks of one limb, hypomotor seizures) and autonomic symptoms gradually replace generalized tonic–clonic seizures (GTCS). In addition, myoclonic and atypical absence seizures (with eyelid myoclonia and a 3-Hz diffuse spike-wave) at times appear up to one hundred times a day with or without strong photosensitivity.53,​56 EEG shows diffuse or multifocal epileptiform EEG abnormalities in combination with a slow background activity. After 5 years of age, myoclonic and atypical absence seizures (also photosensitivity) gradually disappear

and only focal and generalized convulsive seizures sensitive to an elevated body temperature persist to adulthood.53,​54 Psychomotor development stagnates after 1 year of age, resulting in delays in walking and the lack of expressive word acquisition.55 The clinical diagnosis of DS is not as difficult to reach once the aforementioned clinical characteristics are recognized, among which special attention should be placed on seizures extremely sensitive to an elevated body temperature. Genetic confirmation, in which SCN1A mutations are positive in approximately 80% of DS patients,57 can be reliably reached. The treatment strategy for DS is VPA, clobazam (CLB), STP, KBr, TPM, and a ketogenic diet in various combinations and the avoidance of phenytoin, CBZ, and lamotrigine (LTG).53,​54,​58 Although the combination of STP, VPA, and CLB is the current mainstay of treatment,59 reductions in frequent and disabling seizures are beyond satisfactory. Seizure and intellectual prognoses are generally poor because of the persistence of refractory seizures.54 Morbidity due to sudden unexpected death and acute encephalopathy was reported be peak at 1 to 4 years and 7 years of age, respectively.60

PCDH19-Related Epilepsy PCDH19-related epilepsy was initially reported as epilepsy limited to females, and mental retardation was reported in familial case series61 and caused by the mutation of PCDH19 located in Xq22.1.62 The same de novo mutations were identified among female patients with DS lacking SCN1A mutations.63 Therefore, electroclinical features are shared with those of DS. The onset of epilepsy is younger than 2 years of age (average 8–11 months).64,​65,​66 Otherwise-healthy female infants develop clusters of brief focal seizures (mostly with affective symptoms) or FBTCS lasting for a number of days to 1 week, followed by a second cluster after a seizure-free period of approximately 6 months. The subsequent clusters of seizures continue with age with a monthly or yearly frequency.65,​67 They are often triggered by fever. KBr, CLB, TPM, VPA, and CZP were shown to be effective as chronic treatments.64,​66 As seizures continue, cognitive impairments and autistic behavior gradually progress in 70 to 80% and 30 to 40% of patients, respectively.65,​66 Although epileptic seizures are reasonably controlled after adolescence, neurobehavioral issues remain significant.

Childhood Period (Aged 24+ Months) Among ILAE-recognized epilepsy syndromes, EMAS, LGS, epilepsy with CSWS, and LKS, and RS develop in this age period and run a catastrophic clinical course. In contrast to the catastrophic epilepsies with an onset during the neonatal and infant periods, metabolic and genetic analyses are less helpful, while detailed neuroimaging and long-term video-EEG studies are more useful for reaching a syndromic diagnosis.

Epilepsy with Myoclonic–Atonic Seizures EMAS was once termed epilepsy with myoclonic-astatic seizures or Doose syndrome.68 Seizure onset is between 2 and 5 years of age in a previously healthy child (boys > girls).69 Patients start to have recurrent GTCS and, after a median of 1 month,

9  Infantile and Childhood-Onset Catastrophic Epilepsy Syndromes start having frequent drop attacks. Drop attacks are caused either by myoclonic-flexor or atonic seizures with or without preceding minor myoclonia.70 The former manifests with sudden propulsion of the upper trunk forward, at times hitting the face on the desk, while the latter manifests with the sudden collapse of the body forward, backward, or straight downward onto the buttocks.71 Patients instantaneously recover without postictal confusion. These seizures are refractory to AEDs and may increase in frequency up to 100 times per day. Patients have multiple injuries, particularly on their faces. Approximately 50% patients also have atypical absence seizures. Patients with unfavorable outcomes have nocturnal generalized tonic vibrating seizures in the middle of or later in the clinical course.69 Interictal EEG shows the slowing of background activity, often with centroparietal predominance and frequent diffuse slow spike-wave or polyspike-wave discharges at 1 to 3 Hz without focal or lateralized features. Drop attacks always correspond to diffuse high amplitude spike-wave or polyspike-wave discharges, which may be used to distinguish those caused by ES in LGS.72 Brain CT and MRI are generally normal. Regarding seizure outcomes, myoclonic/atonic drop attacks in most cases disappear within 1 to 3 years despite initial resistance; however, GTCS or tonic seizures may continue.73 The most effective treatment for drop attacks appears to be a ketogenic diet, followed by ACTH/steroids and ethosuximide (ESM).73,​74 EMAS may be classified into favorable, intermediate, and unfavorable forms according to ultimate seizure outcomes and favorable outcomes account for two-thirds of cases.69

Lennox–Gastaut Syndrome LGS was originally characterized by: (1) generalized tonic seizures (GTS) and atypical absence seizures (Video 9.2); (2) pronounced mental retardation; and (3) interictal EEG records showing pseudo-rhythmical (1.5–2.5 c/s) diffuse slow spikewave discharges. Astatic or drop attacks, and generalized fast rhythms at approximately 10 Hz during sleep EEG were added to these criteria.55,​75 The onset age of the first seizure in LGS ranges between 1 and 8 years, depending on the etiology. One-third of patients with LGS evolved from IS.14 It takes time to show the full-blown electroclinical features of LGS described in the criteria.75 In the early form of LGS, or transition from IS to LGS, patients manifest with frequent ES with or without periodic clusters and a diffuse disorganized slow spike-wave with multifocal accentuation (Fig. 9.3). In a differential diagnosis, atypical benign partial epilepsy in childhood and EMAS are two important epilepsy syndromes to distinguish from LGS because both syndromes share some of the electroclinical features of LGS.75,​76,​77 The former was previously Video 9.2 Lennox-Gastaut Syndrome and generalized tonic seizures. (This video is provided courtesy of Hirokazu Oguni.) ht t ps://www.thieme.de/de/q.ht m?p=opn/ tp/255910102/9781626238176_c009_v002&t=video

termed pseudo-LGS or LGS borderland,78 which is characterized by the combinations of focal motor seizures and epileptic negative myoclonus associated with CSWS, and more favorable ­prognosis.77 The latter needs to be distinguished from cryptogenic LGS, which manifests with drop attacks caused by axial spasms (a special form of ES) and an organized diffuse slow spike-wave developing in previously healthy children.76 A differential diagnosis sometimes requires long-term video-EEG monitoring.72 Long-term seizure and intellectual prognoses are generally poor.55 VPA, LTG, TPM, rufinamide, and CLB in combination are choices for AEDs.75,​79,​80 Ketogenic diet therapy is recommended early in the clinical course if seizures are pharmacoresistant (Fig. 9.3).81,​82 Corpus callosotomy is indicated when drop attacks incapacitate daily activities.55,​75 Vagal nerve stimulation is also worthwhile before or at the same time as corpus callosotomy.55,​75,​83,​84

Continuous Spike-Wave During Slow Sleep and Landau–Kleffner Syndrome Epilepsy with CSWS is an age-related, self-limiting disorder characterized by epilepsy with different seizure types, global or selective neuropsychological regression, motor impairment, and the typical EEG pattern of continuous epileptiform activity occupying more than 85% of non-rapid eye movement sleep.85,​86,​87 In LKS, CSWS is considered to affect the auditory- and ­language-related perisylvian cortex, giving rise to verbal ­auditory agnosia (word deafness) followed by language ­ regression.88,​ 89,​90 In contrast, acquired epileptiform opercular syndrome causes dysarthria and dysphasia because CSWS presumably ­disturbs the function of the operculum.91 CSWS was formerly or alternatively termed as electrical status epilepticus during slow sleep, which may potentially impair the cerebral functions of the corresponding cerebral regions.85,​86 The onset of first seizures varies from 2 to 12 years with a peak at approximately 4 to 5 years.55 They are frequently nocturnal focal motor seizures and FBTCS, at times lasting more than 30 minutes. The frequencies of attacks vary in cases and also sometimes in rare cases of LKS.55 In other cases, epileptic negative myoclonus, atonic seizures, or atypical absence seizures develop during the period of CSWS.77,​85 The precedent neurological abnormality as well as neuroradiological abnormalities including polymicrogyria, early thalamic lesions, and hydrocephalus are found in approximately one-third to one-half of cases of CSWS.92,​93 Interictal EEG findings initially show focal or multifocal spike-waves in the centrotemporal, frontopolar, and occipital regions, which later become diffuse and continuous during sleep, satisfying the EEG criteria of CSWS between the ages of 4 and 14 years.85 The degree of the spike-wave index defining CSWS varies among investigators, ranging from 50 to 85% of non-rapid eye movement sleep.86,​94 Neurobehavioral and neurocognitive regression develops with the persistence of CSWS.85,​93 Thus, long-term video-EEG monitoring is recommended for patients with recent-onset neurobehavioral regression and sleep EEG suggesting CSWS. The treatment for epilepsy with CSWS and also LKS includes AEDs, high-dose corticosteroids/ACTH, a ketogenic diet, intravenous immunoglobulin (IVIG), and epilepsy surgery.85,​87,​95 VPA, ESM, benzodiazepines, sulthiame, LEV, and LTG in various

65

66

I  Introduction to Epilepsy in Children

Fig. 9.3  Lennox–Gastaut syndrome (LGS) evolved from infantile spasms. A 13-year-old girl developed epileptic spasms in clusters at 3 months of age, when a diagnosis of infantile spasms (IS) due to tuberous sclerosis was made by brain CT and later by MRI (a, fluid attenuation inversion recovery [FLAIR] image). Thereafter, IS gradually evolved to LGS, characterized by frequent generalized tonic seizures (b), flexor spasms (b), and diffuse slow spike-wave EEG discharges (c). Ketogenic diet (KD) therapy introduced at 2 years and 3 months of age markedly improved her seizures and EEG (d). She has since remained seizure free.

c­ ombinations are the choice of AEDs.93,​95 High-dose corticosteroids/ACTH may control seizures and CSWS, but need to continue for months unless patients relapse. Regarding epilepsy surgery with multiple subpial transection (MST) for patients with LKS, there is currently insufficient evidence to recommend MST over medical therapy.96,​97 Long-term seizure and CSWS outcomes are excellent because they are self-limited, disappearing until the adolescent period in most of cases.98,​99 However, neurobehavioral and cognitive prognoses (language issues in the case of LKS) are not as good as those of seizures and CSWS.90,​93 The varying degree of residual deficits remains at 50 to 90% of patients.93,​100

Rasmussen’s Syndrome RS is characterized by intractable focal epilepsy associated with epilepsia partialis continua (EPC) and slowly progressive neurological deficits affecting one side of the body due to chronic localized encephalitis.101,​102 Onset ages are between 1 and 14 years with a peak at approximately 6 years.103 RS affects healthy developing children and has no gender difference. RS patients start to have focal motor or sensori-motor seizures, complex focal seizures, or FBTCS, which gradually

increase in frequency despite AED therapy.101 EPC commonly develops within 3 years of the onset, and involves the fingers, arms, legs, or mouth. EPC was found in 56 to 100% of RS patients and the absence of EPC cannot exclude the diagnosis of RS.103,​104 Interictal EEG findings are characterized by continuous polymorphous delta activity predominantly over the centroparietal regions with or without epileptiform discharges in the affected hemisphere. Brain MRI initially shows normal findings or unilateral atrophy in the insular and peri-insular regions.102,​ 105 ­Progressive diffuse atrophy as well as increasing high-signal lesions are then detected in the affected hemisphere. Interictal single-photon emission CT and fluorodeoxyglucose positron emission tomography also show decreased perfusion and glucose uptake in the same hemisphere, respectively.102 Although RS is regarded as an immune-mediated encephalitis and epilepsy, there have been no specific autoantibodies or immune-­ mediated products contributing to an early RS diagnosis.105 The clinical diagnosis of RS can be reached based on characteristic clinical features and progressive brain MRI changes, and rarely needs a pathological confirmation.102,​105 Regarding the treatment of RS, a trial on various AEDs is of limited value.103 In cases of status epilepticus or the

9  Infantile and Childhood-Onset Catastrophic Epilepsy Syndromes ­ xacerbation of seizures, high-dose steroid therapy or highe dose IVIG administration may alleviate the condition.102,​106 The identification of cell-mediated immune processes in RS, tacrolimus, or other agents that suppress these mechanisms has been under trial.106,​107 However, these immunomodulatory treat-

ments seem to slow rather than halt disease progression without changing the eventual outcome.105 Regarding surgical treatments, functional hemispherotomy is the final and best option, but needs to be cautiously considered when RS affects the dominant hemisphere.103,​105

References 1. Wyllie E. Surgery for catastrophic localization-related epilepsy in infants. Epilepsia 1996;37(Suppl 1):S22–S25 2. Shields WD. Catastrophic epilepsy in childhood. Epilepsia 2000;41(Suppl 2):S2–S6 3. Oguni H, Otsuki T, Kobayashi K, et al. Clinical analysis of catastrophic epilepsy in infancy and early childhood: results of the Far-East Asia Catastrophic Epilepsy (FACE) study group. Brain Dev 2013;35(8):786–792 4. Gaily E, Lommi M, Lapatto R, Lehesjoki AE. Incidence and outcome of epilepsy syndromes with onset in the first year of life: a retrospective population-based study. Epilepsia 2016;57(10):1594–1601 5. Vignoli A, Peron A, Turner K, et al. Long-term outcome of epilepsy with onset in the first three years of life: findings from a large cohort of patients. Eur J Paediatr Neurol 2016;20(4):566–572 6. Eltze CM, Chong WK, Cox T, et al. A population-based study of newly diagnosed epilepsy in infants. Epilepsia 2013;54(3):437–445 7. Rahman S, Footitt EJ, Varadkar S, Clayton PT. Inborn errors of metabolism causing epilepsy. Dev Med Child Neurol 2013;55(1):23–36 8. Wolf NI, García-Cazorla A, Hoffmann GF. Epilepsy and inborn errors of metabolism in children. J Inherit Metab Dis 2009;32(5):609–617 9. Pearl PL. Inherited metabolic epilepsies. New York, NY: Demos Medical Publishing; 2012:1–13 10. Barkovich AJ, Guerrini R, Kuzniecky RI, Jackson GD, Dobyns WB. A developmental and genetic classification for malformations of cortical development: update 2012. Brain 2012;135 (Pt 5):1348–1369 11. Barkovich AJ, Dobyns WB, Guerrini R. Malformations of cortical development and epilepsy. Cold Spring Harb Perspect Med 2015;5(5):a022392 12. Michaud JL, Lachance M, Hamdan FF, et al. The genetic landscape of infantile spasms. Hum Mol Genet 2014;23(18):4846–4858 13. Nordli Jr. DR, Pellock JM, Sankar R, Wheless JW, eds. Neonatal seizures. In: Pellock’s Pediatric Epilepsy: Diagnosis and Therapy. 4th ed. Washington, DC: Springer Publishing Company; 2016:489–503 14. Yamatogi Y, Ohtahara S. Early-infantile epileptic encephalopathy with suppression-bursts, Ohtahara syndrome; its overview referring to our 16 cases. Brain Dev 2002;24(1):13–23 15. Ohtahara S, Yamatogi Y. Ohtahara syndrome: with special reference to its developmental aspects for differentiating from early myoclonic encephalopathy. Epilepsy Res 2006;70(Suppl 1):S58–S67 16. Coppola G, Plouin P, Chiron C, Robain O, Dulac O. Migrating partial seizures in infancy: a malignant disorder with developmental arrest. Epilepsia 1995;36(10):1017–1024 17. Weckhuysen S, Ivanovic V, Hendrickx R, et al; KCNQ2 Study Group. Extending the KCNQ2 encephalopathy spectrum: clinical and neuroimaging findings in 17 patients. Neurology 2013;81(19):1697–1703 18. Kato M, Yamagata T, Kubota M, et al. Clinical spectrum of early onset epileptic encephalopathies caused by KCNQ2 mutation. Epilepsia 2013;54(7):1282–1287 19. Millichap JJ, Park KL, Tsuchida T, et al. KCNQ2 encephalopathy: features, mutational hot spots, and ezogabine treatment of 11 patients. Neurol Genet 2016;2(5):e96

20. Ohtahara S, Yamatogi Y. Epileptic encephalopathies in early infancy with suppression-burst. J Clin Neurophysiol 2003; 20(6):398–407 21. Ishii A, Kang JQ, Schornak CC, et al. A de novo missense mutation of GABRB2 causes early myoclonic encephalopathy. J Med Genet 2017;54(3):202–211 22. Cohen R, Basel-Vanagaite L, Goldberg-Stern H, et al. Two siblings with early infantile myoclonic encephalopathy due to mutation in the gene encoding mitochondrial glutamate/H+ symporter SLC25A22. Eur J Paediatr Neurol 2014;18(6):801–805 23. Ohtahara S, Ohtsuka Y, Yamatogi Y, Oka E. The early-infantile epileptic encephalopathy with suppression-burst: developmental aspects. Brain Dev 1987;9(4):371–376 24. Saitsu H, Kato M, Okada I, et al. STXBP1 mutations in early infantile epileptic encephalopathy with suppression-burst pattern. Epilepsia 2010;51(12):2397–2405 25. Nakamura K, Kato M, Osaka H, et al. Clinical spectrum of SCN2A mutations expanding to Ohtahara syndrome. Neurology 2013;81(11):992–998 26. Otsuki T, Kim HD, Luan G, et al; FACE Study Group. Surgical versus medical treatment for children with epileptic encephalopathy in infancy and early childhood: Results of an international multicenter cohort study in Far-East Asia (the FACE study). Brain Dev 2016;38(5):449–460 27. Coppola G. Malignant migrating partial seizures in infancy: an epilepsy syndrome of unknown etiology. Epilepsia 2009;50(Suppl 5):49–51 28. McTague A, Appleton R, Avula S, et al. Migrating partial seizures of infancy: expansion of the electroclinical, radiological and pathological disease spectrum. Brain 2013;136(Pt 5):1578–1591 29. Barcia G, Fleming MR, Deligniere A, et al. De novo gain-of-function KCNT1 channel mutations cause malignant migrating partial seizures of infancy. Nat Genet 2012;44(11):1255–1259 30. Chong PF, Nakamura R, Saitsu H, Matsumoto N, Kira R. Ineffective quinidine therapy in early onset epileptic encephalopathy with KCNT1 mutation. Ann Neurol 2016;79(3):502–503 31. Milligan CJ, Li M, Gazina EV, et al. KCNT1 gain of function in 2 epilepsy phenotypes is reversed by quinidine. Ann Neurol 2014;75(4):581–590 32. Djuric M, Kravljanac R, Kovacevic G, Martic J. The efficacy of bromides, stiripentol and levetiracetam in two patients with malignant migrating partial seizures in infancy. Epileptic Disord 2011;13(1):22–26 33. Okuda K, Yasuhara A, Kamei A, Araki A, Kitamura N, Kobayashi Y. Successful control with bromide of two patients with malignant migrating partial seizures in infancy. Brain Dev 2000;22(1):56–59 34. Lux AL, Osborne JP. A proposal for case definitions and outcome measures in studies of infantile spasms and West syndrome: consensus statement of the West Delphi group. Epilepsia 2004;45(11):1416–1428 35. Vigevano F, Fusco L, Pachatz C. Neurophysiology of spasms. Brain Dev 2001;23(7):467–472 36. Hrachovy RA, Frost JD Jr. Infantile spasms. Handb Clin Neurol 2013;111:611–618 37. Watanabe K, Negoro T, Okumura A. Symptomatology of infantile spasms. Brain Dev 2001;23(7):453–466

67

68

I  Introduction to Epilepsy in Children 38. Yamamoto N, Watanabe K, Negoro T, et al. Partial seizures evolving to infantile spasms. Epilepsia 1988;29(1):34–40 39. Sankar R, Curran JG, Kevill JW, Rintahaka PJ, Shewmon DA, Vinters HV. Microscopic cortical dysplasia in infantile spasms: evolution of white matter abnormalities. AJNR Am J Neuroradiol 1995;16(6):1265–1272 40. Watanabe K, Negoro T, Aso K, Matsumoto A. Reappraisal of interictal electroencephalograms in infantile spasms. Epilepsia 1993;34(4):679–685 41. Fujii A, Oguni H, Hirano Y, Shioda M, Osawa M. A long-term, clinical study on symptomatic infantile spasms with focal features. Brain Dev 2013;35(5):379–385 42. Osborne JP, Lux AL, Edwards SW, et al. The underlying etiology of infantile spasms (West syndrome): information from the United Kingdom Infantile Spasms Study (UKISS) on contemporary causes and their classification. Epilepsia 2010;51(10):2168–2174 43. Allen AS, Berkovic SF, Cossette P, et al; Epi4K Consortium. Epilepsy Phenome/Genome Project. De novo mutations in epileptic encephalopathies. Nature 2013;501(7466):217–221 44. Chugani HT, Ilyas M, Kumar A, et al. Surgical treatment for refractory epileptic spasms: The Detroit series. Epilepsia 2015;56(12):1941–1949 45. Chugani HT, Shewmon DA, Shields WD, et al. Surgery for intractable infantile spasms: neuroimaging perspectives. Epilepsia 1993;34(4):764–771 46. Lux AL, Edwards SW, Hancock E, et al; United Kingdom Infantile Spasms Study. The United Kingdom Infantile Spasms Study (UKISS) comparing hormone treatment with vigabatrin on developmental and epilepsy outcomes to age 14 months: a multicentre randomised trial. Lancet Neurol 2005;4(11):712–717 47. O’Callaghan FJ, Edwards SW, Alber FD, et al; Participating Investigators. Safety and effectiveness of hormonal treatment versus hormonal treatment with vigabatrin for infantile spasms (ICISS): a randomised, multicentre, open-label trial. Lancet Neurol 2017;16(1):33–42 48. Knupp KG, Coryell J, Nickels KC, et al; Pediatric Epilepsy Research Consortium. Response to treatment in a prospective national infantile spasms cohort. Ann Neurol 2016;79(3):475–484 49. Coppola G, Pascotto A. The medical treatment of infantile spasms. In: Guzzetta F, Dalla Bernadina B, Guerrini R, eds. Progress in epileptic spasm and West syndrome. London, UK: John Libbey Eurotext; 2007:153–173 50. Hirano Y, Oguni H, Shiota M, Nishikawa A, Osawa M. Ketogenic diet therapy can improve ACTH-resistant West syndrome in Japan. Brain Dev 2015;37(1):18–22 51. Jonas R, Asarnow RF, LoPresti C, et al. Surgery for symptomatic infant-onset epileptic encephalopathy with and without infantile spasms. Neurology 2005;64(4):746–750 52. Harvey AS, Cross JH, Shinnar S, Mathern GW, Taskforce IPESS; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epilepsia 2008;49(1):146–155 53. Oguni H, Hayashi K, Osawa M, et al. Severe myoclonic epilepsy in infancy: clinical analysis and relation to SCN1A mutations in a ­Japanese cohort. Adv Neurol 2005;95:103–117 54. Dravet C, Oguni H. Dravet syndrome (severe myoclonic epilepsy in infancy). Handb Clin Neurol 2013;111:627–633 55. Dravet C, Bureau M, Oguni H, Cokar O, Guerrini R. Dravet syndrome (severe myoclonic epilepsy in infancy). In: Michelle B, Pierre G, Charlotte D, eds. Epileptic syndromes in infancy, childhood and adolescence. London, UK: John Libbey Eurotext; 2012:125–156 56. Tsuda Y, Oguni H, Sakauchi M, Osawa M. An electroclinical study of absence seizures in Dravet syndrome. Epilepsy Res 2013;103(1):88–96 57. Wang JW, Kurahashi H, Ishii A, et al. Microchromosomal deletions involving SCN1A and adjacent genes in severe myoclonic epilepsy in infancy. Epilepsia 2008;49(9):1528–1534

58. Oguni H, Hayashi K, Oguni M, et al. Treatment of severe myoclonic epilepsy in infants with bromide and its borderline variant. Epilepsia 1994;35(6):1140–1145 59. Aras LM, Isla J, Mingorance-Le Meur A. The European patient with Dravet syndrome: results from a parent-reported survey on antiepileptic drug use in the European population with Dravet syndrome. Epilepsy Behav 2015;44:104–109 60. Sakauchi M, Oguni H, Kato I, et al. Retrospective multiinstitutional study of the prevalence of early death in Dravet syndrome. Epilepsia 2011;52(6):1144–1149 61. Scheffer IE, Turner SJ, Dibbens LM, et al. Epilepsy and mental retardation limited to females: an under-recognized disorder. Brain 2008;131(Pt 4):918–927 62. Dibbens LM, Tarpey PS, Hynes K, et al. X-linked protocadherin 19 mutations cause female-limited epilepsy and cognitive impairment. Nat Genet 2008;40(6):776–781 63. Depienne C, Bouteiller D, Keren B, et al. Sporadic infantile epileptic encephalopathy caused by mutations in PCDH19 ­resembles Dravet syndrome but mainly affects females. PLoS Genet 2009;5(2):e1000381 64. Higurashi N, Nakamura M, Sugai M, et al. PCDH19-­ related female-limited epilepsy: further details regarding early clinical features and therapeutic efficacy. Epilepsy Res 2013;106(1–2):191–199 65. Marini C, Darra F, Specchio N, et al. Focal seizures with affective symptoms are a major feature of PCDH19 gene-related epilepsy. Epilepsia 2012;53(12):2111–2119 66. Lotte J, Bast T, Borusiak P, et al. Effectiveness of antiepileptic therapy in patients with PCDH19 mutations. Seizure 2016;35:106–110 67. Higurashi N, Shi X, Yasumoto S, et al. PCDH19 mutation in­ Japanese females with epilepsy. Epilepsy Res 2012; 99(1–2):28–37 68. Doose H, Gerken H, Leonhardt R, Völzke E, Völz C. Centrencephalic myoclonic-astatic petit mal. Clinical and genetic investigation. Neuropadiatrie 1970;2(1):59–78 69. Oguni H, Hayashi K, Imai K, et al. Idiopathic myoclonic-astatic epilepsy of early childhood—nosology based on electrophysiologic and long-term follow-up study of patients. Adv Neurol 2005;95:157–174 70. Oguni H, Fukuyama Y, Imaizumi Y, Uehara T. Video-EEG analysis of drop seizures in myoclonic astatic epilepsy of early childhood (Doose syndrome). Epilepsia 1992;33(5):805–813 71. Oguni H, Uehara T, Imai K, Osawa M. Atonic epileptic drop attacks associated with generalized spike-and-slow wave complexes: video-polygraphic study in two patients. Epilepsia 1997;38(7):813–818 72. Itoh Y, Oguni H, Hirano Y, Osawa M. Study of epileptic drop attacks in symptomatic epilepsy of early childhood—differences from those in myoclonic-astatic epilepsy. Brain Dev 2015;37(1):49–58 73. Oguni H, Tanaka T, Hayashi K, et al. Treatment and long-term prognosis of myoclonic-astatic epilepsy of early childhood. Neuropediatrics 2002;33(3):122–132 74. Wiemer-Kruel A, Haberlandt E, Hartmann H, Wohlrab G, Bast T. Modified Atkins diet is an effective treatment for children with Doose syndrome. Epilepsia 2017;58(4):657–662 75. Arzimanoglou A, French J, Blume WT, et al. Lennox–Gastaut syndrome: a consensus approach on diagnosis, assessment, management, and trial methodology. Lancet Neurol 2009;8(1):82–93 76. Kaminska A, Oguni H. Lennox–Gastaut syndrome and epilepsy with myoclonic-astatic seizures. Handb Clin Neurol 2013;111:641–652 77. Fujii A, Oguni H, Hirano Y, Osawa M. Atypical benign partial epilepsy: recognition can prevent pseudocatastrophe. Pediatr Neurol 2010;43(6):411–419 78. Hahn A. Atypical benign partial epilepsy/pseudo-Lennox ­syndrome. Epileptic Disord 2000;2(Suppl 1):S11–S17

9  Infantile and Childhood-Onset Catastrophic Epilepsy Syndromes 79. Montouris GD, Wheless JW, Glauser TA. The efficacy and tolerability of pharmacologic treatment options for Lennox–Gastaut syndrome. Epilepsia 2014;55(Suppl 4):10–20 80. Ohtsuka Y, Yoshinaga H, Shirasaka Y, Takayama R, Takano H, Iyoda K. Long-term safety and seizure outcome in Japanese patients with Lennox–Gastaut syndrome receiving adjunctive rufinamide therapy: An open-label study following a randomized clinical trial. Epilepsy Res 2016;121:1–7 81. Kossoff EH, Shields WD. Nonpharmacologic care for patients with Lennox–Gastaut syndrome: ketogenic diets and vagus nerve stimulation. Epilepsia 2014;55(Suppl 4):29–33 82. Cross JH. The ketogenic diet in the treatment of Lennox–Gastaut syndrome. Dev Med Child Neurol 2012;54(5):394–395 83. Lancman G, Virk M, Shao H, et al. Vagus nerve stimulation vs. corpus callosotomy in the treatment of Lennox–Gastaut syndrome: a meta-analysis. Seizure 2013;22(1):3–8 84. Katagiri M, Iida K, Kagawa K, et al. Combined surgical intervention with vagus nerve stimulation following corpus callosotomy in patients with Lennox–Gastaut syndrome. Acta Neurochir (Wien) 2016;158(5):1005–1012 85. Tassinari CA, Cantalupo G, Dalla Bernardina B, et al. Encephalopathy related to status epilepticus during slow sleep (ESES) including Landau–Kleffner syndrome. In: Epileptic Syndromes in Infancy, Childhood and Adolescence. London, UK: John Libbey Eurotext; 2012:255–275 86. Patry G, Lyagoubi S, Tassinari CA. Subclinical “electrical status epilepticus” induced by sleep in children. A clinical and electroencephalographic study of six cases. Arch Neurol 1971;24(3):242–252 87. Veggiotti P, Pera MC, Teutonico F, Brazzo D, Balottin U, Tassinari CA. Therapy of encephalopathy with status epilepticus during sleep (ESES/CSWS syndrome): an update. Epileptic Disord 2012;14(1):1–11 88. Paetau R. Magnetoencephalography in Landau–Kleffner syndrome. Epilepsia 2009;50(Suppl 7):51–54 89. Hirsch E, Valenti MP, Rudolf G, et al. Landau–Kleffner syndrome is not an eponymic badge of ignorance. Epilepsy Res 2006;70(Suppl 1):S239–S247 90. Robinson RO, Baird G, Robinson G, Simonoff E. Landau–Kleffner syndrome: course and correlates with outcome. Dev Med Child Neurol 2001;43(4):243–247 91. Tachikawa E, Oguni H, Shirakawa S, Funatsuka M, Hayashi K, Osawa M. Acquired epileptiform opercular syndrome: a case report and results of single photon emission computed tomography and computer-assisted electroencephalographic analysis. Brain Dev 2001;23(4):246–250 92. Galanopoulou AS, Bojko A, Lado F, Moshé SL. The spectrum of neuropsychiatric abnormalities associated with electrical status epilepticus in sleep. Brain Dev 2000;22(5):279–295 93. Kramer U, Sagi L, Goldberg-Stern H, Zelnik N, Nissenkorn A, Ben-Zeev B. Clinical spectrum and medical treatment of

c­ hildren with electrical status epilepticus in sleep (ESES). Epilepsia 2009;50(6):1517–1524 94. Fernández IS, Peters JM, Hadjiloizou S, et al. Clinical staging and electroencephalographic evolution of continuous spikes and waves during sleep. Epilepsia 2012;53(7):1185–1195 95. Sánchez Fernández I, Chapman K, Peters JM, et al. Treatment for continuous spikes and waves during sleep (CSWS): survey on treatment choices in North America. Epilepsia 2014;55(7):1099–1108 96. Morrell F, Whisler WW, Smith MC, et al. Landau–Kleffner syndrome. Treatment with subpial intracortical transection. Brain 1995;118(Pt 6):1529–1546 97. Downes M, Greenaway R, Clark M, et al. Outcome following multiple subpial transection in Landau–Kleffner syndrome and related regression. Epilepsia 2015;56(11):1760–1766 98. Smith MC, Hoeppner TJ. Epileptic encephalopathy of late childhood: Landau–Kleffner syndrome and the syndrome of continuous spikes and waves during slow-wave sleep. J Clin Neurophysiol 2003;20(6):462–472 99. Paquier PF, Van Dongen HR, Loonen CB. The Landau–Kleffner syndrome or ‘acquired aphasia with convulsive disorder’. Longterm follow-up of six children and a review of the recent literature. Arch Neurol 1992;49(4):354–359 100. Soprano AM, Garcia EF, Caraballo R, Fejerman N. Acquired epileptic aphasia: neuropsychologic follow-up of 12 patients. Pediatr Neurol 1994;11(3):230–235 101. Oguni H, Andermann F, Rasmussen T. The natural history of the syndrome of chronic encephalitis and epilepsy: a study of the MNI series of forty-eight cases. In: Andermann F, ed. Chronic Encephalitis and Epilepsy—Rasmussen’s Syndrome. Boston, MA: Butterworth–Heinemann; 1991:7–35 102. Bien CG, Granata T, Antozzi C, et al. Pathogenesis, diagnosis and treatment of Rasmussen encephalitis: a European consensus statement. Brain 2005;128(Pt 3):454–471 103. Oguni H, Andermann F, Rasmussen TB. The syndrome of chronic encephalitis and epilepsy. A study based on the MNI series of 48 cases. Adv Neurol 1992;57:419–433 104. Muto A, Oguni H, Takahashi Y, et al. Nationwide survey (incidence, clinical course, prognosis) of Rasmussen’s encephalitis. Brain Dev 2010;32(6):445–453 105. Varadkar S, Bien CG, Kruse CA, et al. Rasmussen’s encephalitis: clinical features, pathobiology, and treatment advances. Lancet Neurol 2014;13(2):195–205 106. Takahashi Y, Yamazaki E, Mine J, et al. Immunomodulatory therapy versus surgery for Rasmussen syndrome in early childhood. Brain Dev 2013;35(8):778–785 107. Bien CG, Tiemeier H, Sassen R, et al. Rasmussen encephalitis: incidence and course under randomized therapy with tacrolimus or intravenous immunoglobulins. Epilepsia 2013;54(3):543–550

69

10

  Epilepsy Surgery for Congenital or Early Brain Lesions Ahsan Moosa Naduvil Valappil, Tobias Loddenkemper, and Elaine Wyllie

Summary Identification of candidates for epilepsy surgery was originally derived from experience in adult and adolescent patients, often with epileptogenic lesions acquired later in life. In this patient population, the hallmark features predicting postoperative seizure freedom included focal lesion seen on neuroimaging, together with congruent focal features on seizure semiology, ictal, and interictal electroencephalography (EEG). In the 1990s, with successful surgical treatment of group of children with infantile spasms due to focal cortical dysplasia, it became accepted that in very young surgical candidates, the seizure semiology and the EEG may lack the focal features that are characteristic of epileptogenic lesions acquired later in life. Congenital or early-acquired brain lesions that occur during the phase of rapid brain growth and development in the environment of immature nervous system have special implications for the clinical and electrophysiological phenotype of epilepsy, with generalized and sometimes contralesional EEG discharges in up to one-fourth of children. In these children with a congenital or early-acquired lesion, the evaluation and decision for epilepsy surgery is no longer based on traditional unilateral or focal findings on EEG and seizure semiology. Instead it is based on the comprehensive picture of a child with severe epilepsy, a potentially epileptogenic unilateral lesion seen on MRI, and focal or generalized EEG and seizure types. Early surgery may be warranted in selected children with severe epilepsy to reduce the seizure and medication burden and capitalize on the period of plasticity to maximize developmental potential. Keywords:  epilepsy surgery, epileptogenic lesions, maladaptive plasticity, hemispheric lesions

„„ Introduction Identification of candidates for epilepsy surgery was originally derived from experience in adult and adolescent patients, often with epileptogenic lesions acquired later in life.1,​2 In this patient population, the hallmark features predicting postoperative seizure freedom included focal lesion seen on neuroimaging, together with congruent focal features on seizure semiology, ictal, and interictal EEG.3 As MRI became more sensitive for identification of more subtle lesions such as malformations of cortical development (MCDs),4,​5 and as epilepsy surgery came into the mainstream as a treatment

modality for infants and children,6,​7,​8,​9,​10 certain challenges to this selection paradigm emerged. In the 1990s, with successful surgical treatment of group of children with infantile spasms due to focal cortical dysplasia, it became accepted that in very young surgical candidates, the seizure semiology and the EEG may lack the focal features that are characteristic of epileptogenic lesions acquired later in life.11 Thus, these early lesions pose unique challenges in selection of surgical candidates. Issues related to selection of pediatric epilepsy surgical candidates are discussed in several chapters throughout the book. In this chapter, we will address: (1) the special challenges presented by early epileptogenic lesions in older children and adolescents, as well as infants; and (2) issues related to timing of surgery in pediatric patients.

„„ Types of Congenital or EarlyAcquired Lesions Causing Epilepsy From the standpoint of surgical planning, congenital or early-acquired lesions may be categorized into two broad ­ groups: hemispheric lesions and focal or multifocal lesions.12 The former group requires hemispherectomy and the latter group needs lobar or multilobar resection. Additionally, localization of the epileptogenic zone, distribution of eloquent areas, age, and plasticity potential needs to be considered in the ­equation.

Hemispheric Lesions Epileptogenic lesions affecting the whole or most of a hemisphere are more common in pediatric than adult epilepsy surgery candidates. MCDs, Sturge–Weber syndrome, encephalomalacia due to a variety of insults, and Rasmussen’s encephalitis constitute the four major groups. Such lesions are typically associated with hemiparesis, with or without hemianopia.

Malformation of Cortical Development First, MCDs can affect the whole or most of a hemisphere referred here as hemispheric malformations.13 The prototype of hemispheric MCD is hemimegalencephaly. Hemimegalencephaly may occur as an isolated abnormality or with neurocutaneous syndromes such as epidermal nevus syndrome,14 hypomelanosis of Ito,15 Klippel–Trenauney–Weber syndrome,16

10  Epilepsy Surgery for Congenital or Early Brain Lesions Proteus syndrome,17 neurofibromatosis type 1,18 and tuberous sclerosis.19 Precise diagnosis is important in such cases to address other systemic problems in these syndromes. In some cases, megalencephaly may not affect the whole hemisphere and may be confined to posterior or anterior regions of affected hemisphere. Some authors refer to these cases as hemi-­ hemimegalencephaly.20 Some of the hemispheric MCD may have associated atrophy and are distinct from megalencephalic malformations.21 Other less common malformations in surgical series include schizencephaly and polymicrogyria.21,​22 Subtle or overt abnormalities are frequent in the contralateral hemisphere and may be present in up to two-thirds of children who undergo hemispherectomy.13 Gross bilateral malformations like lissencephaly and subcortical band heterotopia are excluded from this group as they are usually not amenable to epilepsy surgery at present.

Sturge–Weber Syndrome Second, Sturge–Weber syndrome was one of the first disorders to undergo surgery in an early series of hemispherectomy in infants.23 This syndrome can be easily recognized by the clinical triad of facial nevus flammeus, contralateral hemiparesis, visual field defect, and characteristic neuroradiological abnormalities including leptomeningeal and intraparenchymal angiomatosis typically in the posterior quadrant, choroid plexus hypertrophy, gyriform calcification, and progressive regional or hemispheric cerebral atrophy.24,​25

Encephalomalacia Third, extensive cystic encephalomalacia or gliosis affecting most of the hemisphere form an important group in recent pediatric surgical series.26,​27,​28 Encephalomalacia acquired early in life is most often due to pre- or perinatal cerebral artery infarction, intraventricular hemorrhage, or hypoxia–ischemia. Trauma and infection are important etiologies in the postnatal period, infancy, and early childhood.

Rasmussen’s Syndrome Lastly, Rasmussen’s syndrome is a progressive disorder characterized by severe unilateral focal epilepsy, often with epilepsia partialis continua and progressive neurological deficits including hemiparesis, cognitive decline, and (if the dominant hemisphere is involved) aphasia.29,​30,​31 MRI in early stages is often normal. Few patients show transient focal cortical swelling at the onset of disease. As disease advances, progressive hemispheric atrophy most prominent in insular and peri-insular regions with increased cortical and subcortical signal on T2-weighted and fluid-attenuated inversion recovery sequences appear.30,​32 Involvement of ipsilateral caudate head and putamen is common.33 Rarely both hemispheres may be affected.34

Neoplasms Rarely tumors may involve a large part of one hemisphere but surgery in this situation may not be for the sole purpose of seizure control. Gliomatosis cerebri may be misdiagnosed as hemimegalencephaly, especially when it occurs in very young children with refractory epilepsy.35

Focal Lesions The single major group of early focal epileptogenic lesions ­requiring epilepsy surgery in children is focal cortical dysplasia.7,​36 Other focal malformations would include heterotopia, polymicrogyria, schizencephaly, and hypothalamic hamartomas.37 Dysembryoblastic neuroectodermal tumors and multifocal cortical tubers of tuberous sclerosis also fall in the spectrum of dysplasias on the border zone of neoplasms.37 Other early tumors include ganglioglioma, gangliocytoma, and pleomorphic xanthoastrocytoma.7,​38 Focal areas of gliosis due to remote vascular insult, trauma, and infection constitute the rest of early focal lesions.9,​39 Mesial temporal sclerosis (MTS) is an uncommon lesion in early life9 although cases as young as 4 months have been described.40 In pediatric epilepsy surgical candidates, MTS is more likely to exist as dual pathology with ipsilateral anterior temporal cortical dysplasia.9,​38,​41

Nonlesional MRI Despite advances in structural neuroimaging, a significant number of patients with drug-resistant focal epilepsy do not have an identifiable structural lesion on MRI. Successful surgical excision has been performed in many of these apparently nonlesional patients based on convergence of clinical and electrophysiological data7 or supported by functional imaging studies such as positron emission tomography (PET) or ictal single-photon emission CT.11 Invasive monitoring is frequently needed in this group to map the potential seizure onset zone. Histopathology often reveal evidence of cortical dysplasia, neuronal heterotopia, and focal gliosis as the substrates of epilepsy in majority of MRI-negative lesions.7,​42

„„ Impact of Early Lesions Early Lesions: Role of Adaptive Plasticity Unlike in adults, plasticity of the young brain serves as a cushion against new postoperative neurological deficits after resection of eloquent regions, including language, motor, and primary visual cortex areas.43,​44,​45 Plasticity, in this context, in simple terms refers to recruitment of neurons to perform an eloquent function which is otherwise not destined to. Such gain of function through plasticity is referred to as adaptive plasticity.44 The classical example is transfer of language functions to right hemisphere with early left hemispheric injuries.43,​46–50 Dominant handedness can also transfer effectively with early injuries. Handedness is likely determined by the presence of praxis center in the inferior parietal lobule rather than language areas.51 However, language and praxis centers often exist in the same hemisphere and hence handedness is a surrogate marker for language dominance.51,​52 Motor, sensory, and visual functions have limited plasticity and the contribution of neuronal plasticity to recovery is limited. Plasticity of cognitive functions is difficult to evaluate separately but likely to be significantly more than sensorimotor functions. Plasticity and the resultant transfer of functions are best studied in the language domain. The anatomic location of B ­ roca’s area and Wernicke’s area are fairly consistent in the majority of normal subjects. Epileptogenic lesions in and around these locations

71

72

I  Introduction to Epilepsy in Children displace the language functions to adjacent areas or in extreme instances to the homologous area in opposite h ­ emisphere.49,​ 50,​53 Three major factors influence the transfer of the language functions: age at insult, the size of the lesion, and the nature of the lesion. Younger age at insult, large lesions, and destructive lesions are more likely to shift the language areas. Of these the most important factor is age at insult and 6 years is generally the cut off for effective language transfer to occur.50 However, many such early lesions may not manifest with epilepsy until later, technically beyond the period of effective plasticity. In these instances, the age at the time of brain injury is more important than the age at onset of epilepsy or age at surgery. Hence, these plasticity rules may apply even in adults provided the lesion is acquired within the period of plasticity window, typically before 6 years of age. Even though effective transfer of functions occurs when the brain injury occurs early in life, some degree of plasticity may be possible in late childhood and even in adults.47,​54,​55 Large hemispheric lesions as in perinatal stroke or hemimegalencephaly shift language areas to the opposite hemisphere.49,​50 On the contrary, smaller lesions such as tumors tend to displace language areas to adjacent areas in the same hemisphere.53 This information is critical when operating on such lesions, as peritumoral bed may harbor language function. Transferred language function is never as good as ­naturally acquired language, and several studies suggest that the left hemisphere is phylogenetically superior in language acquisition in most patients.43,​56 This applies more so to expressive as compared to receptive language functions. The intracarotid amobarbital test (Wada test) is difficult to perform in children with neurocognitive deficits but has been reported to be of help in selected cases.57 Functional MRI and child-specific language testing paradigms may soon replace the Wada test for language lateralization.58,​59

Early Lesions: Electroencephalography Manifestations Interactions between an early epileptogenic brain lesion and normal developmental processes may result in EEG patterns that are different from, and more diffuse than, those seen in patients with lesions acquired after brain maturity. This was first recognized in young children who presented with the generalized pattern of hypsarrhythmia and infantile spasms, and had seizure-free outcome after resection of a congenital or early-acquired focal or hemispheric brain lesion.11,​42 Subsequently, successful epilepsy surgery was also reported for older children and adolescents with early brain lesions and other generalized EEG patterns, including the generalized pattern of slow spikewave complexes that is traditionally associated with Lennox– Gastaut syndrome (Fig. 10.1) or the contralesional patterns (maximum epileptiform abnormalities over the hemisphere contralateral to the lesion) seen in the setting of extensive unilateral encephalomalacia (Fig. 10.2).26,​60 Neither West syndrome with hypsarrhythmia nor Lennox– Gastaut syndrome with generalized slow spike-wave complexes were recognized as surgically amenable initially.61,​62 These syndromes are not etiology specific and can occur in response to a wide variety of cerebral insults. The occurrence of identifiable focal and multifocal lesions in these disorders prompted several centers to attempt lesionectomies in these patients who otherwise had multidrug-resistant epilepsy with very high ­seizure burden with very little options left.8,​11,​26,​60,​63,​64,​65 ­Successful

surgery in these patients has widened the spectrum of epilepsy phenotypes amenable to surgery. The key factor for manifestation of the generalized patterns in these surgical candidates appeared to be the age at the time of lesion occurrence (78% pre- or perinatal; 90% within the first 2 years of life in one series)26 rather than the age at evaluation for epilepsy surgery (0.2–24 years; median: 8 years). This phenomenon of “generalized patters in focal lesions” follows the rules similar to the adaptive plasticity rules in language transfer described earlier. When we assess an individual for left hemispherectomy at any age, the important factor deciding the side of language dominance is age at lesion occurrence, not age at presurgical evaluation.50 Generalized EEG patterns seen in focal lesions appear to follow the same rules.26 Although not all children or adolescents with early lesions will present with generalized or contralesional epileptiform discharges, it is important to be aware of the phenomenon so that carefully selected children with focal or unilateral epileptogenic lesions are not excluded from surgical consideration. In one series,26 ­eizure-free outcome did not differ between children and adolescents with early brain lesions who had generalized EEG abnormalities (72% of patients were seizure free at the last follow-up) compared to children with ipsilesional epileptiform discharges. In this series, the contralesional EEG abnormalities (24% of patients) appeared to represent an altered expression of diffuse EEG p ­ attern, decreased on the ipsilesional side affected by large destructive lesion. Such false localizing and lateralizing EEG abnormalities are not limited to large destructive hemispheric lesions. Even in typical focal epileptic syndromes such as temporal lobe epilepsy due to neoplasms in childhood may express multiregional and contralesional EEG abnormalities which d ­ isappeared after successful removal of the focal epileptogenic lesions.66,​67 Most of the patients in the latter series had early-­onset brain lesions. The mechanisms underlying the generalized epileptiform abnormalities in focal cerebral lesions are unknown. Some view this as a form of maladaptive plasticity of immaturebrain.44,​60 The lesions in the environment of immature/maturing neural network of young brain alter the neural circuits leading to spontaneous hypersynchrony, resulting in generalized features.68 Some authors suggested involvement of the thalamocortical networks.69 Generalized epileptiform abnormalities have been described with hypothalamic hamartomas as well.65 In a study on the evolution of epilepsy in hypothalamic hamartomas in children, gelastic seizures were noted in infancy followed by generalized tonic seizures at around 6 years of age. Early EEGs were normal but later became progressively abnormal with emergence of generalized epileptiform abnormalities. Intraoperative EEG showed persistence of generalized interictal spike-wave even after removal of hamartoma, but these discharges resolved in postoperative studies. This observation suggests that the generalized epileptiform discharges may be a result of secondary epileptogenesis similar to the kindling phenomenon seen in animals.70 It is hypothesized that these discharges may u ­ ltimately resolve if the kindling primary focus is removed before the ­secondary epileptogenic areas become completely independent.

Timing of Surgery Intractable epilepsy within the first 2 years of life is a significant risk factor for mental handicap, especially if seizures occur daily.71 Also, repeated episodes of prolonged status epilepticus carry the risk of more global damage. Mechanisms of brain

10  Epilepsy Surgery for Congenital or Early Brain Lesions

Fig. 10.1  (a) Axial MRI demonstrates right hemispheric malformation of cortical development in an 8-year-old girl with left hemiparesis, mental impairment, and refractory epilepsy since 18 months of age. Seizures included daily multiple drop attacks and episodes of slumping over with unresponsiveness and head bobbing for 10 to 30 seconds, 50 to 100 per day. (b) Interictal EEG shows generalized slow spike-wave complexes. (c) Ictal EEG illustrates generalized slow spikewave complexes during an episode of slumping over with unresponsiveness and head nodding. (d) Most of the bursts of slow spike-wave complexes were bilaterally synchronous at onset, but a few had an initial maximum or lead in from the left (shown here) or right side. (e) Postoperative EEG, 6 months after right functional hemispherectomy showing expected diminished background and sharp waves on the right and persistent focal sharp waves on the left side. Follow-up EEG at 1, 2, and 3 years after surgery showed no contralateral interictal epileptiform discharges. (Reproduced with permission from Wyllie et al 2007.26)

injury may vary with etiology. For instance, in Sturge–Weber ­syndrome, progressive neurological deficits occur due to two processes: hypoxemia due to thrombosis and venous stasis, which affects the ipsilesional hemisphere and recurrent seizures, which may affect the contralesional hemisphere.72,​73 Early successful surgery may prevent damage to contralesional side. Epilepsy surgery within the period of effective neuronal plasticity window provides best opportunity to maximize developmental potential and minimize postoperative deficit. As noted earlier, surgery involving the dominant hemisphere is less likely to result in major language deficits if performed before 6 years of age,50 even if the eloquent areas had resided in the surgical bed. Developmental quotient may improve for children operated in infancy.74 In another study,75 cognitive function improved in preschool children following epilepsy surgery, with improved catch-up development in children rendered seizure free. Shorter duration of epilepsy was significantly associated with postoperative improvement in developmental quotient.75 Improvement in developmental outcome in these children may be due to multiple

factors including seizure control, resolution of epileptic encephalopathy, and reduction of toxic antiepileptic drugs. As a guiding principle, the ideal age for epilepsy surgery may be the earliest age at which the usual selection criteria for surgery are met.6,​12 These criteria are built on three critical questions: Is surgery warranted? Will it work? Is it safe? If the answer is yes to all three questions, then surgery should be performed at the earliest. For children with epileptic encephalopathy early in life, surgery within the first year of life may be indicated, although this entails special risks and may be most safely performed at specialized pediatric centers.

„„ Selection of Surgical Candidates In the child or adolescent with a congenital or early-acquired lesion, the evaluation and decision for epilepsy surgery is no longer based on traditional unilateral or focal findings on

73

74

I  Introduction to Epilepsy in Children

Fig. 10.2  (a) Axial and sagittal MRI shows cystic encephalomalacia due to perinatal left middle cerebral artery infarction in a 12-yearold boy with severe right hemiparesis, mental impairment, and refractory epilepsy since 9 years of age. He had daily tonic seizures. (b) Interictal awake EEG reveals a run of generalized slow spike-wave complexes. (c) Interictal EEG during sleep shows generalized polyspike-wave complexes, sometimes maximum on the right (shown here) or left side. (d) Interictal EEG demonstrates a spike maximum on the right, and continuous slowing in the left frontal region. (e) Ictal EEG shows initial movement artifact followed by generalized polyspikes, maximum in the right frontal region. Identical pattern was seen in all 11 recorded seizures. (Reproduced with permission from Wyllie et al 2007.26)

EEG and seizure semiology. Instead it is based on the comprehensive picture of a child with severe epilepsy, a potentially ­epileptogenic unilateral lesion seen on MRI, and focal or generalized EEG and seizure types. Surgical decision-making is fairly straightforward in patients with focal lesions with concordant localizing seizures and EEG abnormalities, but it is more challenging when EEG is generalized. Focal clinical features during seizures are helpful when they suggest seizure onset on the side of the lesion; but these may or may not be present, even in children with subsequent seizure-free postoperative outcome.26,​60 Hemiparesis may also be helpful to surgical planning, as long as it is concordant with the side of proposed surgery. Surgical decision-making is challenging when the seizures and EEG are generalized. In the early series of generalized epilepsy phenotypes undergoing surgery, coexisting dominant focal features on EEG and semiology were taken as primary factors for determining epilepsy surgical candidacy.8,​11,​63,​ 76 However, recent experience has shown that surgery can be

successful in children and adolescents with generalized EEG patterns and an early unilateral brain lesion.26,​60 The role of invasive ­monitoring is limited in this group, and was not used to clarify the extent of the epileptogenic zone in any patients in these series26,​60 who were seizure free after surgery. Just as hypsarrhythmia is an accepted diffuse manifestation of an early epileptogenic lesion in infants, slow spike-wave complexes and other forms of generalized discharges may be the manifestation of such lesions later in childhood. In children and adolescents with extensive encephalomalacia due to an early destructive brain injury, the ipsilesional hemisphere may not express normal or abnormal EEG activity (Fig. 10.2). In these cases, ictal and interictal epileptiform discharges may be maximum over the contralesional hemisphere, possibly representing a generalized discharge that is reduced on the ipsilesional side.26 In every case, two key features for consideration of surgery include the nature, timing, and extent of the unilateral brain lesion, together with the severity of the

10  Epilepsy Surgery for Congenital or Early Brain Lesions epilepsy.26,​60 Performing preoperative video-EEG evaluation in this group of patients remains important for the following three reasons: (1) to identify focal EEG features or lateralizing signs during clinical seizures which would help to build a case for surgery in complicated cases; (2) to clarify that all of the reported clinical events are in fact seizures and not nonepileptic events; and (3) to expand our understanding of such cases and their postsurgical evolution. EEG plays an important role in children with multifocal lesions as in tuberous sclerosis. Careful identification of the epileptogenic zone can lead to seizure freedom in some of these patients.77,​78 EEG abnormalities concordant with an isolated or the largest tuber predict good outcome. Some centers have performed serial invasive monitoring in a multistaged approach in order to identify multiple possible epileptogenic lesions in tuberous sclerosis.79,​80 In this approach, a second surgery would include removal of the epileptogenic lesion confirmed by initial invasive monitoring and would also include placing of additional electrodes for further monitoring to map additional epileptogenic zones if any. This is not widely practiced due to significant postoperative morbidity associated with serial multiple surgeries.

Types of Surgeries for Early Lesions Resective epilepsy surgery for early lesions falls under three broad categories: hemispheric surgery, lesionectomy or lobar resection, and multilobar resection.7,​8,​21,​26 Decision-making regarding the type of surgery is a multidisciplinary approach involving pediatric epileptologists, neurosurgeons, neuropsychologists, neuroradiologists, social workers, and bioethicists. Hemispherectomies and multilobar resections are common surgeries performed in children and adolescents with early lesions,6,​7,​8,​9,​21,​26 but in some cases lesionectomy or more limited lobar resection may be appropriate.81,​82 The decision is based on a comprehensive assessment of all lines of evidence including seizure semiology, EEG, MRI, neurological examination, and other features. Infants and young children pose unique risks for major surgeries like hemispherectomy and an experienced pediatric epilepsy surgery team is crucial for better outcome. Callosotomy, multiple subpial transaction, and vagal nerve stimulation are mostly palliative. Readers are referred to dedicated chapters on surgeries for further details on these surgical procedures elsewhere in this book.

dren and adolescents is not significantly different from adults.9 Surgery in infancy can yield similar results in terms of seizure outcome.7,​8,​21,​42 In a series of 170 patients with hemispherectomy, 66% were seizure free at a mean follow-up of 5.3 years; another 5% achieved late remission and 9% had major improvement in seizure burden.87 Overall 80% had a favorable outcome. Similar outcome has been reported by other centers.21,​22,​88–​91 Although it seems counterintuitive, large lesions requiring the most extensive surgery (e.g., hemispheric infarction or posterior quadrant malformations) may in fact be associated with a relatively better chance for seizure-free outcome than smaller lesions near eloquent cortex (e.g., frontocentral malformations associated with little or no hemiparesis) that often lead a smaller and sometimes incomplete resection. In these cases the limitations on extent of surgery may reduce the ability to completely remove the epileptogenic zone. Many studies have looked at predictors of seizure free outcome. In one series,39 the complete resection of the epileptogenic lesion and the electrographically abnormal region was the major determinant of good outcome. The site of resection, lesional status, and pathologic diagnosis were not significant predictors of outcome.39 Intelligence quotient alone is not a predictor of outcome and patients with low intelligence should not be denied surgery if other presurgical data points to a resectable focus.92 Presence of generalized EEG abnormalities does not portend to poor outcome in carefully selected patients with early unilateral brain lesions.26,​60 In a series of 170 patients with hemispherectomy, age at seizure onset, age at surgery, seizure types, etiology, and presence of generalized EEG abnormalities and contralateral hemisphere abnormalities did not affect the seizure outcome. Presence of bilateral fluorodeoxyglucose PET abnormalities and acute postoperative seizures were associated with seizure recurrence. Successful epilepsy surgery has been shown to improve quality of life in children with intractable seizures.93,​94 Improvement in cognition and other developmental aspects following epilepsy surgery is unclear.95 In one series, temporal lobe resections have been shown to have a negative impact on verbal memory skills in high functioning children.96 However, compared to adults, greater functional recovery has been noted with both left and right temporal surgical resections in pediatric surgical candidates.45 Cognitive improvement can be tremendous in individual cases following successful epilepsy surgery.11 In one series a larger increase in developmental quotient was noted after surgery in children operated at younger age and those with epileptic spasms.74

„„ Outcome

„„ Conclusion

The pediatric epilepsy population is heterogeneous and many factors have to be taken into account prior to meaningful interpretation of outcome data. Outcome should emphasize not only seizure outcome but also developmental and cognitive outcome, and quality of life for patients and care givers.83 Risks of epilepsy surgery should be weighed against the benefits of surgery, and against the risk of continued uncontrolled seizures.83,​84 Despite age-related differences in etiology, pediatric surgical results for lobar resection are comparable to adults.85,​86 Seizure outcome after resection of cortical dysplasias in chil-

In conclusion, drug-resistant epilepsy in children is often secondary to congenital or early-acquired brain lesions that occur during the phase of rapid brain growth and development. These early lesions in the environment of immature nervous system have special implications for the clinical and electrophysiological phenotype of epilepsy, with generalized or contralesional EEG discharges in some patients. Early surgery may be warranted in selected children with severe epilepsy, to reduce the seizure and medication burden, and capitalize on the period of plasticity to maximize developmental potential.

75

76

I  Introduction to Epilepsy in Children

References 1. Engel J Jr. Update on surgical treatment of the epilepsies. Summary of the second international palm desert conference on the surgical treatment of the epilepsies (1992). Neurology 1993;43(8):1612–1617 2. Engel J Jr. Surgery for seizures. N Engl J Med 1996;334(10):647–652 3. Berg AT, Walczak T, Hirsch LJ, Spencer SS. Multivariable prediction of seizure outcome one year after resective epilepsy surgery: development of a model with independent validation. Epilepsy Res 1998;29(3):185–194 4. Kuzniecky R, Berkovic S, Andermann F, Melanson D, Olivier A, Robitaille Y. Focal cortical myoclonus and rolandic cortical dysplasia: clarification by magnetic resonance imaging. Ann Neurol 1988;23(4):317–325 5. Sperling MR, Wilson G, Engel J Jr, Babb TL, Phelps M, Bradley W. Magnetic resonance imaging in intractable partial epilepsy: correlative studies. Ann Neurol 1986;20(1):57–62 6. Wyllie E. Catastrophic epilepsy in infants and children: identification of surgical candidates. Epileptic Disord 1999;1(4): 261–264 7. Duchowny M, Jayakar P, Resnick T, et al. Epilepsy surgery in the first three years of life. Epilepsia 1998;39(7):737–743 8. Wyllie E, Comair YG, Kotagal P, Raja S, Ruggieri P. Epilepsy surgery in infants. Epilepsia 1996;37(7):625–637 9. Wyllie E, Comair YG, Kotagal P, Bulacio J, Bingaman W, Ruggieri P. Seizure outcome after epilepsy surgery in children and adolescents. Ann Neurol 1998;44(5):740–748 10. Centeno RS, Yacubian EM, Sakamoto AC, Ferraz AF, Junior HC, Cavalheiro S. Pre-surgical evaluation and surgical treatment in children with extratemporal epilepsy. Childs Nerv Syst 2006;22(8):945–959 11. Chugani HT, Shields WD, Shewmon DA, Olson DM, Phelps ME, Peacock WJ. Infantile spasms: I. PET identifies focal cortical dysgenesis in cryptogenic cases for surgical treatment. Ann Neurol 1990;27(4):406–413 12. Engel J Jr, Cascino GD, Shields WD. Surgically remediable syndromes. In: Engel J, Jr, Pedley TA, eds. Epilepsy: A Comprehensive Textbook. Vol II. Philadelphia, PA: Lippincott-Raven Publishers; 1997:1687–1696

22. Devlin AM, Cross JH, Harkness W, et al. Clinical outcomes of hemispherectomy for epilepsy in childhood and adolescence. Brain 2003;126(Pt 3):556–566 23. Hoffman HJ, Hendrick EB, Dennis M, Armstrong D. Hemispherectomy for Sturge–Weber syndrome. Childs Brain 1979;5(3):233–248 24. Krings T, Geibprasert S, Luo CB, Bhattacharya JJ, Alvarez H, Lasjaunias P. Segmental neurovascular syndromes in children. Neuroimaging Clin N Am 2007;17(2):245–258 25. Bourgeois M, Crimmins DW, de Oliveira RS, et al. Surgical treatment of epilepsy in Sturge–Weber syndrome in children. J Neurosurg 2007;106(1, Suppl):20–28 26. Wyllie E, Lachhwani DK, Gupta A, et al. Successful surgery for epilepsy due to early brain lesions despite generalized EEG findings. Neurology 2007;69(4):389–397 27. Guzzetta F, Battaglia D, Di Rocco C, Caldarelli M. Symptomatic epilepsy in children with poroencephalic cysts secondary to perinatal middle cerebral artery occlusion. Childs Nerv Syst 2006;22(8):922–930 28. Carreño M, Kotagal P, Perez Jiménez A, Mesa T, Bingaman W, Wyllie E. Intractable epilepsy in vascular congenital hemiparesis: clinical features and surgical options. Neurology 2002;59(1):129–131 29. Dubeau F. Rasmussen’s encephalitis (chronic focal encephalitis). In: Wyllie E, Gupta A, Lacchwani DK, eds. The Treatment of Epilepsy. 4th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2006:441–454 30. Bien CG, Granata T, Antozzi C, et al. Pathogenesis, diagnosis and treatment of Rasmussen encephalitis: a European consensus statement. Brain 2005;128(Pt 3):454–471 31. Hart Y. Rasmussen’s encephalitis. Epileptic Disord 2004;6(3): 133–144 32. Vining EP, Freeman JM, Brandt J, Carson BS, Uematsu S. Progres­ sive unilateral encephalopathy of childhood (Rasmussen’s ­syndrome): a reappraisal. Epilepsia 1993;34(4):639–650 33. Rajesh B, Kesavadas C, Ashalatha R, Thomas B. ­ Putaminal involvement in Rasmussen encephalitis. Pediatr Radiol 2006;36(8):816–822

13. Hallbook T, Ruggieri P, Adina C, et al. Contralateral MRI abnormalities in candidates for hemispherectomy for refractory epilepsy. Epilepsia 2010;51(4):556–563

34. Tobias SM, Robitaille Y, Hickey WF, Rhodes CH, Nordgren R, Andermann F. Bilateral Rasmussen encephalitis: postmortem documentation in a five-year-old. Epilepsia 2003;44(1):127–130

14. Pavone L, Curatolo P, Rizzo R, et al. Epidermal nevus syndrome: a neurologic variant with hemimegalencephaly, gyral malformation, mental retardation, seizures, and facial hemihypertrophy. Neurology 1991;41(2 ( Pt 1)):266–271

35. Maton B, Resnick T, Jayakar P, Morrison G, Duchowny M. Epilepsy surgery in children with gliomatosis cerebri. Epilepsia 2007;48(8):1485–1490

15. Tagawa T, Futagi Y, Arai H, Mushiake S, Nakayama M. Hypomelanosis of Ito associated with hemimegalencephaly: a clinicopathological study. Pediatr Neurol 1997;17(2):180–184 16. Cristaldi A, Vigevano F, Antoniazzi G, et al. Hemimegalencephaly, hemihypertrophy and vascular lesions. Eur J Pediatr 1995;154(2):134–137 17. Griffiths PD, Welch RJ, Gardner-Medwin D, Gholkar A, McAllister V. The radiological features of hemimegalencephaly including three cases associated with proteus syndrome. Neuropediatrics 1994;25(3):140–144 18. Cusmai R, Curatolo P, Mangano S, Cheminal R, Echenne B. Hemimegalencephaly and neurofibromatosis. Neuropediatrics 1990;21(4):179–182 19. Maloof J, Sledz K, Hogg JF, Bodensteiner JB, Schwartz T, Schochet SS. Unilateral megalencephaly and tuberous sclerosis: related disorders? J Child Neurol 1994;9(4):443–446 20. D’Agostino MD, Bastos A, Piras C, et al. Posterior quadrantic dysplasia or hemi-hemimegalencephaly: a characteristic brain malformation. Neurology 2004;62(12):2214–2220 21. González-Martínez JA, Gupta A, Kotagal P, et al. Hemispherectomy for catastrophic epilepsy in infants. Epilepsia 2005;46(9):1518–1525

36. Sisodiya SM. Surgery for malformations of cortical development causing epilepsy. Brain 2000;123(Pt 6):1075–1091 37. Raymond AA, Fish DR, Sisodiya SM, Alsanjari N, Stevens JM, Shorvon SD. Abnormalities of gyration, heterotopias, tuberous sclerosis, focal cortical dysplasia, microdysgenesis, dysembryoplastic neuroepithelial tumour and dysgenesis of the archicortex in epilepsy. Clinical, EEG and neuroimaging features in 100 adult patients. Brain 1995;118(Pt 3):629–660 38. Maton B, Jayakar P, Resnick T, Morrison G, Ragheb J, Duchowny M. Surgery for medically intractable temporal lobe epilepsy during early life. Epilepsia 2008;49(1):80–87 39. Paolicchi JM, Jayakar P, Dean P, et al. Predictors of outcome in pediatric epilepsy surgery. Neurology 2000;54(3):642–647 40. DeLong GR, Heinz ER. The clinical syndrome of early-life bilateral hippocampal sclerosis. Ann Neurol 1997;42(1):11–17 41. Mohamed A, Wyllie E, Ruggieri P, et al. Temporal lobe epilepsy due to hippocampal sclerosis in pediatric candidates for epilepsy surgery. Neurology 2001;56(12):1643–1649 42. Chugani HT, Shewmon DA, Shields WD, et al. Surgery for intractable infantile spasms: neuroimaging perspectives. Epilepsia 1993;34(4):764–771 43. Helmstaedter C, Kurthen M, Linke DB, Elger CE. Right hemisphere restitution of language and memory functions in right

10  Epilepsy Surgery for Congenital or Early Brain Lesions hemisphere language-dominant patients with left temporal lobe epilepsy. Brain 1994;117(Pt 4):729–737 44. Johnston MV. Clinical disorders of brain plasticity. Brain Dev 2004;26(2):73–80 45. Gleissner U, Sassen R, Schramm J, Elger CE, Helmstaedter C. Greater functional recovery after temporal lobe epilepsy surgery in children. Brain 2005;128(Pt 12):2822–2829 46. Helmstaedter C, Kurthen M, Linke DB, Elger CE. Patterns of language dominance in focal left and right hemisphere epilepsies: relation to MRI findings, EEG, sex, and age at onset of epilepsy. Brain Cogn 1997;33(2):135–150 47. Loddenkemper T, Wyllie E, Lardizabal D, Stanford LD, Bingaman W. Late language transfer in patients with Rasmussen encephalitis. Epilepsia 2003;44(6):870–871 48. Liégeois F, Connelly A, Cross JH, et al. Language reorganization in children with early-onset lesions of the left hemisphere: an fMRI study. Brain 2004;127(Pt 6):1229–1236 49. Rausch R, Walsh GO. Right-hemisphere language dominance in right-handed epileptic patients. Arch Neurol 1984;41(10): 1077–1080 50. Rasmussen T, Milner B. The role of early left-brain injury in determining lateralization of cerebral speech functions. Ann N Y Acad Sci 1977;299:355–369 51. Meador KJ, Loring DW, Lee K, et al. Cerebral lateralization: relationship of language and ideomotor praxis. Neurology 1999;53(9):2028–2031 52. Papagno C, Della Sala S, Basso A. Ideomotor apraxia without aphasia and aphasia without apraxia: the anatomical support for a double dissociation. J Neurol Neurosurg Psychiatry 1993;56(3):286–289 53. DeVos KJ, Wyllie E, Geckler C, Kotagal P, Comair Y. Language dominance in patients with early childhood tumors near left hemisphere language areas. Neurology 1995;45(2):349–356 54. Hertz-Pannier L, Chiron C, Jambaqué I, et al. Late plasticity for language in a child’s non-dominant hemisphere: a pre- and postsurgery fMRI study. Brain 2002;125(Pt 2):361–372 55. Boatman D, Freeman J, Vining E, et al. Language recovery after left hemispherectomy in children with late-onset seizures. Ann Neurol 1999;46(4):579–586 56. de Bode S, Curtiss S. Language after hemispherectomy. Brain Cogn 2000;43(1–3):135–138 57. Jansen FE, Jennekens-Schinkel A, Van Huffelen AC, et al. Diagnostic significance of Wada procedure in very young children and children with developmental delay. Eur J Paediatr Neurol 2002;6(6):315–320 58. Gaillard WD, Balsamo LM, Ibrahim Z, Sachs BC, Xu B. fMRI identifies regional specialization of neural networks for reading in young children. Neurology 2003;60(1):94–100 59. Gaillard WD, Berl MM, Moore EN, et al. Atypical language in lesional and nonlesional complex partial epilepsy. Neurology 2007;69(18):1761–1771 60. Gupta A, Chirla A, Wyllie E, Lachhwani DK, Kotagal P, Bingaman WE. Pediatric epilepsy surgery in focal lesions and generalized electroencephalogram abnormalities. Pediatr Neurol 2007;37(1):8–15 61. Nolte R, Christen HJ, Doerrer J. Preliminary report of a multicenter study on the West syndrome. Brain Dev 1988;10(4):236–242 62. Dulac O, N’Guyen T. The Lennox–Gastaut syndrome. Epilepsia 1993;34(Suppl 7):S7–S17 63. Kramer U, Sue WC, Mikati MA. Focal features in West syndrome indicating candidacy for surgery. Pediatr Neurol 1997; 16(3):213–217 64. Wyllie E, Comair Y, Ruggieri P, Raja S, Prayson R. Epilepsy surgery in the setting of periventricular leukomalacia and focal cortical dysplasia. Neurology 1996;46(3):839–841 65. Freeman JL, Harvey AS, Rosenfeld JV, Wrennall JA, Bailey CA, Berkovic SF. Generalized epilepsy in hypothalamic hamartoma: evolution and postoperative resolution. Neurology 2003;60(5):762–767

66. Blume WT, Girvin JP, Kaufmann JC. Childhood brain tumors presenting as chronic uncontrolled focal seizure disorders. Ann Neurol 1982;12(6):538–541 67. Wyllie E, Chee M, Granström ML, et al. Temporal lobe epilepsy in early childhood. Epilepsia 1993;34(5):859–868 68. Sutula TP. Mechanisms of epilepsy progression: current theories and perspectives from neuroplasticity in adulthood and development. Epilepsy Res 2004;60(2–3):161–171 69. Van Hirtum-Das M, Licht EA, Koh S, Wu JY, Shields WD, Sankar R. Children with ESES: variability in the syndrome. Epilepsy Res 2006;70(Suppl 1):S248–S258 70. Bertram E. The relevance of kindling for human epilepsy. Epilepsia 2007;48(Suppl 2):65–74 71. Vasconcellos E, Wyllie E, Sullivan S, et al. Mental retardation in pediatric candidates for epilepsy surgery: the role of early seizure onset. Epilepsia 2001;42(2):268–274 72. Jansen FE, van der Worp HB, van Huffelen A, van Nieuwenhuizen O. Sturge–Weber syndrome and paroxysmal hemiparesis: epilepsy or ischaemia? Dev Med Child Neurol 2004;46(11):783–786 73. Comi AM. Pathophysiology of Sturge-Weber syndrome. J Child Neurol 2003;18(8):509–516 74. Loddenkemper T, Holland KD, Stanford LD, Kotagal P, Bingaman W, Wyllie E. Developmental outcome after epilepsy surgery in infancy. Pediatrics 2007;119(5):930–935 75. Freitag H, Tuxhorn I. Cognitive function in preschool children after epilepsy surgery: rationale for early intervention. Epilepsia 2005;46(4):561–567 76. Asano E, Chugani DC, Juhász C, Muzik O, Chugani HT. Surgical treatment of West syndrome. Brain Dev 2001;23(7):668–676 77. Lachhwani DK, Pestana E, Gupta A, Kotagal P, Bingaman W, Wyllie E. Identification of candidates for epilepsy surgery in patients with tuberous sclerosis. Neurology 2005;64(9):1651–1654 78. Jansen FE, van Huffelen AC, Algra A, van Nieuwenhuizen O. Epilepsy surgery in tuberous sclerosis: a systematic review. Epilepsia 2007;48(8):1477–1484 79. Romanelli P, Najjar S, Weiner HL, Devinsky O. Epilepsy surgery in tuberous sclerosis: multistage procedures with bilateral or multilobar foci. J Child Neurol 2002;17(9):689–692 80. Weiner HL, Carlson C, Ridgway EB, et al. Epilepsy surgery in young children with tuberous sclerosis: results of a novel approach. Pediatrics 2006;117(5):1494–1502 81. Cataltepe O, Turanli G, Yalnizoglu D, Topçu M, Akalan N. Surgical management of temporal lobe tumor-related epilepsy in children. J Neurosurg 2005;102(3, Suppl):280–287 82. Minkin K, Klein O, Mancini J, Lena G. Surgical strategies and seizure control in pediatric patients with dysembryoplastic neuroepithelial tumors: a single-institution experience. J Neurosurg Pediatr 2008;1(3):206–210 83. Gilliam F, Wyllie E, Kashden J, et al. Epilepsy surgery outcome: comprehensive assessment in children. Neurology 1997;48(5):1368–1374 84. Donner EJ, Smith CR, Snead OC III. Sudden unexplained death in children with epilepsy. Neurology 2001;57(3):430–434 85. Mizrahi EM, Kellaway P, Grossman RG, et al. Anterior temporal lobectomy and medically refractory temporal lobe epilepsy of childhood. Epilepsia 1990;31(3):302–312 86. Adler J, Erba G, Winston KR, Welch K, Lombroso CT. Results of surgery for extratemporal partial epilepsy that began in childhood. Arch Neurol 1991;48(2):133–140 87. Moosa ANV, Gupta A, Jehi L, et al. Longitudinal seizure outcome and prognostic predictors after hemispherectomy in 170 children. Neurology 2013;80(3):253–260 88. Basheer SN, Connolly MB, Lautzenhiser A, Sherman EM, ­Hendson G, Steinbok P. Hemispheric surgery in children with refractory epilepsy: seizure outcome, complications, and ­adaptive function. Epilepsia 2007;48(1):133–140

77

78

I  Introduction to Epilepsy in Children 89. Jonas R, Nguyen S, Hu B, et al. Cerebral hemispherectomy: hospital course, seizure, developmental, language, and motor outcomes. Neurology 2004;62(10):1712–1721

93. Sabaz M, Lawson JA, Cairns DR, et al. The impact of epilepsy­ surgery on quality of life in children. Neurology 2006;66 (4):557–561

90. Kossoff EH, Vining EP, Pillas DJ, et al. Hemispherectomy for intractable unihemispheric epilepsy etiology vs outcome. Neurology 2003;61(7):887–890

94. van Empelen R, Jennekens-Schinkel A, van Rijen PC, Helders PJ, van Nieuwenhuizen O. Health-related quality of life and self-perceived competence of children assessed before and up to two years after epilepsy surgery. Epilepsia 2005;46(2): 258–271

91. Delalande O, Bulteau C, Dellatolas G, et al. Vertical parasagittal hemispherotomy: surgical procedures and clinical long-term outcomes in a population of 83 children. Neurosurgery 2007; 60(2, Suppl 1):ONS19–ONS32, discussion ONS32 92. Gleissner U, Clusmann H, Sassen R, Elger CE, Helmstaedter C. Postsurgical outcome in pediatric patients with epilepsy: a comparison of patients with intellectual disabilities, ­subaverage intelligence, and average-range intelligence. Epilepsia 2006;47(2):406–414

95. Lah S. Neuropsychological outcome following focal cortical removal for intractable epilepsy in children. Epilepsy Behav 2004;5(6):804–817 96. Szabó CA, Wyllie E, Stanford LD, et al. Neuropsychological effect of temporal lobe resection in preadolescent children with epilepsy. Epilepsia 1998;39(8):814–819

II

Part II Preoperative Assessment

IIa.  Preoperative Clinical and ­Neuropsychological Assessment

  11 Intractable Epilepsy in Children and Selection of Surgical Candidates 81

  12 Clinical Semiology in Preoperative Assessment

92

  13 Preoperative Neuropsychological and Cognitive Assessment

103

IIb.  Preoperative ­Electrophysiological Assessment   14 Electroencephalography and Noninvasive Electrophysiological ­Assessment

110

  15 Invasive Electrophysiological Monitoring

127

  16 Extra- and Intraoperative Electrocortical Stimulation

134

  17  Stereoelectroencephalography 

145

 18  Magnetoencephalography

153

IIc.  Preoperative ­Neuroimaging

  19 Structural Brain Imaging in Pediatric ­Epilepsy 161

  20 Functional Magnetic Resonance Imaging in Pediatric Epilepsy Surgery

171

79

II   21 Application of Positron Emission Tomography and Single-Photon Emission Computed Tomography in Pediatric Epilepsy Surgery 180   22 Multimodality Imaging and Coregistration

194

I

11

  Intractable Epilepsy in Children and Selection of Surgical Candidates Çiğdem Inan Akman and James J. Riviello Jr.

Summary Epilepsy surgery is acknowledged as a safe and effective treatment modality for intractable epilepsy in children and the time to surgery from epilepsy onset is critical for postsurgical outcome and cognitive functioning of these children. Various surgical procedures are available, depending upon ­certain characteristics: etiology, location of the epileptic focus, and the function of the cortex in which the epileptic focus is located. The selection of a surgical candidate starts with an exact description of the clinical manifestations of the seizure, called seizure semiology. This is done by the history, followed by the general physical and neurological examinations, basic and computerized neuropsychological testing, electroencephalography (EEG), magnetoencephalography (MEG), structural (magnetic resonance imaging or MRI), functional neuroimaging [single-photon emission computed tomography (SPECT), positron emission tomography (PET) scans] and a neuropsychological examination. The careful review of the test results is necessary for the selection of surgical candidates. If the information gathered during the presurgical evaluation is ­ congruent showing a single epileptic focus located in non-­ eloquent cortex adjacent to a structural lesion, then no further functional imaging or invasive monitoring is required, and the patient may proceed to surgery. If the focus is not clearly identified or could be within or overlapping with eloquent cortex, additional imaging and neurophysiology modalities may be required for localization. Keywords:  epilepsy, children, presurgical evaluation

„„ Introduction The incidence of new onset epilepsy in children in the general population is about 5 cases per 10,000 per year.1,​2 Approximately 23 to 33% of children with epilepsy develop intractable epilepsy.3,​4 The definition of intractable epilepsy is wide ranging and has been arbitrary. The International League Against Epilepsy (ILAE) subcommission defines intractable epilepsy as ongoing seizures despite adequate trials of two antiepileptic drugs (AEDs). Epilepsy surgery is now a recognized treatment for medical refractory epilepsy in all pediatric age groups. Various surgical procedures are available, depending upon certain characteristics: etiology, location of the epileptic focus, and the function

of the cortex in which the epileptic focus is located. It is of utmost importance to consider both the control of seizures and the quality of life after epilepsy surgery. If the epileptic focus is located in a cortical area with a vital neurological function (typically language, motor, primary sensory, or memory, referred to as eloquent cortex), and resection sacrifices this function, then the results will cause an unacceptable compromise in quality of life despite seizure freedom. The best outcome, complete seizure freedom without a deficit, is possible when a single epileptic focus exists in non-eloquent cortex that can undergo complete resection. The identification of the epileptic focus and function of its underlying cortex require data from multiple modalities: clinical, neurophysiological, and neuroanatomical. Modalities to identify the epileptic focus include the following: yy Clinical: ◦◦ Semiology. ◦◦ Physical and neurological examination. yy Neurophysiological: ◦◦ EEG (interictal). ◦◦ EEG (ictal). ◦◦ MEG (source analysis). yy Structural neuroimaging: ◦◦ MRI. ◦◦ MRI: diffusion tensor imaging (DTI). ◦◦ MRI: diffusion-weighted image (DWI). ◦◦ Magnetic resonance spectroscopy (MRS). ◦◦ CT scan. yy Functional neuroimaging: ◦◦ SPECT (interictal/ictal). ◦◦ PET (interictal). ◦◦ MEG (with evoked potentials). ◦◦ Functional MRI. ◦◦ Resting state functional MRI (interictal). yy Additional invasive tests: ◦◦ Intracarotid amobarbital procedure (IAP, Wada test). ◦◦ Invasive monitoring, electrocorticography, cortical stimulation. ◦◦ Evoked potentials during invasive monitoring. –– Somatosensory. –– Visual-evoked potentials. The commission on neurosurgery of the ILAE developed standards for epilepsy surgery. The ILAE formed the pediatric epilepsy surgery subcommission in 1998. The goal was to establish the standards and guidelines for epilepsy surgery in the

82

IIa  Preoperative Clinical and Neuropsychological Assessment pediatric age group. A survey was conducted among 20 epilepsy centers to examine the international practice for pediatric epilepsy surgery. Among 543 children (age < 18 years) who underwent epilepsy surgery, more than half (60%) had seizure onset before 2 years of age. Despite a seizure onset at a young age, only 29% had a short epilepsy duration (≤ 2 years) prior to surgery.5 Epilepsy surgery is acknowledged as a safe and effective treatment modality for intractable epilepsy in children and the time to surgery from epilepsy onset is critical for postsurgical outcome and cognitive functioning of these children. Despite a steady increase in the number of children that have had epilepsy surgery in last decade, the surgical treatment of intractable epilepsy in children is still underutilized. A community based study in Connecticut reveals that 25% of children with newly diagnosed epilepsy develop refractory epilepsy, but only 50% are referred for a comprehensive evaluation and eventually only 16% undergo any surgical procedure.6 A community-based cohort of children with newly diagnosed epilepsy identified several clinical factors for drug ­resistant epilepsy. When the underlying etiology is considered, children with nonidiopathic epilepsy syndromes are at higher risk to develop intractable epilepsy compared to children with idiopathic syndromes (32.9 vs. 9.3%).7 Moreover, an abnormal MRI is an independent risk factor for a drug resistant course in children. Abnormal MRI increases the risk for intractable epilepsy significantly in nonidiopathic epilepsy compared to ­idiopathic epilepsy (54.5 vs. 0%).6

„„ When to Refer Children for Epilepsy Surgery? Despite advances in epilepsy care, the precise definition of intractable epilepsy has remained elusive (Table 11.1).1,​8,​9,​10,​11 In order to address this concern, the ILAE recently appointed a Task Force under the Commission of Therapeutic Strategies to ­formulate a proposal for a consensus how to define intractable Table 11.1  Definitions of intractable epilepsy

epilepsy. In a prospective community based study of 613 children with epilepsy, 13% were considered refractory. Despite ­early remission periods in two-third of children, ­catastrophic epilepsy syndromes remain as a risk factor including “­cryptogenic” or “symptomatic” generalized epilepsy ­syndromes.12 Initial high seizure frequency, epilepsy onset at young age, and focal EEG slowing are the other factors for a drug resistant course. Similar to adults, children with drug resistant epilepsy, stereotyped or lateralized seizures, the presence of a potentially resectable epileptogenic lesion on MRI study should be referred for surgical epilepsy.13 Moreover, specific epilepsy etiologies require ­attention in children because of a higher risk for an intractable course including focal cortical dysplasia, tuberous sclerosis complex, p ­olymicrogyria, hypothalamic hamartoma, hemispheric ­syndromes (hemimegalencephaly), Sturge–Weber syndrome, Rasmussen’s encephalitis and Landau–Kleffner syndrome.13,​14 A number of barriers exist for an epilepsy surgery evaluation. A national survey conducted among neurologists identified a knowledge gap for the definition of intractable epilepsy and indications for epilepsy surgery. Only 51% of participants correctly identified the definition of “intractable” epilepsy and only 54% reported the need to refer the patient for an epilepsy surgery evaluation even once intractable epilepsy is ­diagnosed.15 Reasons include physician reluctance to discuss a surgical evaluation and treatment with caregivers, partly due to an unwillingness to introduce the idea of “brain surgery,” optimism for spontaneous remission, and a desire to give another chance to medication trials.16 The access to specialized care is another factor delaying referral, especially for individuals living in rural locations with limited health care resources.17 These inequalities also exist for low income populations, and ethnic and racial minorities that also have limited access to specialized care.18 Children with public insurance often face delay in presurgical evaluation.19 Other than insurance coverage, clinical severities of epilepsy and seizure frequency are the other contributing factors for shorter time to epilepsy surgery evaluation and surgical treatment in children. For example, children with active infantile spasms and daily seizures are referred for specialized care.

Source

Definition

Subcommission ILAE5

Failure of either two or three appropriate AEDs, disabling seizure side effects, or disabling AED side effects

„„ Process of Evaluation for Selection of Surgical Candidates

Connecticut12

Failure of more than two appropriate firstline AEDs, with an average of more than one seizure per month over 18 months, and not seizure free for more than 3 consecutive months in this time interval

The selection of appropriate treatment is determined by the presurgical evaluation. This is divided into three phases: phase 1, the noninvasive presurgical evaluation; phase 2, invasive monitoring; and phase 3, the surgical resection. All three phases may not be needed in every patient. There are three major aims:

Halifax/Canada13

Two or more seizures in each 2-month period during the last year of follow-up, despite treatment with at least three AEDs as monotherapy or polytherapy

Holland14

Failure to achieve more than 3 months seizure freedom and an epileptiform EEG at 6 months after diagnosis

Philadelphia15

Persistence of any seizures between 18 and 24 months after onset epilepsy and despite at least two maximally tolerated AEDs

Abbreviations: AED, antiepileptic drug; EEG, electroencephalography; ILAE, International League Against Epilepsy.

1. Lateralize and localize the epileptic focus. 2. Determine the function of the cortex with the epileptic focus (brain mapping). 3. Determine which surgical procedure has the greatest chance of controlling seizures without causing a neurological deficit. The selection of a surgical candidate starts with an exact description of the clinical manifestations of the seizure, called seizure semiology. This is done by the history, followed by the general physical and neurological examinations, basic and computerized neuropsychological testing, EEG and MEG, structural (MRI and functional neuroimaging [SPECT and PET scans] and a

11  Intractable Epilepsy in Children and Selection of Surgical Candidates Table 11.2  Cortical zones

Epileptogenic zone

Cortical area indispensible for seizure generation Epileptogenic zone may include portions or all of the following zones

Functional deficit zone

Cortical region with abnormal function in the interictal period Defined by neurologic examination, ­neuropsychologic examination, EEG, and ­functional ­neuroimaging

Irritative zone

Cortical area generating interictal spikes and sharp waves

Symptomatogenic zone

Cortical area producing the ictal symptoms when activated Primary or secondarily activated area from propagation of an epileptic discharge

Ictal (seizure) onset zone

Cortical area from which the seizure is actually generated Silent, if originates from a silent cortical area

Epileptogenic lesion

Neuroradiologic lesion causing epilepsy Important in the presurgical evaluation, but not all lesions are necessarily the lesion causing the refractory seizures (such as encephalomalacia)

Eloquent cortex

Cortex related to a given function For epilepsy surgery, typically refers to primary motor, primary sensory, language, or memory functions The term “silent cortex” is a misnomer: it really means that its function is unknown, possibly because correct paradigm was not used for testing

Abbreviation: EEG, electroencephalography.

neuropsychological examination. The results are then carefully analyzed to determine if there is evidence of focal, multifocal, or diffuse neurologic dysfunction. Further invasive monitoring may be required, depending on the specifics of each case. Although noninvasive diagnostic techniques are used to map specific cortical functions (speech, vision, motor, and sensory functions) using functional MRI, EEG-guided functional MRI, and MEG, the IAP (Wada test) remains the gold standard for language mapping in children. Invasive EEG monitoring may be needed not only to identify the exact location of seizure onset but also to map eloquent cortex. The seizure focus itself can consist of several cortical zones, which each modality examining a different cortical zone (Table 11.2).

„„ The Presurgical Clinical Evaluation The process begins with an initial outpatient evaluation. Several questions need to be addressed before any invasive procedure: 1. Does the child truly have epilepsy? Have nonepileptic seizures, vasovagal events, periodic movement disorders, and hyperekplexia been ruled out? 2. Is surgery warranted for the case? Is there a metabolic or degenerative condition or benign rolandic epilepsy? 3. What is the underlying etiology (lesional, nonlesional, channelopathies, metabolic, degenerative, tumor, infection, etc.)? 4. Is epilepsy truly refractory? Have the appropriate AEDs been used to therapeutic maximum levels? 5. Is remission still a possibility? Have all nonsurgical avenues been exhausted?

6. How prepared are the child and his family for surgery? Have they been counseled on the possible side effects of psychological aspects? We find it best to review this in an outpatient visit before the admission so that the family knows what to expect and we know how to best plan the phase 1 evaluation based upon the child’s individual needs and expectations. The epileptologist must first determine whether the initial diagnosis of epilepsy was correct. Many nonepileptic paroxysmal events are easily mistaken for epilepsy, treated with AEDs, and ultimately referred for presurgical evaluation. In a study of 223 children referred to a tertiary epilepsy center, 87 (39%) did not have epilepsy.20 We have seen children referred for refractory epilepsy with other diagnosis, especially nonepileptic seizures, vasovagal events, including stretch syncope, periodic moment disorders, and hyperekplexia. Capturing the habitual seizure on long-term EEG monitoring excludes these conditions. Appropriate AED treatment for the seizure type or epilepsy syndrome is needed before considering surgery. If an underlying metabolic or degenerative condition is responsible for the seizures, respective surgery may not be appropriate. Epilepsy syndromes are divided into benign or malignant. This distinction refers to the ultimate course of the actual seizures in additional to the developmental outcome. The term catastrophic epilepsy applies to early onset epilepsy syndromes, in which the outcome is poor unless seizures can be controlled, commonly occurring when a structural lesion causes refractory epilepsy. Alternatively, children with benign syndromes, such as Rolandic epilepsy, may have seizures that, although difficult to control, will ultimately remit.21

Seizure Semiology and Neurological Examination The seizure semiology and neurological examination are the first steps in the process of lateralizing and localizing the symptomatogenic and functional deficit zones, with the ultimate purpose of answering the main question, where is the ictal onset? Therefore, a semiology classification system is advised to better aid this process (Table 11.3).22,​23 The clinical history, the description of the seizure type, the age of seizure onset, the history of neurodevelopment, and EEG and neuroimaging findings provide the information needed to characterize pediatric epilepsy, including the epilepsy syndrome and underlying etiology. Clinical description (semiology) and sequential evolution of seizures are of the utmost importance to localize the ictal onset zone combined with the ictal EEG findings. However, semiology may not always identify the ictal onset zone, as in some cases, the ictal onset zone may occur in a silent area with clinical manifestations occurring only when the seizure has spread, which is called the symptomatogenic zone. In children, the seizure semiology may change as a function of age. Focal seizures may present with generalized features.24 Behavioral arrest mimicking absence seizures, bilateral tonic posturing of extremities, bilateral clonic activity, and spasms are the clinical features described in infants diagnosed with temporal and extratemporal epilepsy.25,​26 Secondary generalization is unlikely to occur in infants and young children.25 On the other hand, complex clinical features such as confusion, auras, automatism, and versive head movements are common in older children, after 6 years of age.22,​27,​28

83

84

IIa  Preoperative Clinical and Neuropsychological Assessment Table 11.3  Seizure semiology by lobe

Temporal

Frontal

Parietal

Occipital

Mesial temporal Hippocampus/Amygdala: rising epigastric sensation; nausea; fear, panic; autonomic symptoms; experiential symptoms, such as déjà vu feeling

Lateral frontal Contralateral tonic head and eye deviation; speech arrest

Positive or negative sensory phenomenon, pain, numbness, electric feeling; spreading paresthesias (Jacksonian March); desire to move, feeling of moving or loss of awareness of a body part; palinopsia

Simple or complex visual phenomenon; sparks, flashes, phosphenes; scotoma, hemianopia, amaurosis; perceptual illusions, sensation oscillation; clonic or tonic contraversion of head and eyes; palpebral jerks, forced eyelid closure

Posterior lateral temporal Auditory hallucinations or illusions; vertigo, visual misperceptions; language disturbance

Cingulate gyrus Absence seizures, with CPS and complex motor gestures, GTC seizures; autonomic signs; mood changes Frontopolar Drop attacks, tonic seizures, adversive seizures (head and eyes); secondarily GTC seizures; forced ideation Orbitofrontal CPS with motor automatisms, olfactory hallucinations and illusions, autonomic signs Opercular Simple partial seizures with clonic facial movements; mastication, salivation, swallowing; speech arrest, fear, epigastric aura, gustatory hallucinations, autonomic signs Motor (central lobe) Simple partial seizures with ­clonic movements; epilepsia partialis ­continua

Abbreviations: CPS, complex partial seizures; GTC, generalized tonic clonic.

Physical Examination Physical examination offers clues for certain epilepsy syndromes accompanied by neurocutaneus syndromes such as tuberous sclerosis, neurofibromatosis, epidermal nevus, incontinentia pigmenti. Unilateral weakness or a difference in the size of the extremities or face may suggest the acquired or congenital structural abnormality in the contralateral brain hemisphere, including hemimegalencephaly, hemiatrophy, or cyctic encephalomalacia. Other focal findings in the neurological examination are similarly important to identify focal cortical dysfunction in the presence of motor weakness, unilateral impaired fine motor coordination, apraxia, and visual field defect. Hand preference should be evaluated and confirmed during the examination. The majority of children demonstrate a hand preference by 2 years of age; however, persistent hand preference before 1 year of age should raise a red flag and require careful examination for weakness of contralateral upper extremity.

„„ Presurgical Evaluation Techniques and Modalities The ILAE survey of 20 pediatric epilepsy centers confirms that video-EEG and brain MRI are universal diagnostic

tests for the presurgical evaluation whereas other imaging modalities are applied unevenly. PET and SPECT scans are the functional imaging modalities used in 17 (85%) and 16 (80%) centers, respectively, in this survey. For language lateralization, functional MRI is used more often, in 14 (70%) centers compared to the Wada test, used in only 10 (50%) centers. MEG continues to have a limited role for pediatric presurgical evaluation, and is used in only a few centers, only 7 (35%).5 The combination of an ictal SPECT and PET scan is used in 14 centers. In lieu of this study, the minimum presurgical evaluation for pediatric epilepsy surgery should include the following diagnostic modalities: (1) neurophysiological studies (interictal EEG with sleep recording and ictal video-EEG); (2) high resolution anatomical imaging with a specific epilepsy protocol; and (3) neuropsychological assessment. Each modality serves to identify areas of cortical dysfunction, and some map eloquent cortex. In an ideal epilepsy surgery patient, one localized cortical area contains the epileptic focus and has no eloquent cortex. In the ideal situation, the data for cortical dysfunction would all be congruent to the same cortical area with no mapping modality showing the focus within eloquent cortex. Congruent data imply neurological dysfunction in only one area, which predicts a better chance for seizure control. Alternatively, incongruent data such as multiple dysfunctional areas, possibly with multifocal seizure onset, imply a lower chance of achieving complete seizure control.11,​29

11  Intractable Epilepsy in Children and Selection of Surgical Candidates

Noninvasive Electrophysiology Monitoring The arrival of video-EEG monitoring into the clinical practice has enhanced our ability to diagnose seizures and distinguish them from other clinical events. Particularly, paroxysmal events other than seizures such as dystonia, chorea, and stereotype are seen often in young children and children with disabilities. A number of epilepsy syndromes also presents with multiple seizure types in children such as Lennox–Gestaut ­syndrome. Electrographic seizures without clinical manifestation was ­ reported in 20% of pediatric patients diagnosed with focal epilepsy.30 Electrographic seizures were frequent in infants and young age groups, with presence of neurological disability and high seizure burden. In summary, video-EEG monitoring has indispensable role and appropriate EEG recording tool to characterize and differentiate clinical and EEG features of the clinical events to describe “symptomatogenic zone” in pediatric age group. Self-report of seizures can be misleading in adults living with epilepsy. Seizures were underreported in 23% of adult patients presenting with focal epilepsy. The under-report was attributed to older age at the onset of epilepsy, presence of independent bihemipsheric epileptiform activity, bilateral ­ epileptogenic lesion and temporal lobe epilepsy.31,​32 A study describing the role of ambulatory EEG monitoring in outpatient setting also underscored the fact that 38% of the clinical seizures were not ­reported by the patient.33 In pediatric age group, a review of 1,244 seizures in 78 patients admitted to video-EEG monitoring revealed that 50% of seizures were inaccurately reported in children, particularly in infants and young children.32 Motor seizures were reported more often compared to absence or complex partial seizures. Seizure under-report was attributed to several factors: (1) inability to identify seizures because of subtle clinical features; (2) absence of witness for seizure report; and (3) seizure without any clinical features, only recognized by EEG recording. The authors also noted that despite of co-sleeping with a parent, or sleeping in the same room during video-EEG monitoring, a number of seizures were missed by parents. Although the majority of seizures (72%) were reported in first 2 days of video-EEG monitoring,28 a prolonged EEG monitoring beyond 3 days may be required for some children for accurate diagnosis.34 Noninvasive monitoring is also essential to define “irritative zone” based on the localization and distribution of spikes and sharp waves on the interictal EEG recording. The presence of focal slowing is complimentary to the definition of “functional deficit zone.” The ictal EEG has been considered the sine qua non for localizing the “ictal onset region.” Other than video-EEG monitoring, MEG imaging and dipole modeling is another neurophysiological tool that identifies the magnetic field generated by the epileptic focus. EEG and MEG are complementary: EEG best identifies a vertical dipole and MEG best identifies a horizontal dipole. The MEG dipoles can be superimposed onto MRI, which is useful for localization. The role of EEG monitoring and other methods for noninvasive electrophysiology monitoring are discussed with more detail in other related chapters of this book.

Structural Neuroimaging The goal of epilepsy surgery in all age groups is the complete removal of the epileptogenic cortex without causing a

­ ermanent neurological deficit. Outcome of epilepsy surgery p depends on the underlying etiology and extend of cortical involvement. Favorable outcome, therefore, is correlated if an epileptogenic lesion is present on MRI and location of MRI findings is concordant with ictal EEG findings. A structural lesion is considered one of the best indicators of the epileptogenic zone because specific lesions have a high association with epilepsy. However, a structural lesion may not always contain the focus, or dual pathology may be present. Brain MRI is the imaging method of choice to identify lesions in children and adults with medically intractable epilepsy.10,​35 There are many epileptogenic lesions, as detailed below: yy Hemispheric lesions: ◦◦ Acquired: –– Rasmussen’s encephalitis. –– Cystic encephalomalacia. ◦◦ Congenital: –– Hemimegelancephaly. –– Hemispheric polymicrogyria. –– Hemispheric atrophy. yy Focal lesions: ◦◦ Focal cortical dysplasia. ◦◦ Glioneuronal tumors: –– Ganglioglioma. –– Dysembryoplastic neuroepithelial tumor.

yy yy

yy yy

yy

◦◦ Glial tumors: –– Oligodendroglioma. ◦◦ Focal encephalomalacia. ◦◦ Porencephalic cyst. Mesial temporal sclerosis. Vascular abnormalities: ◦◦ Cavernous malformations. ◦◦ Arteriovenous malformations. ◦◦ Sturge–Weber Syndrome. Encephalitis. Neurocutaneous syndromes: ◦◦ Tuberous sclerosis complex. ◦◦ Neurofibromatosis. ◦◦ Epidermal nevus. Others: ◦◦ Neurocysticercosis. ◦◦ Calcifications.

Focal cortical dysplasia (FCD) is the most common lesion found in children undergoing epilepsy surgery for intractable ­epilepsy.5 Contrary to adult patients with refractory epilepsy undergoing epilepsy surgery, mesial temporal sclerosis (MTS) is reported in only 13% of children who had temporal lobectomy, and often coexists with another lesion (dual-pathology) such as FCD, encephalomalacia, porencephaly, or low-grade tumors (Fig. 11.1). Therefore, the presence of MTS on brain MRI can represent an epileptogenic lesion however dual pathology, not limited to mesial temporal sclerosis, but also to neocortical lesions such as malformations, encephalomalacia, and tumors. The imaging findings of FCD can be subtle (Fig. 11.2). The degree of the histopathological changes described in FCD often correlate with the degree of visibility of FCD on a ­ natomical images. Brain MRI may not demonstrate any abnormality in 40% of FCD type 1 and 10% of FCD type 2. Therefore, ­specific

85

86

IIa  Preoperative Clinical and Neuropsychological Assessment

Fig. 11.1  MRI findings of epileptogenic lesions: (a) multiple tubers (arrow) seen in bilateral hemipsheres seen in axial view of a patient diagnosed with tuberous sclerosis complex; (b) polycystic encephalomalacia of right hemipshere in a child presenting with left-sided hemipharesis and focal epilepsy; (c) dysembryoplastic neuroepithelial tumor located in the parietal cortex (arrow) on axial view; (d) nodular heterotopia (arrow) on a child presenting with focal epilepsy and mild learning dissability; (e) bilateral mesial temporal sclerosis in a child with previous history of limbic encephalitis associted with antiGAD65 autoantibodies.

Fig. 11.2  Arrows point the subtle abnormalities detected on brain MRI of the patient diagnosed with focal cortical dysplasia on surgical substrates. (a) Transmantle sign of FCD type II, seen in coronal view of fluid attenuation inversion recovery (FLAIR) images, increase fluid attenuation inversion recovery (FLAIR) signal seen in frontal white matter extending from the bottom of gyrus to the tip of the lateral ventricular. (b) Focal cortical atrophy and increased FLAIR signals in superior parietal cortex. (c) Blurred interface between grey and white matter in the orbitofrontal gyrus on T1 sequences in a child diagnosed with drug resistant focal epilepsy.

11  Intractable Epilepsy in Children and Selection of Surgical Candidates acquisition protocols are needed to increase the sensitivi­ ty to visualize subtle cortical malformations. T1-weighted volumetric acquisition with isotropic, 1 cubic mm voxels increases special resolution.36,​37 Wiggins et al demonstrated that application of higher field magnets equal or more than 3T (a 32-channel head coil) improves the image signal-to-noise 3.5-fold in the cortex and provides images with high resolution compared to the eight-channel commercial head coil.38 Anatomical imaging with 3T brain MRI generates images with a well-defined contrast between white and grey matter junctions in order to detect FCD. If 1.5T MRI fails to identify FCD, then MRI should be repeated using specific epilepsy protocols at 3T. The main MRI findings of FCD include: • T1-weighted sequence: ◦◦ Thickening of cortex. ◦◦ Blurring grey–white matter boundary. ◦◦ Decreased signal in subjacent white matter. yy T2-weighted sequence: ◦◦ Grey matter hyperintensity. yy T2 fluid attenuation inversion recovery (FLAIR): ◦◦ Subcortical white matter hyperintensity. ◦◦ Transmantle sign (funnel-shaped hyperintensity extending from the cortex to the sub-superolateral margin of the lateral ventricle). For children older than 2 years of age, standard MRI protocol for FCD should include the following: (1) Sagittal T1-­weighted images, 5 mm/thickness; (2) Axial FLAIR and T2-weighted images, ≤ 5 mm thickness; (3) Coronal FLAIR ≤ 5 mm slices; (4) Coronal double echo sequence with 5 mm slices; (5) Axial or coronal T2 gradient echo images; and (6) Coronal–sagittal T1-weighted three-dimensional volume acquisition with 1 to 1.5 mm partitions.39 Certain MRI sequences, slice orientation and thickness are better for identifying specific epileptogenic lesions. Epileptogenic tumors are seen best on multiplanar MRI including axial and coronal T1-, T2-weighted images and FLAIR. Gradient-echo imaging is superior to detect cavernous angioma. Mesial temporal sclerosis can be visualized with increased signal intensity on FLAIR and T2 with coronal perpendicular orientation to the long axis of the hippocampus. For cortical dysplasia, T2-weighted FLAIR imaging is also helpful to detect blurring of the grey–white junctions.

FLAIR images increase the sensitivity to detect hippocampal sclerosis, tumors and gliosis. Therefore, FLAIR is a typical screening test and high parenchymal signals and pulsation artifacts can lead to false-positive results and therefore confirmation is needed using T2-weighted images. Before age of 6 months, however, T2-weighted images show brighter white matter compare to the older children because of incomplete myelination. Therefore blurring of grey-white matter interface is expected to be darker (hypointense) on T2 weighted images in children younger than 6 months of age. A phased array surface coil study performed at 3T (3T PA MRI) is reported superior to 1.5T MRI to detect and define the subtle lesions.23 CT scan has limited role in presurgical evaluation, however remains superior to MRI to visualize calcified lesions.

Functional Neuroimaging Functional neuroimaging refers to nuclear medicine studies using either metabolic or blood flow measurements to identify dysfunctional cortex, with the data presented in an anatomical view.40,​41 Functional imaging is especially helpful when MRI is negative (nonlesional). This includes SPECT, both ictal and interictal (Fig. 11.3), PET scan, which is usually interictal (Fig. 11.4 and Fig. 11.5), MRI-PET (Fig. 11.6), and functional MRI. Localization based on functional imaging presumes that abnormal cortical areas generating seizures have abnormal ­perfusion or metabolism and these may also identify the functional deficit zone. The epileptogenic zone and the functional deficit zone may not always be located in the same area. Functional MRI and MEG may also identify eloquent cortex. Each modality has advantages and disadvantages, with none the ­single best modality for localizing the epileptic focus or functional cortex. MRS is a noninvasive measurement of chemical substances especially N-acetylaspartate, choline, creatinine, and lactic acid, which may be abnormal within the epileptogenic zone. DTI, measuring the diffusion of water molecules, identifies white matter tracts, which is helpful in surgical planning and DWIs may identify the epileptogenic zone, if ­performed rapidly after a seizure. EEG-guided functional MRI identifies cortical areas activated by epileptiform activity, although in clinical practice obtaining an ictal study is difficult for most seizure types, except absence epilepsy, because of a ­rtifact.

Fig. 11.3  (a, b) A 13-year-old patient who has frontal lobe epilepsy and a normal brain MRI. Ictal and interictal SPECT scan subtraction using SPM-5 method that was coregistrated on brain MRI. Ictal onset was mapped to the superior frontal cortex.

87

88

IIa  Preoperative Clinical and Neuropsychological Assessment

Fig. 11.4  (a, b) Normal brain MRI of a 17-year-old man with temporal lobe epilepsy. Focal hypometabolism of left temporal cortex (arrow) was seen in resting state despite normal anatomical imaging.

Fig. 11.5  A 2-year-old patient with focal epilepsy and medically refractory infantile spasms. (a) Brain MRI demonstrates diffuse cortical atrophy and focal cortical dysplasia in the right temporal lobe involving mesial structures (arrow). (b) PET images of the same patient demonstrating an increased glucose uptake despite the absence of seizures during the PET imaging in the area of focal cortical dysplasia.

Fig. 11.6  Coregistration of MRI and PET images in a 16-year-old who presented with drug resistant focal epilepsy. (a) Brain MRI T1 sequence demonstrates the enlarged superior temporal gyrus (X) with a thin cortical layer. (b) PET–MRI depicts the focal cortical lesion with improved spatial resolution in the right anterior neocortex.

­Task-based functional mapping techniques are particularly difficult in patients at young ages and developmental disabilities. Resting state functional mapping is reported as a promising tool in children with limited abilities to perform age-appropriate functions.42

Neuropsychological Evaluation The neuropsychological examination may identify functional deficit zones, and determine if there is focal or multifocal cortical dysfunction. Similar to other modalities, the epileptic focus may be located in this dysfunctional cortex. There are also specific neuropsychological profiles for the various focal epilepsies: frontal, temporal, parietal, or occipital.43,​44 The neuropsychological

evaluation also identifies the individual intelligence strengths and weaknesses so that an educational plan can be developed to o ­ ptimize education and help compensate for deficits, and may predict the risk of postoperative deficits, which is especially important in determining the risk–benefit ratio for the surgery. Neuropsychological testing done during the Wada test helps to determine cerebral dominance for language, memory, and visuospatial functions. Language or verbal memory deficits suggest dominant hemisphere dysfunction, visuospatial memory deficits suggest nondominant temporal dysfunction, and deficits in both suggest bitemporal disease. The Wada test is done by a team from neurology, neuropsychology, and neurophysiology. It may be difficult to perform a successful Wada test when the child has a low IQ, is younger than 10 years, or has seizures

11  Intractable Epilepsy in Children and Selection of Surgical Candidates arising from the dominant hemisphere.45 Another study examined the clinical and initial EEG features recorded at the time of the Wada testing in 30 children referred for language lateralization.46 Language function was lateralized in 23 patients, whereas 7 patients could not endure the procedure. Confusion, agitation, and drowsy state were the reasons for failure to complete procedure. Other than young age at the time of IAP, none of the other clinical features correlated with the clinical outcome of IAP. High amplitude frontal slowing soon after amobarbital injection was observed in patients failed to complete the Wada test. Epilepsy causes many stresses on the child and his family. These are especially magnified when epilepsy is refractory and may be further exacerbated by the presurgical evaluation. If intracranial EEG monitoring is required, cooperation of parents and a child is essential in order to accomplish the goal for EEG monitoring safely. Therefore, neuropsychological evaluation is required to detect potential psychological and psychiatric problems and decide whether intervention is needed beforehand.

„„ Epilepsy Surgery Conference and Invasive Monitoring Information gathered during the phase 1 presurgical evaluation is presented at the Epilepsy Surgical Conference, attended by the epileptologists, epilepsy surgeons, neuropsychologists, neuroradiologists, electrodiagnostic technologists, and epilepsy and neurology trainees. It is best to obtain consensus regarding the appropriate surgical procedure. If the data are congruent showing a single epileptic focus located in non-eloquent cortex adjacent to a structural lesion, then no further functional imaging or invasive monitoring is required, and the patient may proceed to surgery. One study evaluated a “streamlined evaluation” such as the need for an ictal EEG, and suggested that other studies may not be needed if MRI shows an epileptogenic lesion; however, this is not currently how epilepsy surgery is practiced.47 If the focus is not clearly identified or could be within eloquent cortex, then functional MRI or MEG can be done, with these results then presented. If seizure onset or eloquent cortex is not clearly localized, the invasive monitoring may be indicated. Usual indications for invasive EEG monitoring include the following: 1. Localizing a focal seizure onset with normal or nonlocalizing imaging (nonlesional). 2. Defining seizure onset around a potential epileptogenic lesion. 3. Gathering more data in cases with noncongruent noninvasive test results. 4. Assessing multiple lesions or multifocal interictal epileptiform activity. 5. Identifying the location of eloquent cortex.

„„ Surgical Treatment and Outcome ILEA survey conducted among the pediatric epilepsy centers suggests that resective surgery is performed more often in

children compared to palliative surgery procedures.5 Among children undergoing resective surgery (n = 440), a focal ­ resection was performed in 261 (48%), hemispherectomy in 86 (15.8%), and multilobar resections in 70 (12.9%). Of the lobar/ focal resections, a temporal lobe resection was performed more frequently, in 126 (23.2%) compared to frontal lobe in 95 (17.5%), a parietal lobe resection in 15 (2.8%). In the presence of a lesion such as MTS, vascular malformation, low-grade tumors, lesionectomy, and/or lobectomy are the preferred surgical approach. In that case, complete removal of the lesion and perilesional region is vital to achieve a favorable outcome. In addition to lesionectomy and lobectomy, hemispherotomy or functional disconnection of the hemispheres, functional and anatomical hemispherectomy are the other surgical modalities recommended for children diagnosed with an epileptogenic lesion extending to the entire hemisphere. Children diagnosed with Rasmussen encephalitis, hemimegalencephaly or Sturge–Weber syndrome are a few examples of this abnormality referred for hemispherotomy. Palliative surgical procedures include multifocal cortical resection, hemispherectomy, corpus callosotomy, and multiple subpial transection (MST). These procedures are considered with either multifocal or generalized seizures. Multifocal cortical resection or hemispherectomy are considered when the epileptogenic zone is primarily multifocal but unilateral, corpus callosotomy is done for either bilateral or generalized seizure onset, and MST is considered when epileptic focus is within eloquent cortex. Neurostimulation techniques, such as vagus nerve stimulation, or responsive neurostimulation, can be performed in those patients not considered ideal candidates for a focal resection, such as, patients with multifocal seizures, or patients with epileptic focus with an eloquent cortex, and when the patient or family are not interested in resective surgery. In an outcome study of 75 children from Miami Children’s Hospital, a favorable outcome was reported in 92% of patients when there was complete resection of the epileptogenic lesion compared with only 50% with an incomplete resection.48 Even in nonlesional refractory focal epilepsy, a good outcome can be achieved after complete resection of the epileptogenic zone using a multimodality approach.49

„„ Postsurgical Follow-up The postsurgical follow-up period, with routine visits, should last at least 5 years. Visits should assess seizure outcome using a recognized scale, AED use, quality of life, and the neuropsychological status. The postoperative MRI should be performed at a minimum of 3 months after surgery, especially when there is a surgical failure or complication. To consider tapering of AEDs, children should be seen regularly. We typically see these children back at 1 month, 6 months, 12 months, and 2 years, and assess them using a modified Engel Scale, we perform a complete neuropsychological examination at 1 year, and dependent upon case specifics, we consider an EEG or EEG monitoring at 6 months.50

89

90

IIa  Preoperative Clinical and Neuropsychological Assessment

References 1. Camfield PR, Camfield CS. Antiepileptic drug therapy: when is epilepsy truly intractable? Epilepsia 1996;37(Suppl 1):S60–S65 2. Olafsson E, Ludvigsson P, Gudmundsson G, Hesdorffer D, ­Kjartansson O, Hauser WA. Incidence of unprovoked seizures and epilepsy in Iceland and assessment of the epilepsy syndrome classification: a prospective study. Lancet Neurol 2005;4(10):627–634

20. Uldall P, Alving J, Hansen LK, Kibaek M, Buchholt J. The misdiagnosis of epilepsy in children admitted to a tertiary epilepsy centre with paroxysmal events. Arch Dis Child 2006;91(3):219–221 21. Ong HT, Wyllie E. Benign childhood epilepsy with centrotemporal spikes: is it always benign? Neurology 2000;54(5):1182–1185 22. Lüders H, Acharya J, Baumgartner C, et al. Semiological seizure classification. Epilepsia 1998;39(9):1006–1013

3. Berg AT, Shinnar S, Levy SR, Testa FM, Smith-Rapaport S, ­ 23. Knake S, Triantafyllou C, Wald LL, et al. 3T phased array MRI Beckerman B. Early development of intractable epilepsy in improves the presurgical evaluation in focal epilepsies: a pro­children: a prospective study. Neurology 2001;56(11):1445–1452 spective study. Neurology 2005;65(7):1026–1031 4. Berg AT, Vickrey BG, Testa FM, et al. How long does it take for 24. Brockhaus A, Elger CE. Complex partial seizures of tempoepilepsy to become intractable? A prospective investigation. Ann ral lobe origin in children of different age groups. Epilepsia Neurol 2006;60(1):73–79 1995;36(12):1173–1181 5. Harvey AS, Cross JH, Shinnar S, Mathern GW; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epilepsia 2008;49(1):146–155 6. Berg AT, Mathern GW, Bronen RA, et al. Frequency, prognosis and surgical treatment of structural abnormalities seen with magnetic resonance imaging in childhood epilepsy. Brain 2009;132 (Pt 10):2785–2797 7. Berg AT, Levy SR, Testa FM, D’Souza R. Remission of epilepsy after two drug failures in children: a prospective study. Ann Neurol 2009;65(5):510–519 8. Arts WF, Geerts AT, Brouwer OF, Boudewyn Peters AC, Stroink H, van Donselaar CA. The early prognosis of epilepsy in childhood: the prediction of a poor outcome. The Dutch study of epilepsy in childhood. Epilepsia 1999;40(6):726–734 9. Camfield PR, Camfield CS, Gordon K, Dooley JM. If a first antiepileptic drug fails to control a child’s epilepsy, what are the chances of success with the next drug? J Pediatr 1997;131(6):821–824 10. Commission on Neuroimaging of the International League Against Epilepsy. Guidelines for neuroimaging evaluation of patients with uncontrolled epilepsy considered for surgery. Epilepsia 1998;39(12):1375–1376 11. Kurian M, Spinelli L, Delavelle J, et al. Multimodality imaging for focus localization in pediatric pharmacoresistant epilepsy. Epileptic Disord 2007;9(1):20–31

25. Korff C, Nordli DR Jr. Do generalized tonic-clonic seizures in infancy exist? Neurology 2005;65(11):1750–1753 26. Nordli DR. Varying seizure semiology according to age. Handb Clin Neurol 2013;111:455–460 27. Alqadi K, Sankaraneni R, Thome U, Kotagal P. Semiology of hypermotor (hyperkinetic) seizures. Epilepsy Behav 2016;54:137–141 28. Kotagal P, Arunkumar G, hammel J, Mascha E. Complex partial seizures of frontal lobe onset statistical analysis of ictal semiology. Seizure 2003;12(5):268–281 29. Labiner DM, Weinand ME, Brainerd CJ, Ahern GL, Herring AM, Melgar MA. Prognostic value of concordant seizure focus localizing data in the selection of temporal lobectomy candidates. Neurol Res 2002;24(8):747–755 30. Akman CI, Montenegro MA, Jacob S, et al. Subclinical seizures in children diagnosed with localization-related epilepsy: clinical and EEG characteristics. Epilepsy Behav 2009;16(1):86–98 31. Heo K, Han SD, Lim SR, Kim MA, Lee BI. Patient awareness of complex partial seizures. Epilepsia 2006;47(11):1931–1935 32. Montenegro MA, Sproule D, Mandel A, et al. The frequency of non-epileptic spells in children: results of video-EEG monitoring in a tertiary care center. Seizure 2008;17(7):583–587 33. Tatum WO IV, Winters L, Gieron M, et al. Outpatient seizure identification: results of 502 patients using computer-assisted ambulatory EEG. J Clin Neurophysiol 2001;18(1):14–19

12. Ko TS, Holmes GL. EEG and clinical predictors of medically intractable childhood epilepsy. Clin Neurophysiol 1999;110(7):1245–1251

34. Asano E, Pawlak C, Shah A, et al. The diagnostic value of initial video-EEG monitoring in children—review of 1000 cases. Epilepsy Res 2005;66(1–3):129–135

13. Cross JH, Jayakar P, Nordli D, et al; International League against Epilepsy, Subcommission for Paediatric Epilepsy Surgery. Commissions of Neurosurgery and Paediatrics. Proposed criteria for referral and evaluation of children for epilepsy surgery: recommendations of the Subcommission for Pediatric Epilepsy Surgery. Epilepsia 2006;47(6):952–959

35. Commission on Neuroimaging of the International League Against Epilepsy. Recommendations for neuroimaging of patients with epilepsy. Epilepsia 1997;38(11):1255–1256 36. Barkovich AJ, Dobyns WB, Guerrini R. Malformations of cortical development and epilepsy. Cold Spring Harb Perspect Med 2015;5(5):a022392

14. Perry MS, Duchowny M. Surgical management of intractable childhood epilepsy: curative and palliative procedures. Semin Pediatr Neurol 2011;18(3):195–202

37. Barkovich AJ, Rowley HA, Andermann F. MR in partial epilepsy: value of high-resolution volumetric techniques. AJNR Am J Neuroradiol 1995;16(2):339–343

15. Roberts JI, Hrazdil C, Wiebe S, et al. Neurologists’ knowledge of and attitudes toward epilepsy surgery: a national survey. Neurology 2015;84(2):159–166

38. Wiggins GC, Triantafyllou C, Potthast A, Reykowski A, Nittka M, Wald LL. 32-channel 3 Tesla receive-only phased-array head coil with soccer-ball element geometry. Magn Reson Med 2006;56(1):216–223

16. Jehi L, Mathern GW. Who’s responsible to refer for epilepsy surgery? We all are! Neurology 2015;84(2):112–113 17. Meyer AC, Dua T, Boscardin WJ, Escarce JJ, Saxena S, Birbeck GL. Critical determinants of the epilepsy treatment gap: a cross-national analysis in resource-limited settings. Epilepsia 2012;53(12):2178–2185 18. Pestana Knight EM, Schiltz NK, Bakaki PM, Koroukian SM, Lhatoo SD, Kaiboriboon K. Increasing utilization of pediatric epilepsy surgery in the United States between 1997 and 2009. Epilepsia 2015;56(3):375–381 19. Baca CB, Vickrey BG, Vassar S, et al. Time to pediatric epilepsy surgery is related to disease severity and nonclinical factors. Neurology 2013;80(13):1231–1239

39. Woermann FG, Vollmar C. Clinical MRI in children and adults with focal epilepsy: a critical review. Epilepsy Behav 2009;15(1):40–49 40. So EL. Role of neuroimaging in the management of seizure disorders. Mayo Clin Proc 2002;77(11):1251–1264 41. Kim SK, Na DG, Byun HS, et al. Focal cortical dysplasia: comparison of MRI and FDG-PET. J Comput Assist Tomogr 2000;24(2):296–302 42. Vadivelu S, Wolf VL, Bollo RJ, Wilfong A, Curry DJ. Resting-state functional MRI in pediatric epilepsy surgery. Pediatr Neurosurg 2013;49(5):261–273

11  Intractable Epilepsy in Children and Selection of Surgical Candidates 43. Lassonde M, Sauerwein HC, Jambaqué I, Smith ML, Helmstaedter C. Neuropsychology of childhood epilepsy: pre- and postsurgical assessment. Epileptic Disord 2000;2(1):3–13

47. Patil SG, Cross JH, Kling Chong W, et al. Is streamlined evaluation of children for epilepsy surgery possible? Epilepsia 2008;49(8):1340–1347

44. Helmstaedter C. Neuropsychological aspects of epilepsy surgery. Epilepsy Behav 2004;5(Suppl 1):S45–S55

48. Paolicchi JM, Jayakar P, Dean P, et al. Predictors of outcome in pediatric epilepsy surgery. Neurology 2000;54(3):642–647

45. Hamer HM, Wyllie E, Stanford L, Mascha E, Kotagal P, Wolgamuth B. Risk factors for unsuccessful testing during the intracarotid amobarbital procedure in preadolescent children. Epilepsia 2000;41(5):554–563

49. Jayakar P, Gaillard WD, Tripathi M, Libenson MH, Mathern GW, Cross JH; Task Force for Paediatric Epilepsy Surgery, Commission for Paediatrics, and the Diagnostic Commission of the International League Against Epilepsy. Diagnostic test utilization in evaluation for resective epilepsy surgery in children. Epilepsia 2014;55(4):507–518

46. Akman CI, Micic V, Quach M, et al. Application of envelope trend to analyze early EEG changes in the frontal regions during intracarotid amobarbital procedure in children. Epilepsy Behav 2015;43:66–73

50. Lachhwani DK, Loddenkemper T, Holland KD, et al. Discontinuation of medications after successful epilepsy surgery in children. Pediatr Neurol 2008;38(5):340–344

91

12

  Clinical Semiology in Preoperative Assessment András Fogarasi

Summary Clinical semiology is an important piece of the preoperative assessment puzzle. Observing a live seizure, analyzing its details during video-electroencephalography (EEG) monitoring, or simply watching a smartphone-recorded seizure should give us many clues. Is it of an epileptic or nonepileptic origin? Is it temporal or extratemporal? Where did it start? Which parts of the brain were involved? Is there any lateralizing sign during the seizure? Is it congruent with the patient’s EEG and MRI data? There are many questions we can only answer with a proper knowledge of pediatric seizure semiology. In this chapter, we go through the most important factors of childhood seizure semiology analyzing localizing, lateralizing and age-dependent aspects, as well as the most typical semiology-related pitfalls. Two video files—a collection of childhood peri-ictal lateralizing signs and a compilation of various epileptic spasms—illustrate the semiology discussed in this chapter. Keywords:  seizure semiology, localization, lateralization, age-dependent feature, temporal, extratemporal, insular, ­multilobar, seizure onset zone, pitfalls

„„ Introduction Epilepsy is a broad category of recurrent paroxysmal episodes of brain dysfunction manifested by stereotyped alterations in behavior.1 Seizure semiology describes the clinical features of such an episode supporting its precise classification.2 Proper knowledge and frequent use of seizure semiology can greatly help in the adequate categorization of epileptic seizures and the use of their features in the localization and lateralization of the seizure onset zone. For example, a typical patient with temporal lobe epilepsy (TLE) may have an epigastric aura, followed by unresponsiveness, staring, oral and manual automatisms, contralateral dystonic posturing, and postictal confusion. In the case of a dominant hemisphere onset, the seizure might be followed by transient aphasia. A typical frontal lobe epilepsy (FLE) seizure will begin in sleep with no warning signs, followed by bilateral complex automatisms or asymmetric tonic posturing and will end quickly with immediate recovery. Occipital lobe seizures often have a visual aura, and frequently spread to the temporal lobe according to their semiology. Parietal lobe seizures are the least common, may have a sensory aura, and tend to mimic frontal lobe seizures.3 As we will see, however, childhood seizures may show many exceptions.

Since the advent of presurgical evaluation, especially longterm video-EEG monitoring, our knowledge of seizure semiology has become much more exact by getting to know the less typical semiological features. The repeated observation of a videotaped seizure can help us to notice even subtle elements (i.e., fine myoclonus, autonomic signs, etc.), which might be easily missed by lay people or even experienced epileptologists. Moreover, the exact duration of a seizure, the level of consciousness during the attack, mild postictal hemiparesis, or sensorial aphasia can be objectively judged only by video-recorded and adequately tested seizures. During presurgical evaluation, neuroimaging methods, EEG recordings, and semiological features of the seizures are used to localize the seizure onset zone; therefore, different s­ emiological studies use different methods in localizing the epileptic focus.4 However, a patient’s seizure onset zone can be confirmed most reliably by the “gold standard” of assessing patients with seizure free status after a surgical intervention, in whom the previously resected cortical tissue contained the seizure onset zone.5,​6,​7 The widespread usage of smart phones makes even more seizure recordings available.8 These recordings often miss the first few seconds of the episode, but are still very useful in refining the semiology of a certain seizure. In this chapter we will go through the most important factors of childhood seizure semiology by analyzing its localizing, lateralizing, and age-dependent aspects as well as the most typical semiology-related pitfalls.

„„ Localization-Dependent Seizure Semiology Frontal Lobe Seizures FLE is the most frequent extratemporal epilepsy and has a prominent place in presurgical evaluation. The first and largest adulthood FLE semiological studies were already completed in the early 1980s. Rasmussen et al reported the ictal features of 40 patients with FLE who had at least 5 years of seizure-free status postoperatively.9 They found that frequent warnings before the attacks, ictal automatisms, and secondary generalized tonic–clonic seizures (SGTCS) were typical. This first comprehensive study dealt with the frontal lobe as an anatomical entity and did not differentiate between the subregions within it.

12  Clinical Semiology in Preoperative Assessment More than a decade later, Salanova and Mihara observed three different seizure types in adults with FLE.10,​11 Localization of the epileptogenic lesions enabled them to conclude that in both the focal motor seizure group and the supplementary motor seizure group, the epileptogenic zones were restricted to the posterior third of the frontal lobe. Seizures originating from posterior third of the frontal lobe generated somatomotor manifestations as compared to the psychomotor seizure group, wherein the epileptogenic zones were largely in the anterior two-thirds. Using deep electrodes, many more anatomical regions within the frontal lobe were assessed and their typical seizure semiologies were distinguished.12–​14 According to the localization of the seizure onset zone, primary motor, mesial frontal, lateral frontal, frontopolar, central, orbitofrontal, and opercular seizures were differentiated (Fig. 12.1). However, the semiology of these seizures frequently overlapped, thus, creating a new challenge. Laskowitz et al stated that unfortunately the relation between clinical manifestations and the precise site of origin of a particular seizure remained problematic.15 Many patients have rather sizable areas of cortex implicated in seizure initiation, and once the epileptogenic activity begun, frontal lobe seizures are rarely confined to a small volume of cortex. Consequently, many ictal manifestations may result from seizure spread either to adjacent or distant cortex, and in several patients, ictal behavior was difficult to correlate with any specific frontal lobe location.15 Therefore, instead of splitting the frontal lobe seizures to further subregions, they discussed it as FLE syndrome. We can frequently face this overlapping semiology in young children whose spreading tendency is even more prominent. Typical features of frontal lobe epileptic seizures in adults include the following: • yy yy yy yy yy yy yy

Frequent, short seizures. Nocturnal tendency. Sudden onset and offset. Early motor signs. Frequent vocalization. Tonic seizures (tonic posturing and versive seizures). Hypermotor seizures with bimanual-bipedal automatism. Little or no postictal confusion.

In addition to adults’ FLE features, nocturnal tendency is also frequently seen in childhood FLE seizures.16 Motor signs are especially typical in infants and young children. In a series of children with FLE (< 7 years), all children had at least one motor seizure and only 6 of the analyzed 111 attacks (5%) showed no

motor signs at all. Tonic posturing (64%), clonic seizures (36%), and epileptic spasms (36%) are very characteristic. However, versive and hypermotor seizures as well as SGTCS can rarely be observed at this age. Postictal reorientation can be tested with difficulties in the preverbal group, but those young children who could cooperate well with the EEG technicians show a short postictal confusion, usually less than 10 seconds in duration.17 An important differential diagnostic challenge is distinguishing childhood FLE and sleep disorders. In a research study18 on 22 patients less than 16 years of age, seizures were brief (30 seconds to 2 minutes), stereotypic, nocturnal (17/21), and frequent (3–22/night). Clinical features included explosive onset, screaming, agitation, stiffening, kicking or bicycling of the legs, and incontinence. Therefore, diagnosis of FLE was not made in any of these children before referral; instead, they were diagnosed with sleep disturbance (10), psychiatric problems (6), or other seizure types (6). Adding to the challenge of diagnosis, interictal routine EEG as well as MRI were normal in 18/21.18

Temporal Lobe Seizures TLE is the major focal epilepsy in presurgical series and its semiology is very well defined. Typical features of temporal lobe epileptic seizures in adults include the following: • yy yy yy yy yy yy yy yy

Olfactory/epigastric/psychic aura. Arrest reaction. Staring or looking around. (Partial) loss of consciousness. Oral and hand automatisms. Facial grimace. Dystonic posturing. Agitation. Versive components.

Childhood TLEs can also be operated with good postoperative results therefore there are many studies examining pediatric TLE based on the gold standard rule.5,​6,​7,​19,​20,​21,​22,​23 Temporal lobe seizure semiology appears to be significantly influenced by age-related mechanisms so that ictal features in young children may not give much insight into the presence of this type of localization-related epilepsy.5,​6,​22 The diagnosis and referral of these patients for further evaluation, particularly for epilepsy surgery, may therefore be unduly delayed. Knowledge about the age-related temporal evolution of these seizures is critical

Fig. 12.1  Anatomical regions in the frontal lobe of the brain: anterolateral (a), inferior (b), and medial (c) views.

93

94

IIa  Preoperative Clinical and Neuropsychological Assessment because immature ictal manifestations of a young child may be transformed into a typically mature adult-like semiology with advancing age. Duchowny et al studied 16 children under 12 years of age with therapy-resistant TLE.6 They found typical complex partial seizures in 75% of them, with only a few young (< 5 years old) children showing motor attacks, as well. Brockhaus and Elger investigated children with TLE to compare their attacks to temporal lobe seizures occurring in adults.5 This study was based on video recordings of 29 children with TLE and ages from 18 months to 16 years. They found that children older than 6 years of age had ictal features similar to those of adults. In younger children, typical semiology included symmetric motor phenomena of the limbs, postures similar to the frontal lobe seizures in adults, and head nodding like those in infantile spasms. Jayakar and Duchowny further specified that this semiology transformation happens between the second and sixth year of life.22 An etiology-based study assessing patients exclusively with hippocampal sclerosis (HS) also found that the clinical features of mesial TLE in children as young as 4 years were similar to those in adults.24 Another study assessing only preschool children with TLE showed that this transformation occurred in a linear fashion as a function of age so that from the fourth year of life, the nonmotor component of psychomotor seizures, as the hallmark of limbic epilepsy as seen in adults, was the dominant seizure manifestation. By contrast, all patients under 42 months had a high ratio of motor features including tonic, clonic, and myoclonic components, as well as epileptic spasms compared to the overall observed seizure components. Beyond 42 months of age, the rate of complex partial seizure semiology with behavioral arrest and automatisms increased and became the predominant feature in half of the children.25 Animal studies investigating the ontogenetic expression of drug-induced limbic epilepsy in immature young rats showed a comparable age-dependent ictal behavior. Investigating

kainic acid and pilocarpine-induced seizures in young rats during the first 2 postnatal weeks, which corresponded to a maturational age of the human infants, demonstrated these rat pups developed hyperactivity, scratching, hyperextension of the limbs, tremor, head bobbing, and myoclonic movements.26,​27 More mature rats, age more than 2 weeks, produced limbic seizures consisting of rearing, akinesia, and masticatory movements in addition to prominent motor signs. Further studies in ­hippocampal-kindled rat pups demonstrated that after-­discharge thresholds were highest during the second to third postnatal week, suggesting resistance of the limbic system to synchronization.28 These findings from animal studies appear to offer a reasonable explanation for why temporal lobe seizures in immature humans only manifest more clearly with typical psychomotor features beyond the fourth year of life, once the limbic system has matured. Beside motor seizure components, there are other age-­ dependent semiological features described in the only seizure semiology study assessing postoperatively seizure free TLE patients from infancy to adulthood.29 In this study the same epileptologists analyzed both childhood and adult seizures and found that not only motor elements but also secondary generalization show age-dependent distribution. SGTCS were not observed in the preschool groups and appeared rarely (< 3%) in the school-aged group (Fig. 12.2). Later, during adolescence and adulthood, secondary generalization reached an overall incidence of 20%. This may be explained by the gradual maturation of the cortex,30 the immature dentritic development and myelin formation, and the imperfect synchronization of both hemispheres,31 as well as the shorter interval between the age of epilepsy onset and video-EEG monitoring of our younger patients.32 Peri-ictal autonomic symptoms are typical TLE features, and among them, epigastric aura and postictal coughing are significantly more frequent in children with TLE compared to extratemporal epilepsy. Other autonomic elements (e.g., nausea, vomiting, apnea, hyperventilation, and flushing) are typical

Fig. 12.2  Frequency of secondarily generalized tonic–clonic seizures (p = 0.003) and ictal automatisms (p < 0.001) as well as the etiology of hippocampal sclerosis (p < 0.001) showed agedependent frequency in patients with temporal lobe epilepsy. (Reproduced with permission from Fogarasi et al 2007.29)

12  Clinical Semiology in Preoperative Assessment childhood semiological features but without any localizing ­value.33 However, infantile focal seizures can typically start with exclusively vegetative forms (e.g., apnea or tachycardia), thus preventing easy recognition of early-onset seizures among newborns and preterm babies.34 Emotional signs—especially negative emotions like fear, anxiety, and crying—can be characteristic peri-ictal features of TLE in adults. Analyzing their appearance in children we found that the overall frequency of negative emotional signs was high (~50%) in focal childhood epilepsies, but they were more frequent in extratemporal than TLEs.35

Posterior Cortex Seizures Studies on seizures arising from the rear part of the brain usually mix occipital and parietal lobe epilepsy because of the low number of cases in each center. The parietal lobe is subdivided into the two lobules. Superior and inferior parietal lobules are responsible for integrating somatosensory information from the whole body as well as visuomotor coordination. The occipital lobe is the main center of the visual system, including the medially located primary visual cortex and the surrounding extrastriate region responsible for color discrimination, motion perception and visuospatial processing.36 Typical features of posterior cortex seizures mirror the normal function of these regions including somatosensory (pain or paresthesia), vestibular, or visual auras as well as oculomotor features such as nystagmus, eyelid myoclonia, and eye deviation.37,​38,​39,​40,​41,​42 Typical features of posterior cortex epileptic seizures in adults include the following: • yy yy yy yy yy

Sensory (mostly visual) aura. Oculomotor features. Nystagmus. Eyelid myoclonia. Eye deviation. Semiology of seizure spreading: ◦◦ “temporal-like” semiology. ◦◦ “SMA-like” semiology. yy Oral, manual, or complex automatisms. Visual auras can be elementary visual hallucinations or amaurosis originating from the primary visual cortex, or visual illusions, and complex visual hallucinations in the extrastriate cortex. However, distinguishing between such subtle features of visual auras are usually challenging in young children. Peri-ictal headache is a frequent symptom of both occipital epilepsy and migraines making differentiation more difficult.43 In the case of sustained eye deviation and ictal vomiting, one must rule out Panayiotopoulos syndrome, a nonlesional childhood occipital lobe epilepsy.44 The first video-based pediatric semiological study assessed 20 children under 16 years of age with occipital lobe epilepsy and found that 70% had visual auras consisting of mostly elementary visual hallucinations, ictal blindness, and even status epilepticus amauroticus.45 Visual loss is mostly bilateral, but homonymous hemianopsia contralateral to the seizure onset zone may also occur.39 Another study of 19 patients aged between 4 and 22 years with symptomatic occipital epilepsy described very brief (< 30 second) seizures with frequent spread to the temporal regions, therefore, indicating a very variable semiology.46 A

summary of seizures of 12 neonates and infants with occipital lobe epilepsy found bilateral seizures (including spasms in 58%) as well as abnormal ictal ocular movements.47 Our gold standard semiology study on preschool children found frequent (13/18) oculomotor features in the form of nystagmus, eye deviation, eyelid myoclonia, and rapid, repetitive blinking. Ictal flush (5/18) as well as ictal smile (5/18) were also typical semiological features rarely seen in other childhood seizures.48 It is important to note that eye deviation itself had neither a lateralizing49 nor a localizing significance, as it can be a manifestation of both the occipital lobe50 and frontal eye field activation.51 Comparing seizure semiology of children with pure FLE and posterior cortex epilepsy (PCE), we found that characteristic features described in adults’ extratemporal epilepsies were frequently missing in childhood seizures. Myoclonic seizures, epileptic spasms, psychomotor seizures, atonic seizures, oral and manual automatisms, as well as vocalization and eye deviation appeared in both groups without significant difference showing fast spreading of seizures in both extratemporal and temporal regions. While ictal features provide only minor assistance in differentiating childhood FLE from PCE, nocturnal appearance (more frequent in FLE) and the type of aura (visual in PCE) have high localizing value emphasizing the importance of accurate history taking (Table 12.1).52

Seizures with Insular Origin Insular epilepsy and its surgical treatment, especially in childhood, is a special and very challenging field. This is not only because of difficulty of surgical assessment, but also because of the high ratio of under-diagnosed patients. Because it is a hidden area under opercula, EEG features are usually poorly recorded and insular lesions are difficult to find if clinicians don’t orientate their radiologist colleagues. There are no large surgical series for insular epilepsy, and the information related to insular epilepsy is mainly derived from stereoelectroencephalography (SEEG) studies of children. Dylgjeri et al reported 10 children with insular epilepsy demonstrated by SEEG electrodes in the insular cortex.53 They concluded that insular epilepsy is a severe focal epilepsy characterized by an early onset (mean: 1.3 years) and with a significant impact on neuropsychological development in children. The absent or insufficient description of the auras in young children delayed the diagnosis. They suggest that in the case of special semiological features such as a strange or painful reaction, arrest, suffering, and neurovegetative symptoms, an insular onset should be considered. They also found that the semiology of insular seizures is very similar to frontal lobe semiology (unilateral tonic and clonic as well as hypermotor seizures), creating a differential diagnostic challenge in these children. Nevertheless, we do not know the exact semiology of pure insular cases, as all of their patients needed insula-plus resections: insulo-opercular (3), insulo-operculofrontal (1), insulo-operculo-central (2), insulo-frontal (1), and insulotemporal (in one patient). Another recent study applied nonlinear regression analysis to SEEG-recorded seizures in five patients and found that the frontal semiology was expressed by strong functional couplings between the insula and mesial frontal regions.54 These data

95

96

IIa  Preoperative Clinical and Neuropsychological Assessment Table 12.1  Frequency and localizing value of different peri-ictal signs observed in 35 children < 12 years with extratemporal epilepsy

Observed sign

Present in FLE patients (n = 20)

Present in PCE patients (n = 15)

p value

Somatosensory aura

5 (25%)

0

0.06

Visual aura

0

4 (27%)

0.01

Epigastric aura

1 (5%)

1 (7%)

1.00

Tonic

14 (70%)

3 (20%)

< 0.01

Versive

0

3 (20%)

0.07

Clonic

10 (50%)

3 (20%)

0.07

Myoclonic

5 (25%)

3 (20%)

1.00

Epileptic spasm

6 (30%)

4 (27%)

0.83

Hypermotor

4 (20%)

0

0.12

Atonic

2 (10%)

1 (7%)

1.00

Psychomotor

6 (30%)

8 (53%)

0.16

SGTCS

3 (15%)

2 (13%)

1.00

Oral automatism

8 (40%)

8 (53%)

0.43

Manual automatism

3 (15%)

5 (33%)

0.25

Vocalization

6 (30%)

3 (20%)

0.70

Eye deviation

8 (40%)

7 (47%)

0.69

Nystagmus

0

7 (47%)

< 0.01

Nose wiping

2 (10%)

6 (40%)

0.051

Todd’s paralysis

1 (5%)

0

1.00

Dysphasia

0

1 (7%)

0.43

Auras

Seizure components

Other ictal signs

Postictal signs

Abbreviations: FLE, frontal lobe epilepsy; PCE, posterior cortex epilepsy; SGTCS, secondarily generalized tonic–clonic seizure. Source: Reproduced with permission from Fogarasi et al 2005.52

suggest considering the insular origin of seizures also in cases of cryptogenic mesial frontal epilepsies.

Seizure Semiology of Multilobar and Hemispheric Epilepsies Early-onset childhood catastrophic epilepsy, especially due to focal cortical dysplasia, frequently manifest in multilobar form. In a recent series, one-third of children less than 5 years of age with therapy-resistant epilepsy had multilobar dysplasia and needed invasive presurgical evaluation.55 Most typical forms are frontotemporal and temporo-parieto-occipital cases showing a wide variety of seizure semiology. Another recent analyzing chart of 225 consecutive pediatric epilepsy patients undergoing video-EEG monitoring found multilobar etiology in 91 (40%) of them, with a significantly frequent semiology of tonic, hypermotor, and versive elements.56 The patients with more than one seizure onset zone are especially challenging cases. A masked reviewer study of 17 patients with dual seizure foci concluded that seizure semiology was not as useful as invasive EEG in localizing seizure onset in these patients.57

„„ Epileptic Spasms: When Age Overcomes Localization Infants with any etiology, localization, or extension of cerebral lesions might develop a unique seizure semiology: epileptic spasm clusters. Since the advent of functional neuroimaging, many infants with nonlesional epilepsy manifesting in infantile spasms have been operated on. In a study of 23 patients under 3 years of age who underwent cortical ­resection or hemispherotomy for intractable infantile spasms, the functional deficit zone was very large: in half of them, it included the temporo-parieto-occipital region, and, in an ­additional six children, the whole hemisphere.58 In another study, a close temporal association between spasms and partial seizures was found.59 Others hypothesized that spasms might be facilitated (or even induced) by ­partial seizures.60 In a long-term follow-up of 214 children with ­epileptic spasms, 60% of them developed new focal seizures, mostly from the temporal lobe.61 The data support a model in which the spasms, although possibly generated at a subcortical level, are induced by focal discharges from cortical pathology.

12  Clinical Semiology in Preoperative Assessment There is also data supporting both a cortical and subcortical origin of spasms.62 Because of the hyperexcitability and the high tendency of spread of paroxysmal activity to the basal ganglia, a focal cortical lesion may trigger the subcortical structures manifesting in clusters of epileptic spasms.62 The current spectrum of disorders associated with clinical spasms with an onset in infancy is more extensive than previously thought. Therefore, the terminology of spasms has changed.63 In some centers, the term infantile spasms syndrome defines an epileptic syndrome with an onset during infancy, with clusters of clinical spasms and hypsarrhythmia on EEG. According to this classification, West syndrome is a subgroup characterized by the combination of clustered spasms, hypsarrhythmia, and cognitive delay or regression. We can use the term infantile spasms single-spasm variant if spasm occurs with no clusters. Finally, hypsarrhythmia without infantile spasms as well as infantile spasms without hypsarrhythmia also exist. It is unknown why infants with similar lesions will develop either spasms (and other severe symptoms of West syndrome) or “only” pure focal epilepsy. There is growing evidence that infantile spasms may result from disturbances in key genetic pathways of brain development. Dysmorphic features, autism, movement disorders, or systemic malformations might be connected with these genetic associations.63 Because of their semiological diversity, many subtle spasms are almost impossible to distinguish from normal infantile motor behavior. The paroxysmal events can be so delicate that only eye movement, a very fine head nod, or a minimal lift of shoulders happens during the seizure. Even in these cases, clusterization of the subtle elements might help to recognize epileptic spasms (see Video 12.1).64

„„ Does Etiology Play a Role in Childhood Seizure Semiology? There is no study supporting the concept that seizure semiology would depend on etiology itself. However, etiology of therapy resistant epilepsies shows age-dependent features. While malformation of cortical development is more frequent among infants and young children, one can find more vascular and tumor etiology in the elder population.65 Within TLE, HS is the main etiological factor in adolescence and adults bringing domination of mesial TLE syndrome and its typical semiology in adulthood. In our study of a wide age cohort of TLE showed that the frequency of ictal automatisms, secondary generalization, and the number of different lateralizing signs increased while the ratio of motor seizure component decreased by age. However, using HS etiology adjusted linear models revealed that the frequency of automatisms and secondarily generalized seizures as well as the number of different lateralizing signs were HS independent significant variables while incidence of motor seizure elements depended on both age and etiology.29 Video 12.1 Epileptic spasms. (This video is provided courtesy of András Fogarasi.) ht t ps://www.thieme.de/de/q.ht m?p=opn/ tp/255910102/9781626238176_c012_v001&t=video

On the other hand, etiology often determines localization, for example, HS is always, and benign tumors are frequently, bound to the temporal region. Vascular lesions are usually extratemporal while tuberous sclerosis complex has a characteristically multilobar etiology. Finally, hemispheric epilepsies are typical in Sturge–Weber syndrome, large strokes, ­Rasmussen’s encephalitis, hemimegalencephalia, and in some forms of polymicrogyria.66

„„ Peri-ictal Lateralizing Signs in Children During presurgical evaluation, the knowledge of lateralizing signs greatly helps epileptologists. Lateralizing signs are special ictal or postictal features giving information on which hemisphere is involved in a certain attack. Semiological studies of different adult studies have shown that unilateral dystonic or tonic posturing, unilateral mouth deviation, other unilateral facial or distal motor patterns (i.e., clonic and myoclonic seizures), and the forced versive head rotation immediately before secondarily generalization are contralateral, while unilateral eye blinking, hand automatism, and postictal nose wiping are ipsilateral lateralizing features—especially in the most frequently analyzed TLE group. Typical features of the language-dominant hemisphere seizures include postictal aphasia and dysphasia; however, orgasmic aura, ictal speech, peri-ictal autonomic signs, and automatisms with preserved consciousness originate mostly from nondominant seizure onset zones.67,​68,​69,​70 Our blinded, multiobserver study of 100 children showed that childhood lateralizing signs occur less frequently than those of adults, but achieve very good interobserver agreement and high predictive value. Most signs are consistent, and only few of them appear too rarely to be reliable. The most frequent and reliable childhood lateralizing signs are summarized in Table 12.2 and also shown in Video 12.2. Some signs (e.g., unilateral manual automatism or postictal nose wiping) were very frequent (> 20%) and most of them showed a high predictive value (90–100%). It is important to note that eye deviation has no lateralizing value at all, even if it is the most frequent asymmetrical peri-ictal sign.49 A number of different lateralizing signs within each patient showed a linear relation with the age at monitoring, suggesting an ontogenic expression changing parallel with the maturation of the developing brain. Although infants and young children frequently present bilateral synchronous ictal motor phenomena rather than unilateral motor seizures, there are also frequent and reliable signs appearing with similar frequency in all ages (Fig. 12.3). Among them, asymmetric epileptic spasm,71 unilateral clonic movement, ictal nystagmus, as well as postictal nose wiping and Todd paresis are very important in the clinical ­practice. These latter lateralizing signs can also add important Video 12.2 Lateralization of the seizure onset zone. (This video is provided courtesy of András Fogarasi.) ht t ps://www.thieme.de/de/q.ht m?p=opn/ tp/255910102/9781626238176_c012_v002&t=video

97

98

IIa  Preoperative Clinical and Neuropsychological Assessment Table 12.2  Frequency and lateralizing value of different peri-ictal signs in children 60 spikes per minute) were associated with glucose hypermetabolism (Fig. 14.6).56 Advances in neuroimaging include the combination of PET and CT imaging, which improved the anatomical localization of brain glucose metabolism, and also a decreased scan time for children. Simultaneous EEG recording at the time of PET-CT imaging raises the possibility of artifacts originating from the EEG electrodes.58

EEG and Magnetoencephalography EEG offers a sampling of brain activity from a more extensive surface with a high temporal resolution; however, EEG is less sensitive in detecting deeper and smaller cortical areas.59 On the contrary, MEG provides excellent spatial resolution of the signals from the interhemispheric, orbitofrontal surfaces, insula or basal aspect of temporal or frontal lobes, the areas that are not usually detected using scalp EEG.60 EEG spikes and MEG spikes differ by orientation, location, timing, and size of the cortical surface.61 Dendrites are positioned parallel to the cortical surface, which can generate brain activity with a tangential field. MEG spikes can be detected if the magnetic field is oriented tangentially to the skull. On the other hand, EEG can detect spikes from a larger cortical surface and is sensitive to all source orientations. This activity can be detected by MEG, whereas EEG detects activity that has a radial field. Moreover, synchronized brain activity originated from a smaller focus (3–4 cm2) can be detected by MEG. EEG can detect all MEG recorded spikes; however, MEG can miss some EEG spikes.62,​63,​64

14  Electroencephalography and Noninvasive Electrophysiological Assessment

Fig. 14.6  A 2-year-old child diagnosed with focal cortical dysplasia type 2a and infantile spasms refractory to the medical treatment. (a) Brain MRI demonstrated cortical dysplasia involving entire right temporal lobe. (b) FDG-PET imaging in the interictal state following generalized anesthesia demonstrated significant increase in glucose uptake in the right temporal lobe. (c) EEG example at interictal state prior to the epilepsy surgery demonstrating occasional bilateral independent epileptiform activity pronounced in the right hemisphere without evidence of electrographic seizure.

Scalp EEG is less sensitive to detect deeper and smaller cortical areas whereas MEG has advantages to identify the smaller and deep epileptogenic focus.64 If a cortical source generates brain activity with two different orientations, radial and tangential to the recording surface, spikes with radial orientation can be detected by EEG and the spikes with tangential orientation by MEG imaging. Therefore, EEG and MEG study combined are complementary to each other to detect the spike dipoles despite their difference in field and orientation. If MEG and EEG spike fields are synchronous, then EEG will detect the radial component of the spike dipole; MEG will detect

the tangential component. If MEG and EEG spikes are asynchronous, the earliest spike field will represent the initial source, and the later spike field will indicate the propagation pattern. For instance, when a MEG spike leads an EEG spike, MEG identifies the spike source and EEG suggests a radial component of propagation. If an EEG spike leads the MEG spikes, EEG will suggest the origin of the spike source and MEG will indicate the tangential component of propagation. In nonlesional TLE, MEG is found more sensitive than EEG in detecting epileptic activity (Fig. 14.7). Therefore, a negative EEG should not preclude requesting MEG for the presurgical evaluation.63,​64 

117

118

IIb  Preoperative Electrophysiological Assessment

Fig. 14.7  An 18-year-old patient diagnosed with left temporal lobe drug-resistant epilepsy. Scalp EEG demonstrates similar seizure onset from the posterior temporal region before the clinical onset (seen at electrodes P7, T7) prior (a) and after (b) the epilepsy surgery. (c) MEG was obtained following epilepsy surgery after seizures recurred: (c) MEG dipole was noted earlier than EEG dipole suggesting the epileptogenic focus located in the basal temporal lobe. EEG dipole indicated the seizure propagation pattern to the posterior lateral temporal surface.

14  Electroencephalography and Noninvasive Electrophysiological Assessment

„„ EEG and Functional MRI EEG provides good temporal resolution; however, spatial resolution is limited for localization of the epileptogenic focus. Therefore, a combination of EEG and functional MRI (fMRI) is recommended to improve spatial resolution for localization of epileptogenic zone in focal epilepsy. fMRI measures changes in local blood flow by detecting blood oxygen level dependent (BOLD) signals.65 The occurrence of an ictal event or frequent epileptiform discharges will affect the local blood flow changes, which will correlate with BOLD signal changes detected by fMRI. Therefore, EEG recording combined with fMRI may be beneficial in the presence of frequent spiking or the presence of electrographic or frequent clinical seizures.

„„ Pitfalls of Localization of Epileptiform Activity Seizure Semiology The clinical features of seizures vary based on the location of the epileptogenic focus. For instance, dystonia or other motor manifestation is seen more often at the onset of seizures arising from frontal lobe whereas oral or hand automatisms are often reported with temporal lobe seizures.66 The clinical features of seizures, “semiology,” is supportive but not definitive for localization. The practitioners must be cautious interpretation of semiology, especially for the seizures originating from the extratemporal regions. For instance, in FLE, clinical signs may localize the seizures to the frontal lobe but not to the specific areas of the frontal lobe.67 Ideally, if there are multiple seizure types, each should be captured and analyzed, especially with a possible multifocal onset, such as with a cortical dysplasia. Bilateral seizures preclude a focal cortical resection if occurring equally from both hemispheres. However, resection can be considered in selected cases if the majority of disabling seizures are from one side. Seizure semiology can be misleading for focal epilepsy in the presence of bilateral motor manifestations and absence of clinical features characteristic for focal epilepsy, such as behavioral arrest or automatisms in young children. Ictal EEG may demonstrate a propagation pattern corresponding to the clinical semiology and may mislead the actual localization of seizure onset. Moreover, the aura (initial seizure onset) could not be detected by scalp EEG if an epileptogenic focus is smaller than 6 square cm.60,​68 In a study of simultaneous surface and subdural EEG recordings of simple partial seizures (auras), only 6 of 55 auras were captured by scalp EEG.69

EEG and Epileptogenic Lesions The presence of an epileptogenic lesion on MRI is a strong predictor for seizure onset and postsurgical outcome. However, MRI may reveal more than one epileptogenic lesion, as often seen in tuberous sclerosis.70,​71,​72 In fact, false localization could also occur in approximately 3 to 5% of patients in TLE. The presence of a unilateral structural lesion, therefore, should be interpreted with caution, particularly in children, in whom the

presence of discordant or ambiguous EEG findings may suggest another epileptogenic focus. A clinical report describing 29 patients diagnosed with tumor-related TLE, interictal epileptiform activity was seen in only 63%; lateralized in 70% but localized in only 55%. Although Ictal EEG lateralized the seizure onset in 80% of the patients, localization was still limited to only 40%.73 On the other hand, concordant ictal and interictal EEG and MRI findings are essential for prognosis and in predicting a favorable postsurgical outcome. The presence of cortical dysplasia in association with other epileptogenic lesions is known as “dual pathology.” In epilepsy, dual pathology refers to the presence of different pathological abnormalities, which may be unrelated or, more likely, the propagation of discharges from the initial epileptogenic lesion. The coexisting epileptogenic lesions include hippocampal sclerosis, developmental tumors, gliosis, vascular malformation, and cortical dysplasia.74,​75,​76 In the new classification system for FCD, coexisting epileptogenic lesions are classified as type 3 FCD.77,​78 Low-grade tumors may be accompanied by cortical dysplasia adjacent to the epileptogenic lesion.71 Cortical dysplasia is diagnosed in 4 to 20% of the patients diagnosed with ganglioglioma or other developmental tumors such as dysembryoplastic neuroepithelial tumor. Moreover, cortical dysplasia was identified in patients with perinatal infarction, up to 10%, confirmed by histopathology.79 Hippocampal sclerosis is also reported in 20% of the patients diagnosed with cortical dysplasia. Therefore, the presence of discordant ictal and interictal EEG findings compared to the MRI findings should raise the possibility of dual pathology, mainly associated with epileptogenic lesions and cortical dysplasia in the pediatric age group. FCD is the most common pathology found in children who underwent epilepsy surgery.80 Interictal and ictal scalp EEG findings of FCD are not diagnostic of underlying dysplastic tissue. However, the presence of focal slowing and interictal paroxysmal fast activity and polyspikes on EEG are often in the region where the dysplastic cortex is located. The scalp EEG may be of limited value to identify epileptic activity. For instance, the UCLA series found that interictal EEG findings can localize the epileptiform activity to one lobe in 49% of cortical dysplasia whereas the ictal EEG may detect the epileptogenic focus in 60%.81,​82,​83 Early brain lesions may result in drug-resistant focal epilepsy in the pediatric age group. The clinical features of seizures and EEG findings may be challenging to localize the epileptogenic focus despite a significant or widespread MRI abnormality such as cystic encephalomalacia, or cortical hemiatrophy. EEG may demonstrate generalized epileptiform discharges or even hypsarrhythmia. Wyllie et al. described 50 children who were initially excluded for a presurgical evaluation based on ictal and interictal EEG findings.84 When the authors reviewed the clinical and EEG features of those initially excluded, 50% of them had no focal clinical or EEG features that could localize the epileptogenic focus. Moreover, all had widespread cortical abnormality suggesting either perinatal or postnatal injury. Following epilepsy surgery when performed based on the MRI abnormalities, 72% became seizure free at the last follow-up.84 Electrical status epilepticus of sleep (ESES) is a unique clinical electrographic syndrome in the pediatric age group that was first described in 1971.85 Focal and generalized seizures are frequent and often reported in sleep.41 Continuous runs

119

120

IIb  Preoperative Electrophysiological Assessment of generalized spike and wave complexes comprise 85% of the non-REM sleep recording. Seizure frequency varies despite the overwhelming sleep EEG findings. Academic difficulties, behavioral problems, regression in developmental milestones are common and often lead to the diagnosis.41,​86,​87 Antiepileptic drug treatment and ketogenic diet could be effective to improve EEG findings of ESES.88,​89 However, recurrence of ESES is common despite the initial improvement in medical treatment. Focal lesions can cause the generalized pattern of ESES.89 PET imaging was found informative to demonstrate the underlying

dysfunctional cortex associated with ESES.90 Therefore, despite the generalized epileptiform abnormality such as ESES should not rule out the epilepsy surgery evaluation in children with drug-resistant epilepsy, particularly the ones presenting with focal MRI abnormalities (Fig. 14.8).91 A recent report describing a meta-analysis of more than 500 patients diagnosed with ESES, epilepsy surgery remains the most effective treatment modality for the remission of ESES and improvement in cognitive function, reported in 65% of patients diagnosed with ESES.87

Fig. 14.8  Ictal EEG recorded from a 19-year-old patient diagnosed with drug-resistant left temporal lobe epilepsy secondary to an undetermined etiology. (a) Scalp EEG demonstrated epileptiform activity arising from the posterior temporal cortex with spread pattern. (b) Intracranial EEG localizes the seizure onset in the mesial basal temporal cortex with a spread pattern to the posterior temporal cortex. Red closed circles on the schematic view of the intracranial electrodes demonstrate the seizures onset from the contacts on the dorsolateral and subtemporal surface.

14  Electroencephalography and Noninvasive Electrophysiological Assessment

False Lateralization/Localization Scalp (surface) EEG is limited to localize the epileptogenic focus accurately. Moreover, the ictal focus was false lateralized to the contralateral hemisphere despite the presence of unilateral hippocampal sclerosis in the patients with TLE.18 This phenomenon is referred to as “burned out hippocampus” syndrome. If the hippocampus is badly damaged, neocortical recruitment of neurons may not be detected by the scalp EEG until propagation of a seizure occurs into the contralateral temporal lobe.19 The spread of temporal seizure from a mesial temporal region to the contralateral hemisphere usually happens within 3 to 62 seconds. Three ictal propagation patterns of temporal

lobe seizures were described: (1) Spread from one mesial temporal region to the neighboring temporal or frontal neocortex; (2) Spread from one mesial temporal region to the other before the involvement of the ipsilateral temporal neocortex; and (3) Spread from one mesial temporal region to the contralateral mesial temporal region and then lateral temporal cortex in the contralateral hemisphere.8 Because of the obscurity to lateralize the ictal activity accurately, the importance of interictal activity has been emphasized to determine the epileptogenic focus in TLE. Because of the limitation of scalp ictal EEG data, or discordant imaging findings, the interictal EEG findings can be concordant and informative to localize the epileptogenic zone (Fig. 14.9).92

Fig. 14.9  Scalp and intracranial ictal EEG examples on a 13-year-old patient with normal MRI findings diagnosed with the left frontal lobe epilepsy associated with focal cortical dysplasia type 2a. (a) Scalp EEG failed to localize and lateralized the seizure onset, obscured by the muscle artifact because of hyper motor movements. (b) Intracranial EEG illustrated the focal spikes discharges arising from the dorsolateral surface of the frontal grid, localizing seizure onset prior to the clinical onset. Red closed circles on the schematic view of the intracranial electrodes indicate the seizures onset from the contacts on the dorsolateral frontal convexity. Arrows point out the specific electrodes involving for the seizures onset prior to the clinical onset.

121

122

IIb  Preoperative Electrophysiological Assessment In the presence of cortical cysts, porencephaly, encephalomalacia, and marked cortical atrophy, the ipsilateral cortical volume may not be enough to generate ictal scalp findings or scalp EEG captures only propagation pattern.19 We have seen this in other disorders where a hemisphere is severely damaged, such as Sturge–Weber syndrome. If there has been previous surgery, such as prior resection with recurrent seizures, or a skull defect (breach rhythm), then there may be a distortion of epileptiform activity because of the skull defect, fluid-filled cavities, or distorted postoperative anatomy with adhesions.

Normal EEG EEG has an essential role in the initial evaluation of seizures and the localization of the epileptogenic focus. The absence of epileptiform abnormalities on scalp EEG recording should not preclude the presurgical evaluation. Scalp EEG (noninvasive) recording may fail to detect cortical spikes (Fig. 14.10).82 Even

when the surface electrode shows well-defined spikes or sharp waves corresponding to seizure semiology, these may represent a more regionalized or propagating paroxysmal activity rather than a localized phenomenon. A cortical area of 10 to 20 square cm is often required to generate a scalp recognizable interictal spike or ictal rhythm. Sufficient cortical source area and synchrony are needed for the corresponding scalp EEG epileptiform recording. The amplitude is primarily dependent on source area and synchrony, and therefore it is a less important factor.68,​93 Data from the simultaneous surface and invasive recordings show that a much more extensive cortical area is involved in scalp EEG electrodes at the onset of spike onset whereas more discrete area is identified in intracranial cortical recording.68

„„ Emerging Noninvasive Neurophysiology Methods A number of novel clinical neurophysiological methods have emerged to localize the epileptogenic focus in the recent years.

EEG Source Imaging

Fig. 14.10  Epileptogenic lesion and EEG findings. (a) Brain MRI of a patient diagnosed with left sided hemiplegic cerebral palsy and drug resistant infantile spasms secondary to the cystic encephalomalacia located in the right hemisphere. (b) EEG demonstrates frequent spikes discharges in sleep with maximum amplitude in the bifrontal and midline central regions suggesting electrical status epilepticus of sleep (ESES).

EEG Source Imaging (ESI) is one of the emerging noninvasive model-based imaging technique that provides information on the temporal and spatial components of EEG to detect the generator of the electrical activity (the epileptogenic focus). The role of ESI to detect the irritative zone and ictal onset zone was examined in focal epilepsy. The data obtained from either scalp or intracranial EEG recording demonstrated that ESI has a high positive predictive value.42,​94,​95,​96 In 152 patients with intractable epilepsy, accuracy of ESI was compared with the other established noninvasive methods. Sensitivity and specificity of ESI were found comparable to that of PET, SPECT, and MRI, and similar in temporal and extra TLE.97 Therefore, ESI is considered as a promising method to contribute a visual analysis of EEG as part of presurgical evaluation.96 Dense array EEG is applied for the analysis to provide good scalp surface coverage using 256 or more electrodes.98 Dense array EEG detects the paroxysmal ictal activity earlier and ictal onset zone can be delineated better using a special software for the ­analysis.98,​99 Because of the limitation of scalp EEG recording, novel noninvasive methods are in the clinical practice as part of presurgical evaluation. In a small cohort of 14 patients with drug-resistant epilepsy, ESI was applied retrospectively on EEG data before the epilepsy surgery. Ictal ESI correctly localized the seizure-onset zone in the resection area in 5 of 6 patients with excellent postsurgery outcome. Ictal and interictal ESI was also found concordant in 9 of 14 patients (64%) whereas discordant in only 1 patient (7%). Another study examined the role of EEG-fMRI and ESI in 53 children with drug-resistant epilepsy to map the interictal discharges. In 29 patients, the epileptogenic zone was well characterized. The epileptogenic zone was well characterized in 29 patients; 26 had an EEG-fMRI localization that was correct in 11, 22 patients had ESI localization that was correct in 17, and 12 patients had combined EEG-fMRI and ESI that was correct in 11. EEG-fMRI correctly predicted seizure outcome following resection in eight of 20 patients, and by the ESI in 13 of 16. The

14  Electroencephalography and Noninvasive Electrophysiological Assessment combined EEG-fMRI/ESI region entirely predicted outcome in nine of nine patients, including three with no lesion visible on MRI.97

Infraslow Activity Introduction of digital EEG and advances in amplified technology have expanded the detection of activity with various frequencies from infraslow (< 0.5 Hz) to high oscillations with high frequency (80–250 Hz). The slow activity with a frequency below the delta activity, also known as “ISA” “DC shift” attracted attention to localize the epileptogenic focus.87 To examine ISA, high pass filter setting should be changed to at least to 0.1 Hz or 0.01 Hz.100,​101,​102 An ISA represents a small electrical field with negative potentials at the site and positive potentials at a distance. Therefore, a negative baseline shift at the seizure onset is an indicator of particular site involving ictal process rather than spread pattern from the other areas.89 Thus, ISA is found informative to localize the ictal onset which may not be detected by traditional scalp EEG recording.100 ISA was studied in the various childhood-onset epilepsy syndromes. Rodin et al applied ISA on ictal EEG data in 29 patients diagnosed with absence seizures and 20 patients with partial seizures. Data were analyzed using conventional filter setting first and then changing the filter setting to 0.01 to 0.1 Hz. ISA is also applied to analyze ictal scalp EEG data.103 Another study examined the role of ISA to localize the ictal onset in the patients with infantile spasms. In the recording of 101 ictal EEG patterns in 13 patients with infantile spasms, ISA was detected in 77% of the seizures, lateralized in 20%, and generalized in 57%.104

High-Frequency Oscillations HFOs denote a paroxysmal activity with frequency in the 80 to 500 Hz range. Moreover, based on the frequency, HFOs are subcategorized as ripples (80–250 Hz) or fast ripples (> 250 Hz). However, the detection of HFOs requires experience to differentiate HFOs from an artifact or a physiologic activity. Thus, HFO analysis should be performed either during the sleep or in the EEG data from the intracranial monitoring to eliminate the risk of artifact contamination. Scalp EEG does not filter HFOs, but the signal amplitude is decreased correlating with the distance between the generator and recorder. Scalp EEG can detect HFOs if the recording EEG electrodes are at the location of spike generator.68 Simultaneous scalp and ECOG from intracranial electrodes can also detect HFOs in both scalp surface and subdural EEG recordings.105 High-density EEG may filter artifacts and provide information from smaller surfaces. Hence, application of dense array EEG is recommended to sample HFOs to eliminate artifacts.94 On the contrary, intracranial EEG recording is less often compromised by muscle or electrode artifacts. For instance, depth electrodes provide recording from mesial structures with excellent spatial resolution and fewer artifacts. Some studies, in fact, highlight the benefit of depth electrode recording for the detection of HFOs distinctly during the ictal activity.105,​106

Functional Near-Infrared Spectroscopy-EEG fNIRS-EEG monitoring is a novel noninvasive multimodal analysis to monitor blood flow changes and changes in oxygenation

using near-infrared light.107 In the case of increasing neuronal activation, the hemodynamic response is required to meet the metabolic demand that can be detected by fNIRS. Hemoglobin is the primary absorber of near-infrared light. Therefore, fNIRS can record changes in hemoglobin concentration, including deoxyhemoglobin, oxyhemoglobin, and the total hemoglobin in the human brain. However, the fNIRS signal has a low temporal resolution compared to fMRI that also measures hemodynamic changes.108 Functional NIRS-EEG examined to monitor brain function in children diagnosed with traumatic brain injury.107 In the cases subjected to a traumatic brain injury, fNIRS-EEG demonstrated the variation in hemoglobin concentration in correlation with worsening intracranial pressure and arterial pressure changes. Moreover, fNIRS also demonstrated cerebral hypoxygenation preceding the seizure onset.107 Similarly, when newborn seizures were analyzed with fNIRS-EEG, cerebral hypoxygenation was also observed before the onset of electrographic seizures.108 Steinhoff and colleagues studied the application of fNIRS and video EEG monitoring as part of the presurgical evaluation.109 Continuous epileptiform activity was correlated with decreasing cerebral oxygen availability in the ipsilateral f­ rontal ­cortex.54 Watanabe examined the role of the fNIRS m ­ ethod to identify the epileptogenic focus in refractory epilepsy patients.110 A noticeable blood flow increase was noted at the site of the epileptogenic focus in nearly all patients (28/29, 96%) compared to SPECT imaging which captured the blood flow changes in only 69% of the patients. A recent study described the role of fNIRS to identify the ­epileptogenic focus in 40 patients diagnosed with neocortical epilepsy.111 Change in hemoglobin level was detected in majority of the patients (62%) in the epileptogenic focus. Moreover, changes in deoxyhemoglobin and oxyhemoglobin levels correlated with interictal epileptiform discharges in the region of epileptogenic focus in 18 patients. The overall sensitivity of fNIRS was found 62% to detect epileptogenic focus in drug-­ resistant epilepsy, and this modest sensitivity of fNIRS was attributed to the possibility of subtle variation in blood flow and oxygenation in the area where epileptiform discharges are located. In refractory mesial TLE, eight seizures recorded from three patients were evaluated by fNIRS. All seizures were associated with significant local and remote hemodynamic changes. Hemodynamic changes coincided with the seizure onset. During the onset of the seizures, a significant decrease in deoxyhemoglobin was observed with a concurrent increase in oxyhemoglobin level which often outlasted the seizure duration. Furthermore, hemodynamic changes were also observed in the contralateral hemisphere at lesser extent, suggesting the possibility of s­eizure propagation despite the absence of ­epileptiform activity in the contralateral hemisphere. In the FLE, fNIRS demonstrated various changes in deoxyhemoglobin level while oxyhemoglobin increased during the seizure. Eighteen seizures were recorded from nine FLE patients.112 Hemodynamic changes preceded the epileptiform activity (2 seconds earlier), outlasted the seizure offset in the epileptogenic focus and correlated with seizure duration. For instance, brief electrographic seizures were associated with subtle changes in hemoglobin level compared to the more prolonged seizures. However, increased local blood flow change was not coupled by the level of changes in hemoglobin level in

123

124

IIb  Preoperative Electrophysiological Assessment the majority of the patients. Similar activation pattern was also seen in the contralateral hemisphere to a lesser extent. The role of fNIRS in posterior epilepsy was described in a small cohort of nine patients. Deoxyhemoglobin levels were more b ­ roadly distributed however lateralized the epileptogenic focus in five

patients. On the contrary, EEG-fNIRS was less informative. Despite the abnormal MRI findings in six patients diagnosed with infantile spasm, EEG-fNIRS failed to localize or lateralized the epileptogenic focus and hemodynamic changes were observed in the frontal regions.113

References 1. Berg AT, Berkovic SF, Brodie MJ, et al. Revised terminology and concepts for organization of seizures and epilepsies: report of the ILAE Commission on Classification and Terminology, 2005– 2009. Epilepsia 2010;51(4):676–685

19. Fonseca VdeC, Yasuda CL, Tedeschi GG, Betting LE, Cendes F. White matter abnormalities in patients with focal cortical dysplasia revealed by diffusion tensor imaging analysis in a voxelwise approach. Front Neurol 2012;3:121

2. Duchowny M, Cross JH. Preoperative evaluation in children for epilepsy surgery. Handb Clin Neurol 2012;108:829–839

20. White NM, Salinas JA. Mnemonic functions of dorsal striatum and hippocampus in aversive conditioning. Behav Brain Res 2003;142(1–2):99–107

3. Beĭer EV, Arushanian EB, Titenok AL, Alferov VV. [The dorsal hippocampus injury influences chronobiological effects of depressants and antidepressants in rats] Eksp Klin Farmakol 2003;66(3):9–12 4. Joca SR, Padovan CM, Guimarães FS. Activation of post-synaptic 5-HT(1A) receptors in the dorsal hippocampus prevents learned helplessness development. Brain Res 2003;978(1–2):177–184 5. Mintzer S, Cendes F, Soss J, et al. Unilateral hippocampal sclerosis with contralateral temporal scalp ictal onset. Epilepsia 2004;45(7):792–802 6. Yang CS, Chow JC, Tsai JJ, Huang CW. Hyperventilation-induced ictal fear in nonlesional temporal lobe epilepsy. Epilepsy Behav 2011;21(1):100–102 7. Kudr M, Krsek P, Maton B, et al. Ictal SPECT is useful in localizing the epileptogenic zone in infants with cortical dysplasia. Epileptic Disord 2016;18(4):384–390 8. Foldvary N, Klem G, Hammel J, Bingaman W, Najm I, Lüders H. The localizing value of ictal EEG in focal epilepsy. Neurology 2001;57(11):2022–2028 9. Guaranha MS, Garzon E, Buchpiguel CA, Tazima S, Yacubian EM, Sakamoto AC. Hyperventilation revisited: physiological effects and efficacy on focal seizure activation in the era of video-EEG monitoring. Epilepsia 2005;46(1):69–75 10. Kudr M, Krsek P, Maton B, et al. Predictive factors of ictal SPECT findings in paediatric patients with focal cortical dysplasia. Epileptic Disord 2013;15(4):383–391 11. Krsek P, Kudr M, Jahodova A, et al. Localizing value of ictal SPECT is comparable to MRI and EEG in children with focal cortical dysplasia. Epilepsia 2013;54(2):351–358 12. Lantz GC. Regional anesthesia for dentistry and oral surgery. J Vet Dent 2003;20(3):181–186 13. Koh S, Jayakar P, Resnick T, Alvarez L, Liit RE, Duchowny M. The localizing value of ictal SPECT in children with tuberous sclerosis complex and refractory partial epilepsy. Epileptic Disord 1999;1(1):41–46 14. Krsek P, Jahodova A, Maton B, et al. Low-grade focal cortical dysplasia is associated with prenatal and perinatal brain injury. Epilepsia 2010;51(12):2440–2448 15. Krsek P, Maton B, Korman B, et al. Different features of histopathological subtypes of pediatric focal cortical dysplasia. Ann Neurol 2008;63(6):758–769 16. Lantz G, Grave de Peralta R, Spinelli L, Seeck M, Michel CM. Epileptic source localization with high density EEG: how many electrodes are needed? Clin Neurophysiol 2003;114(1):63–69 17. Pfund Z, Chugani DC, Juhász C, et al. Evidence for coupling between glucose metabolism and glutamate cycling using FDG PET and 1H magnetic resonance spectroscopy in patients with epilepsy. J Cereb Blood Flow Metab 2000;20(5):871–878 18. Lantz G, Spinelli L, Seeck M, de Peralta Menendez RG, Sottas CC, Michel CM. Propagation of interictal epileptiform activity can lead to erroneous source localizations: a 128-channel EEG ­mapping study. J Clin Neurophysiol 2003;20(5):311–319

21. Mirsattari SM, Steven DA, Keith J, Hammond RR. Pathophysiological implications of focal cortical dysplasia of end folium for hippocampal sclerosis. Epilepsy Res 2009;84(2–3):268–272 22. Obeid M, Wyllie E, Rahi AC, Mikati MA. Approach to pediatric epilepsy surgery: State of the art, Part II: approach to s­ pecific epilepsy syndromes and etiologies. Eur J Paediatr Neurol 2009;13(2):115–127 23. Wyllie E, Lachhwani DK, Gupta A, et al. Successful surgery for epilepsy due to early brain lesions despite generalized EEG findings. Neurology 2007;69(4):389–397 24. Lüders H, Dinner DS, Morris HH III, Wyllie E, Godoy J. EEG evaluation for epilepsy surgery in children. Cleve Clin J Med 1989;56(Suppl Pt 1):S53–S61, discussion S79–S83 25. Obeid M, Wyllie E, Rahi AC, Mikati MA. Approach to pediatric epilepsy surgery: state of the art, Part I: general principles and presurgical workup. Eur J Paediatr Neurol 2009;13(2):102–114 26. Mohamed A, Wyllie E, Ruggieri P, et al. Temporal lobe epilepsy due to hippocampal sclerosis in pediatric candidates for epilepsy surgery. Neurology 2001;56(12):1643–1649 27. Loddenkemper T, Cosmo G, Kotagal P, et al. Epilepsy surgery in children with electrical status epilepticus in sleep. Neurosurgery 2009;64(2):328–337, discussion 337 28. Gupta A, Chirla A, Wyllie E, Lachhwani DK, Kotagal P, Bingaman WE. Pediatric epilepsy surgery in focal lesions and generalized electroencephalogram abnormalities. Pediatr Neurol 2007;37(1):8–15 29. Battaglia D, Lettori D, Contaldo I, et al. Seizure semiology of lesional frontal lobe epilepsies in children. Neuropediatrics 2007;38(6):287–291 30. Nordli DR. Varying seizure semiology according to age. Handb Clin Neurol 2013;111:455–460 31. Rezayof A, Zarrindast MR, Sahraei H, Haeri-Rohani A. Involvement of dopamine receptors of the dorsal hippocampus on the acquisition and expression of morphine-induced place preference in rats. J Psychopharmacol 2003;17(4):415–423 32. Gilbert PE, Kesner RP. Localization of function within the dorsal hippocampus: the role of the CA3 subregion in paired-associate learning. Behav Neurosci 2003;117(6):1385–1394 33. Bhardwaj SK, Beaudry G, Quirion R, Levesque D, Srivastava LK. Neonatal ventral hippocampus lesion leads to reductions in nerve growth factor inducible-B mRNA in the prefrontal cortex and increased amphetamine response in the nucleus accumbens and dorsal striatum. Neuroscience 2003;122(3):669–676 34. Michel CM, Murray MM, Lantz G, Gonzalez S, Spinelli L, Grave de Peralta R. EEG source imaging. Clin Neurophysiol 2004;115(10):2195–2222 35. Kun Lee S, Young Lee S, Kim DW, Soo Lee D, Chung CK. Occipital lobe epilepsy: clinical characteristics, surgical outcome, and role of diagnostic modalities. Epilepsia 2005;46(5):688–695 36. Sperli F, Spinelli L, Seeck M, Kurian M, Michel CM, Lantz G. EEG source imaging in pediatric epilepsy surgery: a new perspective in presurgical workup. Epilepsia 2006;47(6):981–990

14  Electroencephalography and Noninvasive Electrophysiological Assessment 37. Sinclair DB, Wheatley M, Snyder T, Gross D, Ahmed N. Posterior resection for childhood epilepsy. Pediatr Neurol 2005;32(4):257–263

60. van den Munckhof B, van Dee V, Sagi L, et al. Treatment of electrical status epilepticus in sleep: a pooled analysis of 575 cases. Epilepsia 2015;56(11):1738–1746

38. Aykut-Bingol C, Bronen RA, Kim JH, Spencer DD, Spencer SS. Surgical outcome in occipital lobe epilepsy: implications for pathophysiology. Ann Neurol 1998;44(1):60–69

61. Hauptman JS, Mathern GW. Surgical treatment of epilepsy associated with cortical dysplasia: 2012 update. Epilepsia 2012;53(Suppl 4):98–104

39. Beleza P, Pinho J. Frontal lobe epilepsy. J Clin Neurosci 2011;18(5):593–600

62. Liu Z, Nagao T, Desjardins GC, Gloor P, Avoli M. Quantitative evaluation of neuronal loss in the dorsal hippocampus in rats with long-term pilocarpine seizures. Epilepsy Res 1994;17(3):237–247

40. Gotman J. Interhemispheric relations during bilateral spikeand-wave activity. Epilepsia 1981;22(4):453–466 41. Kellinghaus C, Lüders HO. Frontal lobe epilepsy. Epileptic Disord 2004;6(4):223–239 42. Salanova V, Andermann F, Rasmussen T, Olivier A, Quesney LF. Parietal lobe epilepsy. Clinical manifestations and outcome in 82 patients treated surgically between 1929 and 1988. Brain 1995;118(Pt 3):607–627 43. Michel CM, Lantz G, Spinelli L, De Peralta RG, Landis T, Seeck M. 128-channel EEG source imaging in epilepsy: clinical yield and localization precision. J Clin Neurophysiol 2004;21(2):71–83 44. Klein KM, Knake S, Hamer HM, Ziegler A, Oertel WH, Rosenow F. Sleep but not hyperventilation increases the sensitivity of the EEG in patients with temporal lobe epilepsy. Epilepsy Res 2003;56(1):43–49 45. Adachi N, Alarcon G, Binnie CD, Elwes RD, Polkey CE, Reynolds EH. Predictive value of interictal epileptiform discharges during non-REM sleep on scalp EEG recordings for the lateralization of epileptogenesis. Epilepsia 1998;39(6):628–632 46. Giorgi FS, Perini D, Maestri M, et al. Usefulness of a simple sleep-deprived EEG protocol for epilepsy diagnosis in de novo subjects. Clin Neurophysiol 2013;124(11):2101–2107 47. Taylor I, Scheffer IE, Berkovic SF. Occipital epilepsies: identification of specific and newly recognized syndromes. Brain 2003;126(Pt 4):753–769 48. Modur PN, Vitaz TW, Zhang S. Seizure localization using broadband EEG: comparison of conventional frequency activity, high-frequency oscillations, and infraslow activity. J Clin Neurophysiol 2012;29(4):309–319 49. Díaz-Negrillo A. Influence of sleep and sleep deprivation on ictal and interictal epileptiform activity. Epilepsy Res Treat 2013;2013:492524 50. Sammaritano M, Gigli GL, Gotman J. Interictal spiking during wakefulness and sleep and the localization of foci in temporal lobe epilepsy. Neurology 1991;41(2 ( Pt 1)):290–297 51. Mirkovic N, Adjouadi M, Yaylali I, Jayakar P. 3D source localization of epileptic foci integrating EEG and MRI data. Brain Topogr 2003;16(2):111–119 52. Lee DS, Lee SK, Kim SK, et al. Late postictal residual perfusion abnormality in epileptogenic zone found on 6-hour postictal SPECT. Neurology 2000;55(6):835–841 53. So EL. Integration of EEG, MRI, and SPECT in localizing the seizure focus for epilepsy surgery. Epilepsia 2000;41(Suppl 3):S48–S54 54. Chugani HT. PET in preoperative evaluation of intractable epilepsy. Pediatr Neurol 1993;9(5):411–413 55. Chugani HT. The role of PET in childhood epilepsy. J Child Neurol 1994;9(Suppl 1):S82–S88 56. Chugani HT, Shields WD, Shewmon DA, Olson DM, Phelps ME, Peacock WJ. Infantile spasms: I. PET identifies focal cortical dysgenesis in cryptogenic cases for surgical treatment. Ann Neurol 1990;27(4):406–413 57. Nickels K, Wirrell E. Electrical status epilepticus in sleep. Semin Pediatr Neurol 2008;15(2):50–60 58. Viñas FC, Zamorano L, Mueller RA, et al. [15O]-water PET and intraoperative brain mapping: a comparison in the localization of eloquent cortex. Neurol Res 1997;19(6):601–608 59. Scholtes FB, Hendriks MP, Renier WO. Cognitive deterioration and electrical status epilepticus during slow sleep. Epilepsy Behav 2005;6(2):167–173

63. Guerrini R, Genton P, Bureau M, et al. Multilobar polymicrogyria, intractable drop attack seizures, and sleep-related electrical status epilepticus. Neurology 1998;51(2):504–512 64. Erbayat Altay E, Fessler AJ, Gallagher M, et al. Correlation of severity of FDG-PET hypometabolism and interictal regional delta slowing in temporal lobe epilepsy. Epilepsia 2005;46(4):573–576 65. Mathern GW. Epilepsy surgery patients with cortical dysplasia: present and future therapeutic challenges. Neurology 2009;72(3):206–207 66. Cepeda C, André VM, Flores-Hernández J, et al. Pediatric cortical dysplasia: correlations between neuroimaging, electrophysiology and location of cytomegalic neurons and balloon cells and glutamate/GABA synaptic circuits. Dev Neurosci 2005;27(1):59–76 67. Cepeda C, Hurst RS, Flores-Hernández J, et al. Morphological and electrophysiological characterization of abnormal cell types in pediatric cortical dysplasia. J Neurosci Res 2003;72(4):472–486 68. Tassinari CA, Rubboli G. Cognition and paroxysmal EEG activities: from a single spike to electrical status epilepticus during sleep. Epilepsia 2006;47(Suppl 2):40–43 69. Cammarota M, Bevilaqua LR, Kerr D, Medina JH, Izquierdo I. Inhibition of mRNA and protein synthesis in the CA1 region of the dorsal hippocampus blocks reinstallment of an extinguished conditioned fear response. J Neurosci 2003;23(3):737–741 70. Hahn B, Shoaib M, Stolerman IP. Involvement of the prefrontal cortex but not the dorsal hippocampus in the attention-­ enhancing effects of nicotine in rats. Psychopharmacology (Berl) 2003;168(3):271–279 71. Kelley SA, Kossoff EH. How effective is the ketogenic diet for electrical status epilepticus of sleep? Epilepsy Res 2016;127:339–343 72. Thornton R, Laufs H, Rodionov R, et al. EEG correlated functional MRI and postoperative outcome in focal epilepsy. J Neurol ­Neurosurg Psychiatry 2010;81(8):922–927 73. Kim DW, Lee SK, Nam H, et al. Epilepsy with dual pathology: ­surgical treatment of cortical dysplasia accompanied by hippocampal sclerosis. Epilepsia 2010;51(8):1429–1435 74. Harvey AS, Cross JH, Shinnar S, Mathern GW; ILAE Pediatric ­Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. ­Epilepsia 2008;49(1):146–155 75. Blumcke I, Spreafico R, Haaker G, et al; EEBB Consortium. Histopathological findings in brain tissue obtained during epilepsy surgery. N Engl J Med 2017;377(17):1648–1656 76. Lerner JT, Salamon N, Hauptman JS, et al. Assessment and ­surgical outcomes for mild type I and severe type II cortical dysplasia: a critical review and the UCLA experience. Epilepsia 2009;50(6):1310–1335 77. Lee I, Kesner RP. Time-dependent relationship between the dorsal hippocampus and the prefrontal cortex in spatial memory. J Neurosci 2003;23(4):1517–1523 78. van der Stelt HM, Broersen LM, Olivier B, Westenberg HG. Effects of dietary tryptophan variations on extracellular serotonin in the dorsal hippocampus of rats. Psychopharmacology (Berl) 2004;172(2):137–144 79. Okuyaz C, Aydin K, Gücüyener K, Serdaroğlu A. Treatment of ­electrical status epilepticus during slow-wave sleep with highdose corticosteroid. Pediatr Neurol 2005;32(1):64–67

125

126

IIb  Preoperative Electrophysiological Assessment 80. Veggiotti P, Pera MC, Teutonico F, Brazzo D, Balottin U, Tassinari CA. Therapy of encephalopathy with status epilepticus during sleep (ESES/CSWS syndrome): an update. Epileptic Disord 2012;14(1):1–11 81. Ferbinteanu J, Ray C, McDonald RJ. Both dorsal and ventral ­hippocampus contribute to spatial learning in Long-Evans rats. ­Neurosci Lett 2003;345(2):131–135 82. Wang WT, Han D, Zou ZY, Zeng J. [Epileptiform activity of the anterior dorsal hippocampal network induced by acute tetanization of the right posterior dorsal hippocampus of the rat] Sheng Li Xue Bao 2003;55(3):339–348 83. White NM, Holahan MR, Goffaux P. Involuntary, unreinforced (pure) spatial learning is impaired by fimbria-fornix but not by dorsal hippocampus lesions. Hippocampus 2003;13(3):324–333 84. Blümcke I, Thom M, Aronica E, et al. The clinicopathologic spectrum of focal cortical dysplasias: a consensus classification proposed by an ad hoc Task Force of the ILAE Diagnostic Methods Commission. Epilepsia 2011;52(1):158–174 85. Patry G, Lyagoubi S, Tassinari CA. Subclinical “electrical status epilepticus” induced by sleep in children. A clinical and electroencephalographic study of six cases. Arch Neurol 1971;24(3):242–252 86. Jayakar PB, Seshia SS. Electrical status epilepticus during slowwave sleep: a review. J Clin Neurophysiol 1991;8(3):299–311 87. Francois D, Roberts J, Hess S, Probst L, Eksioglu Y. Medical management with diazepam for electrical status epilepticus during slow wave sleep in children. Pediatr Neurol 2014;50(3): 238–242 88. Inutsuka M, Kobayashi K, Oka M, Hattori J, Ohtsuka Y. Treatment of epilepsy with electrical status epilepticus during slow sleep and its related disorders. Brain Dev 2006;28(5):281–286 89. Sakata-Haga H, Sawada K, Ohta K, Cui C, Hisano S, Fukui Y. Adverse effects of maternal ethanol consumption on development of dorsal hippocampus in rat offspring. Acta Neuropathol 2003;105(1):30–36 90. Jeong A, Strahle J, Vellimana AK, Limbrick DD Jr, Smyth MD, ­Bertrand M. Hemispherotomy in children with electrical status epilepticus of sleep. J Neurosurg Pediatr 2017;19(1):56–62 91. Sammaritano M, de Lotbinière A, Andermann F, Olivier A, Gloor P, Quesney LF. False lateralization by surface EEG of seizure onset in patients with temporal lobe epilepsy and gross focal cerebral lesions. Ann Neurol 1987;21(4):361–369 92. Bast T, Zhang WN, Feldon J. Dorsal hippocampus and classical fear conditioning to tone and context in rats: effects of local NMDA-receptor blockade and stimulation. Hippocampus 2003;13(6):657–675 93. Ray A, Tao JX, Hawes-Ebersole SM, Ebersole JS. Localizing value of scalp EEG spikes: a simultaneous scalp and intracranial study. Clin Neurophysiol 2007;118(1):69–79 94. Brodbeck V, Spinelli L, Lascano AM, et al. Electroencephalographic source imaging: a prospective study of 152 operated epileptic patients. Brain 2011;134(Pt 10):2887–2897 95. Towart LA, Alves SE, Znamensky V, Hayashi S, McEwen BS, ­Milner TA. Subcellular relationships between cholinergic terminals and estrogen receptor-alpha in the dorsal hippocampus. J Comp Neurol 2003;463(4):390–401 96. Holmes MD, Tucker DM, Quiring JM, Hakimian S, Miller JW, Ojemann JG. Comparing noninvasive dense array and

i­ ntracranial electroencephalography for localization of seizures. Neurosurgery 2010;66(2):354–362 97. Centeno M, Tierney TM, Perani S, et al. Combined electroencephalography-functional magnetic resonance imaging and electrical source imaging improves localization of pediatric focal epilepsy. Ann Neurol 2017;82(2):278–287 98. Sinclair DB, Aronyk KE, Snyder TJ, et al. Pediatric epilepsy surgery at the University of Alberta: 1988–2000. Pediatr Neurol 2003;29(4):302–311 99. Vanhatalo S, Holmes MD, Tallgren P, Voipio J, Kaila K, Miller JW. Very slow EEG responses lateralize temporal lobe seizures: an evaluation of non-invasive DC-EEG. Neurology 2003;60(7):1098–1104 100. Miller JW, Kim W, Holmes MD, Vanhatalo S. Ictal localization by source analysis of infraslow activity in DC-coupled scalp EEG recordings. Neuroimage 2007;35(2):583–597 101. Rodin E, Constantino T, van Orman C, House P. EEG infraslow activity in absence and partial seizures. Clin EEG Neurosci 2008;39(1):12–19 102. Modur PN. High frequency oscillations and infraslow activity in epilepsy. Ann Indian Acad Neurol 2014;17(Suppl 1): S99–S106 103. Tao JX, Baldwin M, Hawes-Ebersole S, Ebersole JS. Cortical substrates of scalp EEG epileptiform discharges. J Clin Neurophysiol 2007;24(2):96–100 104. K  umar A, Asano E, Chugani HT. α-[11C]-methyl-L-tryptophan PET for tracer localization of epileptogenic brain regions: clinical studies. Biomarkers Med 2011;5(5):577–584 105. Jirsch JD, Urrestarazu E, LeVan P, Olivier A, Dubeau F, Gotman J. High-frequency oscillations during human focal seizures. Brain 2006;129(Pt 6):1593–1608 106. A  waad Y, Shamato H, Chugani H. Hemidystonia improved by baclofen and PET scan findings in a patient with glutaric aciduria type I. J Child Neurol 1996;11(2):167–169 107. K  umar A, Muzik O, Chugani D, Chakraborty P, Chugani HT. PET-derived biodistribution and dosimetry of the benzodiazepine receptor-binding radioligand (11)C-(R)-PK11195 in children and adults. J Nucl Med 2010;51(1):139–144 108. K  hanna S, Chugani HT, Messa C, Curran JG. Corpus callosum agenesis and epilepsy: PET findings. Pediatr Neurol 1994;10(3):221–227 109. Engel J Jr, Henry TR, Risinger MW, Sutherling WW, Chugani HT. PET in relation to intracranial electrode evaluations. Epilepsy Res Suppl 1992;5:111–120 110. Juhász C, Nagy F, Muzik O, Watson C, Shah J, Chugani HT. [11C] Flumazenil PET in patients with epilepsy with dual pathology. Epilepsia 1999;40(5):566–574 111. Peng K, Nguyen DK, Tayah T, et al. fNIRS-EEG study of focal interictal epileptiform discharges. Epilepsy Res 2014;108(3):491–505 112. Pouliot P, Tran TP, Birca V, et al. Hemodynamic changes during posterior epilepsies: an EEG-fNIRS study. Epilepsy Res 2014;108(5):883–890 113. Bourel-Ponchel E, Mahmoudzadeh M, Delignières A, Berquin P, Wallois F. Non-invasive, multimodal analysis of cortical activity, blood volume and neurovascular coupling in infantile spasms using EEG-fNIRS monitoring. Neuroimage Clin 2017;15:359–366

15

  Invasive Electrophysiological Monitoring Prasanna Jayakar, Ann Hyslop, and Ian Miller

Summary The indications, limitations, technical considerations, and interpretation of invasive electrophysiological monitoring (IEM) in the surgical epilepsy evaluation are discussed in this chapter. Authors identify patient characteristics in which invasive monitoring, via stereo-electroencephalography (SEEG) or placement of depth, grid or strip electrodes, leads to more accurate identification of the epileptogenic region (ER), minimization of cortical resection, and successful removal of seizure foci in close proximity to eloquent cortex. An in-depth discussion of the intricacies of intracranial EEG interpretation is included and will help readers identify patterns of activity that may differentiate the ictal onset zone from an irritative zone or region of static functional abnormality. Keywords:  invasive monitoring, intracranial EEG, subdural EEG, electrode implantation, stereo-EEG, SEEG

„„ Introduction The primary goal of the presurgical evaluation for intractable epilepsy is to accurately define the ER and its relationship to eloquent cortex. The ER is defined as the minimum amount of tissue that must be resected to ameliorate the target seizure type. This critical mass of tissue is viewed as a function of the region of seizure onset, seizure propagation patterns, the areas that could become epileptogenic later, and the underlying structural lesion and functional abnormality. As in adults, surgical strategies in childhood are guided by semiology, structural and functional imaging abnormalities, and neurophysiological data. This task can be daunting given the heterogeneity of etiopathological substrates and maturational factors that influence the clinical presentation and investigative findings.1 The role of IEM in the evaluation of childhood epilepsy has evolved. With advances in magnetic resonance imaging (MRI), single-photon emission computed tomography (SPECT), and positron emission tomography (PET) imaging, presurgical evaluations and development of successful surgical plans can be accomplished in many children without invasive monitoring. In addition, the increasing use of magnetoencephalography, three-dimensional EEG dipole source localization algorithms, and the use of functional MRI in delineating eloquent cortex has further diminished the need for IEM. In the International League Against Epilepsy (ILAE) study of 543 children, the 20

participating centers worldwide reported using IEM in just over 25% of their surgeries.2 However, this trend has been offset as advanced pediatric centers pursue surgical intervention strategies in increasingly complex cases. This chapter critically examines the advancing role of IEM in presurgical evaluation.

„„ Pragmatic Considerations Since IEM is inherently more costly and morbid than the same resection without IEM, its role must be driven by pragmatic considerations. During the noninvasive portion of the presurgical evaluation, it is prudent to consider whether a more precise definition of ER using IEM will alter the ultimate surgical resection and outcome. IEM may be of little use if noninvasive studies support widespread epileptogenic dysfunction that cannot safely be resected. For example, IEM is unlikely to document discrete seizure onsets when patients with normal imaging studies present with spasms or diffuse ictal patterns on the scalp EEG or in patients with multiple subcortical nodular heterotopias, large infiltrative lesions, or extensive multilobar cortical dysplasia. Thus, the risks imparted by IEM may exceed benefit if the goals of surgery are mainly palliative. While IEM has limitations in sampling and interpretation, our experience suggests that children who are neurodevelopmentally intact, in which a restricted ER is suspected by structural or functional imaging or clinical semiology, with an otherwise normal scalp EEG are likely to accrue the greatest benefit from well-planned IEM. In such cases, the preceding noninvasive evaluation should provide sufficient information to generate a reasonable hypothesis (or hypotheses) regarding the location of the ER and its proximity to eloquent cortex such that data acquired from IEM results in better localization of the ER and minimization of the size of the ultimate resection. Lastly, should an intraoperative electrocorticography (ECoG) reveal near-continuous focal electrographic ictal activity in a previously suspected ER in non-eloquent cortex, benefits of extraoperative IEM are unlikely to outweigh its risks. Extraoperative IEM is often required to define eloquent cortex when intraoperative functional mapping is not possible or yields suboptimal data. Language mapping can rarely be performed intraoperatively in younger children or older children with anxiety or language disorders. Somatosensory responses to median nerve stimulation are often used to successfully identify the central sulcus intraoperatively, but direct electrical stimulation of

128

IIb  Preoperative Electrophysiological Assessment the precentral gyrus for precise definition of critical motor cortex may yield inaccurate data due to effects of general anesthesia. Lastly, family dynamics play an important role when deciding the best surgical approach for each child. Both parents (or guardians) must be educated on the risks and limitations of IEM before the technique is recommended. Psychological support for the caregivers and the child may be needed during intraand ­ operative-IEM and, especially, during attempts to map critical ­cortex.

continues to play a significant role in this patient subgroup, especially when functional imaging data are inconclusive, as removal of all cortex harboring epileptogenic activity identified by IEM often achieves seizure freedom. In a recent series of 25 MRI-negative patients who underwent small resections of cortical tissue, all but one had IEM performed and 60% were seizure free at 2-year follow-up and 80% of the patients experienced reduction in seizure frequency.6

Multiple or Large Structural Lesions

„„ Indications Although IEM usage is guided primarily by the availability of resources at each center, the ILAE has proposed general recommendations for its role in presurgical evaluation.3,​4 All children should undergo video-EEG and brain MRI; those with specific types of epileptogenic MRI lesions who have convergent clinical semiology and scalp EEG data require no further testing. In all others, functional (SPECT and/or PET) imaging, three-dimensional spike source localization, and functional MRI for identification of language and motor cortices are necessary. Multidisciplinary case review with experienced colleagues in the practices of neurology, neurosurgery, neuroradiology, and neuropsychology is recommended. Guided by the pragmatic considerations discussed previously, IEM is generally recommended for the following indications: inconclusive preoperative data; divergent preoperative data; and ER in proximity to eloquent cortex (see Table 15.1).

„„ Inconclusive Preoperative Data Normal or Nonspecific Structural Neuroimaging Findings Despite advances in MRI, many children with localization-­ related intractable epilepsy have normal scans; approximately one in every four children in the ILAE series did not show a definite lesion on MRI.2 In a series of 102 patients with no causative lesion identified on brain MRI, 80 underwent extraoperative IEM, while the remainder were adequately localized based on scalp EEG, functional imaging, and intraoperative EcoG.5 IEM

In general, presence of a discrete structural abnormality is regarded as a reliable marker of the ER and decreases the utilization of IEM. However, there are many documented failures after lesionectomy in children,7,​8 partly because lesional epilepsies are typically not due to homogeneous substrates. Whereas developmental tumors, hippocampal sclerosis, lowflow vascular lesions, or Sturge–Weber syndrome can often be successfully surgically treated after noninvasive evaluation, children with ill-defined cortical dysplasias or multiple lesions, such as those seen in tuberous sclerosis, often have complex and rapid seizure propagation patterns that make interpretation of the scalp EEG and functional imaging data difficult. IEM often helps clarify ambiguities of seizure origin and propagation, thus facilitating surgery in these difficult cases.8,​9,​10 In rare cases, data from a noninvasive evaluation may suggest that ictal onset occurs from a restricted region within a widespread lesion (such as an extensive cortical dysplasia or polymicrogyria); IEM may provide data leading to a successful focal resection and obviate the need for a hemispherectomy in a motorically intact child.11 Fig. 15.1 shows the appearance of overriding fast frequencies in a diffuse ictal onset from IEM of a SEEG from a 10-yearold female with a large peri-insular focal cortical dysplasia and multiple daily seizures. Following resection of only the tissue from which the fast frequencies emanated, she is now 2 years seizure free despite the large amount of residual focal cortical dysplasia.

Divergent Preoperative Data The definition of divergence varies between epilepsy centers; we regard divergence when the clinical semiology, EEG, and

Table 15.1  General indications for invasive electrophysiological monitoring

Indication

Clinical scenarios

To define the ER precisely when noninvasive data is inconclusive

Common scenarios include rapidly “generalized” seizures such as those seen in early childhood, differentiating regional versus l­obar or multilobar involvement (e.g., temporal vs. temporal plus epilepsy), determining the side of mesial temporal onset, mesial versus neocortical temporal involvement, “dual” temporal lobe pathology, defining deep seated or interhemispheric cortical sources especially those related to occult dysplasia not evident on MRI scans

To resolve divergence of noninvasive data pointing to two or more regions

Divergence is not uncommon; scenarios particularly prone include bilateral mesial temporal foci, large lesions such as encephalomalacia, multiple lesions such as those in tuberous sclerosis or nodular heterotopia

To map eloquent cortical function precisely

ER encroaching or involving eloquent cortex. Unlike acquired tumors or early acquired atrophic/ gliotic lesions that tend to displace function, developmental substrates often retain eloquent function and may manifest atypical representations

Secondary ­indications

To further corroborate the ER or provide information of prognostic value, to selectively ablate active regions using radiofrequency thermal coagulation

Abbreviation: ER, epileptogenic region.

15  Invasive Electrophysiological Monitoring

Fig. 15.1  An ictal onset from SEEG of a 10-year-old with a large peri-insular focal cortical dysplasia and multiple daily seizures. Seven depth electrodes were placed orthogonally in the anterior (AF), middle (MF), and posterior (PF) frontal operculum and adjacent insula, the amygdala (AMG), the middle (MT) and posterior (PT) temporal operculum and adjacent insula and, lastly, in an anteriorto-posterior trajectory within the insula (S). The SEEG shows diffuse attenuation of nearly all contacts and immediate emergence of faster frequencies primarily over the posterior frontal operculum and insula which, when resected, resulted in seizure-freedom despite the intended incomplete resection of this ILAE type 2a focal cortical dysplasia.

functional/structural imaging data indicate that the ER is in separate regions. This divergence is likely, in part, due to limitations of scalp EEG and functional imaging techniques in defining complex ictal networks, propagation patterns, and dipole localization in deeply seated ERs. In these cases, well-planned IEM will often correctly identify the true ER and lead to a successful focal resection. Careful planning of electrode positioning is vital in these cases since the divergence may arise from complex and rapid interaction between noncontiguous cortical sites, of which must be adequately sampled. When divergence occurs in the context of large or deep-seated lesions, a combination of subdural and strategically placed depth electrodes is recommended.

Epileptogenic Region in Proximity to Eloquent Cortex In our series of discrete peri-Rolandic foci, aggressive resections tailored to the ER and eloquent motor cortex led to seizure freedom in more than half the cases, even in the absence of a structural lesion.10 In some cases, a calculated decision to remove part of the motor cortex revealing face or proximal limb function enhances the chances of success without incurring risks of significant clinical deficits. Likewise, in patients with occipital foci who have intact visual fields, accurate demarcation of the ER over the occipital convexity or base may allow a restricted corticectomy preserving most of the calcarine cortex and visual pathways, making IEM a worthwhile endeavor.12 The issues surrounding the need to identify and preserve language cortex are more complex. Because language cortex exhibits greater degrees of reorganization under the age

Fig. 15.2  Subdural electrodes (a) over the left temporal region in a 4-year-old patient with cortical dysplasia. Subdural electroencephalography recording (b), showing ictal onset at contacts 3 and 4 of the anterior temporal polar strip, with early involvement of the superior temporal convexity (contacts: 19–20), which were subsequently shown to be critical language areas on functional mapping. Invasive monitoring helped tailor the resection.

of 5, many centers opt for more aggressive large resections in the hope of forcing language transfer. Our presurgical evaluation strategy is driven by the intent to preserve predestined ­language sites unless they are involved at ictal onset. Thus, we used IEM to tailor resections, even in the very young, to maximize language outcomes without compromising post­ ­ operative s­ eizure freedom (Fig. 15.2).

„„ Technical Aspects Spatial Coverage In general, an attempt must be made to place enough electrodes so that there is adequate sampling of the suspected ictal onset zone and demonstration of the margins of the ER. This task is made more challenging by the restricted electrical field sampling of a typical subdural electrode13 and the number of ­electrodes that can be placed in children. When the ER

129

130

IIb  Preoperative Electrophysiological Assessment is expected to encroach on critical cortex, additional coverage must be provided to perform functional mapping. Bilateral placements, commonly performed in adults, are rarely required in children, but can used to regionalize epileptogenic foci with SEEG. Stereo-EEG offers a less invasive method of IEM which has prompted exploration of less well-defined ERs in younger children.14

Type of Electrodes Invasive electrodes are usually made of platinum and are configured for subdural or depth placement. Subdural ­electrodes spaced 5 to 10 mm apart and configured as strips (4–8 contacts) or grids (20–64 contacts) are most suited to c­ over large areas of the neocortical convexity and basal and interhemispheric surfaces.15,​16 Depth electrodes (4–8 contacts; 1.2 mm diameter) used to penetrate parenchyma are critical for monitoring deep-seated regions such as depth of sulcus lesions, the insula, operculae, and the mesial temporal regions, and can be employed in conjunction with strip and grid electrodes following craniotomy or, in contrast, placed through burr holes with stereotactic guidance (4–8 contacts; reduced 0.8 mm diameter). They also provide the means to accurately map eloquent cortex. Using a combination of subdural and depth electrodes results in comprehensive coverage of both the cortical surface and deep l­ ocations.

Surgical Insertion Subdural grid electrodes are implanted under direct observation after craniotomy, although strips may be placed via burr holes. Depth electrodes are inserted either through a craniotomy or under stereotactic MRI- or CT-guided techniques that allow accurate placement at target sites, including symmetric bilateral positioning in homotopic areas of interest. Implantation may be guided by a stereotactic frame but this technique has limited application in the younger child where the calvarium is thin and skull framing is contraindicated. The exact location of the electrodes can be defined extraoperatively on MRI or highresolution CT scans coregistered to the MRI postoperatively.

„„ Risks Dedicated nursing and social intervention facilitates the perioperative care in children undergoing IEM. Prophylactic steroids help minimize the risk of reaction to the implant. Implantation is generally well tolerated, but complications, including wound infection, cerebrospinal fluid leak, intracranial bleeding, or symptomatic pneumocephalus, have all been reported.17,​18,​19 Depth placements may lead to intracerebral microhemorrhage20; subdural electrodes may cause local inflammatory reactions. Risks tend to be higher in children who undergo reoperation19; permanent neurological deficit or death associated with IEM is very rare.

„„ Recording Current digital systems have high sampling rate capability, at 500 to 1,000 Hz, allowing detection of fast frequencies that may provide useful additional information. Seizure capture may be

augmented by withdrawal of antiepileptic medications; albeit rare, due consideration should be given to the chances of ­activating atypical seizures. Spontaneous seizures are usually captured over 4 to 10 days; longer periods of up to 1 month may occasionally be required. Generally, 3 to 10 seizures are considered adequate, although multiple factors may influence the confidence of the reader, including stereotypy of onset and evolution, and their convergence with other data. Once the capture of spontaneous seizures is deemed complete, we reinstate full medication before attempting functional mapping via direct electrical cortical stimulation.

„„ Interpretation The ambiguities and complexity of intracranial EEG interpretation are well-recognized and guidelines for its use in defining the ER are often based on empirical criteria.10,​13,​21,​22,​23,​24

Ictal Onset Zone Defining the region of seizure onset accurately is the strongest justification for IEM, especially in children with multifocal MRI and interictal EEG abnormalities. As illustrated in Fig. 15.3, periodic discharges were seen independently over several regions, but ictal onset consistently occurred from a single focus. The onset zone may be regarded as the region showing the initial transformation from the interictal state: the collective area revealing a group of patterns commonly observed during the initial phase of the seizure including bursts of focal fast activity, spike/polyspikes, runs of spikes, or electrodecrement.10,​25,​26 Rhythmic slow delta, theta, and alpha range frequencies occurring discretely as the first change are considered significant but may not be included in the ictal onset zone if they are observed after onset, even when they achieve high amplitude. Very high frequencies27 or slow direct current shifts have also been observed but may be missed with conventional filter settings. While more study is needed, recent studies suggest that focal high frequency oscillations seen during IEM indicate highly epileptogenic tissue that, when resected, results in reduction of seizure burden postoperatively.28 The types of epileptogenic patterns appearing at or after ictal onset vary between patients and at times from seizure to seizure in the same subject. Thus, objective definition of the ictal onset zone can be challenging, especially in patients with rapid propagation. In general, the longer that ictal activity remains focal within an ER after onset, the greater the likelihood it represents the true ictal onset zone and the more favorable an outcome will be following its resection.25 In our experience, more than half the children with discrete ictal onset on invasive studies were rendered seizure free as opposed to fewer than 10% when the onset was diffuse.29 In some cases, presumably because seizures arise from lesional or severely damaged cortex, the ictal onset region does not generate a typical robust ictal sequence but merely acts as a trigger, activating remote, healthier areas.10 The ictal trigger may be characterized only by a subtle transformation from the ongoing interictal state, such as alteration of the frequency, morphology, and distribution of interictal spikes; alteration of the frequency or content of an interictal burst suppression; or merely a further attenuation of the background. This finding

15  Invasive Electrophysiological Monitoring may be appreciated only after the same change is consistently observed during ictal onset in several seizures. From a practical standpoint, failure to recognize the subtle transformation as the true ictal onset zone may result in false localization or the appearance of data nonconvergence. Fig. 15.4 shows an ictal onset in a 3-year-old child implanted in the lateral and mesial posterior frontal lobe, characterized primarily by drop out of focal interictal spiking and emergence of focal fast frequencies in a region over the lateral frontal convexity. After resection of cortex in the mesial posterior frontal region, she is seizure free.

Irritative Zone To maximize the chances of seizure freedom, the surgical resection should also include cortical regions exhibiting significant potential epileptogenicity. However, the potential risk of seizures attributed to the observed pattern(s) must be weighed against the feasibility and safety of extending the resection beyond the ictal onset zone. Interictal spikes and sharp waves, the patterns most representative of the irritative zone, show considerable variation in morphology, frequency, and distribution, and may persist after resection without adversely affecting outcome.13,​22,​23 They are thus generally considered more significant if they are consistently unifocal or appear in rhythmic runs. Quantitative assessment of spike parameters, such as average frequency or amplitude, may help interpretation.30 In some children, the discharges may be almost continuous and appear in burst or repetitive recruiting and de-recruiting rhythms akin to electrographic seizures. Such patterns are regarded as a hallmark of cortical dysplasia,7 but in our experience, may be seen in other pathological substrates as well.31 Other patterns signifying potential epileptogenicity, albeit infrequent, may also be used in planning the extent of resection. These include focal intermittent bursts of fast activity initially described on the scalp EEG32 and intraictal secondarily activated foci in which, during the course of a seizure, a well-lo-

Fig. 15.3  Subdural electrodes (a) in a child with multifocal MRI abnormalities. The subdural interictal recording (b) revealed independent areas of semiperiodic epileptiform discharges. The discharges over the superior part of the grid (electrodes: 3 and 4) shows gradual attenuation: a subtle ictal transformation that 10 seconds later (c) builds up into a robust seizure discharge. This area corresponded to primary motor cortex on functional mapping. The periodic discharges over inferior region (around electrode 51) remain virtually unchanged during the seizure. The accurate identification of the ictal onset helped make the critical decision to resect the primary motor area.

Fig. 15.4  A stereotyped ictal onset in a MRI-negative 3-year-old with multiple nightly seizures not localizable on scalp EEG. She was implanted in the lateral and mesial posterior frontal lobe, and onset was characterized primarily by drop out of interictal spiking followed by focal attenuation in the mesial frontal lobe, at contacts A10–14. Emergence of focal fast frequencies is seen in a region over the lateral frontal convexity at ictal onset, at contacts G3–6, but indicates propagation to remote, healthier tissue. After resection of cortex in the mesial posterior frontal region (ILAE type 1c) at contacts A10–14, she is seizure free.

131

132

IIb  Preoperative Electrophysiological Assessment calized, independent ictal sequence develops at propagated sites and outlasts the primary seizure sequence.33 In our series, three quarters of secondarily activated foci were documented as being capable of generating independent ­seizures. The role of responses to electrical or pharmacologic provocation is not well studied in children. In our experience with electrical stimulation, the localization value is higher when the early manifestations of provoked seizures are similar to the patient’s habitual auras; late manifestations generally tend to be less reliable. We also do not find afterdischarge thresholds to be useful in defining the ER.

Functional Deficit Zone Burst suppression activity and focal attenuation of fast activity in the background are not conventionally regarded to be epileptiform, but they closely correlate with other markers of the ER. Both patterns remain consistent over time; the focality of beta attenuation can be accentuated by administering drugs

that activate fast frequencies. Polymorphic slowing, by contrast, generally shows considerable temporal variability reflecting nonspecific reactions to anesthesia or electrode implantation and is unlikely to be useful. Postictal background abnormalities, which are shown to be reliable on the scalp EEG, have not been analyzed on invasive studies. Similarly, the role of absent or diminished evoked responses is not established.

„„ Conclusion The role of invasive monitoring in the presurgical evaluation of children continues to evolve as our experience with multimodal noninvasive techniques increases. While improved noninvasive presurgical evaluation techniques have allowed more children to undergo successful resection without IEM, intracranial recording continues to increase the overall number of children eligible for surgical intervention, minimize the size of cortical resections, and decrease the likelihood of postoperative functional deficit.

References 1. Cross JH, Jayakar P, Nordli D, et al; International League against Epilepsy, Subcommission for Paediatric Epilepsy Surgery. Commissions of Neurosurgery and Paediatrics. Proposed criteria for referral and evaluation of children for epilepsy surgery: recommendations of the subcommission for pediatric epilepsy surgery. Epilepsia 2006;47(6):952–959

11. Duchowny M, Jayakar P, Resnick T, et al. Epilepsy surgery in the first three years of life. Epilepsia 1998;39(7):737–743

2. Harvey AS, Cross JH, Shinnar S, Mathern GW; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epilepsia 2008;49(1):146–155

13. Gloor P. Contributions of electroencephalography and electrocorticography to the neurosurgical treatment of the epilepsies. Adv Neurol 1975;8:59–105

3. Jayakar P, Gaillard WD, Tripathi M, Libenson MH, Mathern GW, Cross JH; Task Force for Paediatric Epilepsy Surgery, Commission for Paediatrics, and the Diagnostic Commission of the International League Against Epilepsy. Diagnostic test utilization in evaluation for resective epilepsy surgery in children. Epilepsia 2014;55(4):507–518 4. Jayakar P, Gotman J, Harvey AS, et al. Diagnostic utility of invasive EEG for epilepsy surgery: indications, modalities, and techniques. Epilepsia 2016;57(11):1735–1747 5. Jayakar P, Dunoyer C, Dean P, et al. Epilepsy surgery in patients with normal or nonfocal MRI scans: integrative strategies offer long-term seizure relief. Epilepsia 2008;49(5):758–764 6. Hyslop A, Miller I, Bhatia S, Resnick T, Duchowny M, Jayakar P. Minimally resective epilepsy surgery in MRI-negative children. Epileptic Disord 2015;17(3):263–274 7. Palmini A, Andermann F, Olivier A, et al. Focal neuronal migration disorders and intractable partial epilepsy: a study of 30 patients. Ann Neurol 1991;30(6):741–749 8. Holthausen, H. Teixeira V, Tuxhorn I, et al. Epilepsy surgery in children and adolescents with focal cortical dysplasia. In: Tuxhorn I, Holthausen H, Boenigk H, eds. Paediatric Epilepsy Syndromes and Their Surgical Treatment. London: John Libbey; 1997:199–215 9. Palmini A, Gambardella A, Andermann F, et al. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 1995;37(4):476–487 10. Jayakar P. Invasive EEG monitoring in children: when, where, and what? J Clin Neurophysiol 1999;16(5):408–418

12. Kuzniecky R, Gilliam F, Morawetz R, Faught E, Palmer C, Black L. Occipital lobe developmental malformations and ­epilepsy: clinical spectrum, treatment, and outcome. Epilepsia 1997;38(2):175–181

14. Taussig D, Chipaux M, Lebas A, et al. Stereo-electroencephalography (SEEG) in 65 children: an effective and safe diagnostic method for pre-surgical diagnosis, independent of age. Epileptic Disord 2014;16(3):280–295 15. Wyllie E, Lüders H, Morris HH III, et al. Subdural electrodes in the evaluation for epilepsy surgery in children and adults. Neuropediatrics 1988;19(2):80–86 16. Nespeca M, Wyllie E, Lüders H, Rothner D, et al. EEG recording and functional localization studies with subdural electrodes in infants and young children. J Epilepsy 1990;3:107–117 17. Burneo JG, Steven DA, McLachlan RS, Parrent AG. Morbidity associated with the use of intracranial electrodes for epilepsy surgery. Can J Neurol Sci 2006;33(2):223–227 18. Johnston JM Jr, Mangano FT, Ojemann JG, Park TS, Trevathan E, Smyth MD. Complications of invasive subdural electrode monitoring at St. Louis Children’s Hospital, 1994–2005. J Neurosurg 2006;105(5, Suppl):343–347 19. Musleh W, Yassari R, Hecox K, Kohrman M, Chico M, Frim D. Low incidence of subdural grid-related complications in prolonged pediatric EEG monitoring. Pediatr Neurosurg 2006;42(5):284–287 20. Taussig D, Lebas A, Chipaux M, et al. Stereo-electroencephalography (SEEG) in children surgically cured of their epilepsy. Neurophysiol Clin 2016;46(1):3–15 21. Lieb JP, Joseph JP, Engel J Jr, Walker J, Crandall PH. Sleep state and seizure foci related to depth spike activity in patients with temporal lobe epilepsy. Electroencephalogr Clin Neurophysiol 1980;49(5)(–)(6):538–557

15  Invasive Electrophysiological Monitoring 22. Luders H. Lesser RP, Dinner DS, et al. Commentary: chronic intracranial recording and stimulation with subdural electrodes. In: Engel J Jr, ed. Surgical Treatment of the Epilepsies. New York, NY: Raven Press; 1997:297–321 23. Ajmone-Marsan C. Chronic intracranial recording and electrocorticography. In: Daly DD, Pedley TA, eds. Current Practice of Clinical Electroencephalography. New York, NY: Raven press; 1990:535–560 24. Jayakar P, Duchowny M, Resnick TJ, Alvarez LA. Localization of seizure foci: pitfalls and caveats. J Clin Neurophysiol 1991;8(4):414–431 25. Ikeda A, Terada K, Mikuni N, et al. Subdural recording of ictal DC shifts in neocortical seizures in humans. Epilepsia 1996;37(7):662–674 26. Schiller Y, Cascino GD, Busacker NE, Sharbrough FW. Characterization and comparison of local onset and remote propagated electrographic seizures recorded with intracranial electrodes. Epilepsia 1998;39(4):380–388 27. Fisher RS, Webber WR, Lesser RP, Arroyo S, Uematsu S. High-­ ­ frequency EEG activity at the start of seizures. J Clin ­Neurophysiol 1992;9(3):441–448

28. Frauscher B, Bartolomei F, Kobayashi K, et al. High-frequency oscillations: The state of clinical research. Epilepsia 2017; 58(8):1316–1329 29. Whiting SE, Jayakar P, Duchowny M, et al. The utility of subdural EEG patterns to define the epileptogenic zone in children with cortical dysplasia. American Epilepsy Society Proceedings. ­Epilepsia 1998;39 (Supp 6): 65 30. Asano E, Muzik O, Shah A, et al. Quantitative interictal subdural EEG analyses in children with neocortical epilepsy. Epilepsia 2003;44(3):425–434 31. Turkdogan D, Duchowny M, Resnick T, Jayakar P. Subdural EEG patterns in children with taylor-type cortical dysplasia: comparison with nondysplastic lesions. J Clin Neurophysiol 2005;22(1):37–42 32. Altman K, Shewmon DA. Local paroxysmal fast activity: significance interictally and in infantile spasms. Epilepsia 1990;31:632 33. Jayakar P, Duchowny M, Alvarez L, Resnick T. Intraictal activation in the neocortex: a marker of the epileptogenic region. Epilepsia 1994;35(3):489–494

133

16

  Extra- and Intraoperative Electrocortical Stimulation Ingrid Tuxhorn

Summary Electrocortical stimulation (ECS) is still the gold standard for defining functional cortex when planning a focal resection to treat medically refractory epilepsy. While it is a valuable tool used at all level IV epilepsy surgery centers, it requires in depth knowledge and experience of specific pediatric issues. In this chapter we will review the current principles, practices, and techniques published from various centers, the underlying electrophysiology, pediatric stimulation paradigms and safety issues. The value of extraoperative vs. intraoperative ECS is discussed. We also review in detail the impact that developmental or acquired pathology may have on the functional anatomy of the developing brain, which is highly relevant for surgical resective planning. Electric stimulation of motor, sensory, language, and more specific cortical areas is reviewed in detail. Lastly, we discuss the value of newer noninvasive techniques as adjunctive methods for ECS that may reduce the risks of invasiveness. Keywords:  cortical, electrocortical stimulation, cortical mapping, extraoperative stimulation, intraoperative simulation

„„ Introduction Pediatric epilepsy surgery is no longer a treatment of last resort and surgical resection of epileptic tissue is an established treatment for children with intractable focal epilepsy.1 Outcome researches from leading centers support early pressurgical evaluation and selection of children with intractable focal epilepsy to optimize seizure control and long-term psychosocial outcome.2 Unique developmental and maturational issues necessitate a specific pediatric approach for referral, diagnosis, and management, which has been outlined in the recent recommendations of the International League Against Epilepsy (ILAE) subcommission on pediatric epilepsy surgery.3 Similarly, the pediatric presurgical evaluation requires specific pediatric epilepsy expertise of experienced and knowledgeable pediatric epilepsy centers.4 This includes accurate description and classification of seizure types and specific epilepsy syndromes and etiologies along the lines of the ILAE classification guidelines.5 In addition, diagnostic assessment of developmental impairments and risks for developmental stagnation and decline as seen in the encephalopathies and other progressive epilepsies is important to optimize timing of surgical intervention.

„„ Electrocortical Stimulation Basics Pediatric Aspects of the Presurgical Evaluation The presurgical evaluation is a multimodal diagnostic approach with the goal of localizing the epileptogenic zone, defining the epileptogenic substrate, and delineating neighboring functional cortex to reduce the risk for neurological deficits with surgical removal.6 Outcome studies suggest that the cortical regions that underlie epileptogenicity must be excised entirely because residual epileptogenic tissue increases the risk for persisting postoperative seizures.7 Medically resistant partial epilepsies in children are more heterogeneous with regard to anatomic localization, types and extent of pathologies, and electroclinical functional characteristics compared to adults. The spectrum of epileptogenic lesions spans the gamut from discrete structural changes as seen in type 1 cortical dysplasias extensive multilobar or hemispheric lesions.4 Temporal lobe epilepsy in children is also more heterogeneous compared to adult temporal lobe epilepsy caused mainly by hippocampal sclerosis or tumors.8 Typically temporal lobe epilepsy in childhood arises from more extensive pathology also involving neocortical regions and extending sometimes beyond the margins of the temporal lobe rather than being confined to the mesial temporolimbic ­structures.9 In addition, the seizure semiology is highly age dependent with prominent motor features (tonic, myoclonic, spasms) that are more typical of extratemporal onset and not suggestive of temporal limbic localization.10 Invasive electroencephalography (EEG) and functional mapping may be indicated to properly tailor a cortical resective procedure after the ictal onset zone and eloquent cortex have been clearly localized with these procedures.11 The high incidence of extratemporal neocortical epilepsy in pediatric patients makes this an important consideration.4 The anatomic landmarks of cortical areas subserving important functions, such as sensation, movement, language, and vision, have been well delineated. However, there is sufficient interindividual variation that may be accentuated with associated developmental or acquired pathological conditions through a process of intrahemispheric or interhemispheric reorganization. Thus, careful and individualized presurgical investigation with functional cortical mapping is essential for each patient on a case-by-case basis. Cortical electrical stimulation remains the gold standard procedure in cortical mapping

16  Extra- and Intraoperative Electrocortical Stimulation although functional MRI, transcranial magnetoencephalography (MEG), transcallosal inhibition are helpful adjunctive techniques with low invasive risk.

A Historical Note After extensive studies in animals, cortical stimulation in humans was performed by Victor Horsely in London, Fedor Krause in Berlin, Harvey Cushing in Boston, and Ottfried ­Foerster in Breslau. The first report of a pediatric patient who is evaluated with cortical stimulation was published in the 1950s by Penfield and Jasper in Montreal who reported a 4-year-old with tuberous sclerosis and epilepsy arising from the central region.12 They performed intraoperative electrocorticography (ECoG) and found a well-localized spike focus in the right central region. With cortical stimulation, they reproduced the left clonic seizures of the patient before resecting that area. They also reported a 16-year-old girl who had active spontaneous spikes over the first temporal gyrus and midtemporal region and reproduced her habitual aura of fear by stimulating the anterior insula close to the junction with the uncus.12,​13 The appearance of “dreamy states” on stimulation of the uncus and other clues from animal studies and ECoG led Penfield and Jasper to believe that the mesial and inferior parts of the temporal lobe were the origin of many epileptic attacks. Resection of these regions subsequently improved the outcome significantly. The pioneering work of Penfield and Jasper also led to the mapping of the anterior and posterior language areas and visual and cortical areas and further refined the cortical map of sensory and motor representation.12,​13,​14 Subsequently, many studies have described the utility of cortical stimulation either used extraoperatively or intraoperatively in adults and children to delineate cortical functional areas in relation to areas of epileptogenesis before surgical removal.14 Cortical stimulation still continues to be the gold standard for cortical functional mapping in the neurosurgical management of epilepsy in adults and childhood.

Physiology of Electrocortical Stimulation The neurophysiological effects of extracellular neuronal stimulation have been studied extensively, and were reviewed by Ranck in 1975.15 The voltage distribution in neural tissue after electrical stimulation depends on the current density (which is a function of stimulus frequency and wave forms) applied, as well as the stimulation-induced membrane polarization. The electrical field within brain tissue produced by stimulation of a subdural electrode (SDE) has a complex three-dimensional shape, and underlying neural processes may be subject to depolarizing or hyperpolarizing events from the stimulation, which may depend on stimulation parameters, the cellular geometry of cortical pyramidal neurons, and the position of the neuron in relation to the stimulating ­electrode.16,​17 The effect of the applied stimulus on local neuronal cells near the stimulating electrode may be considered to result in a direct effect of the applied electrical field on the local cell or the indirect stimulation induced transsynaptic excitation and inhibition, which results from activation of a large number of axon terminals leading to increased synaptic activity on the dendritic tree of the local cells. Depending on the types and numbers of synaptic receptors

activated, the transsynaptic activity induced by electrical ­ stimulation may be excitatory, inhibitory, or a blend of both. Generally, with appropriate stimulus conditions, the maximal current density is achieved beneath the stimulated electrodes so that the stimulus-related responses usually represent cortical function in the crown of the gyrus, whereas the banks of the sulcus are not investigated with this method. Potentially, there may be distant current spread to produce positive responses from remote areas. Cortical excitability in children is known to be different from adults, and electrical stimulation at maximal stimulus intensity, as will be discussed later, may not elicit a positive response after stimulating functional cortex. This may result in yielding a false–negative response and the risk of removing potentially functional cortex.11,​18,​19

„„ Extraoperative Electrocortical Stimulation Extraoperative ECS is usualy achieved with direct e ­lectrical stimulation of the cerebral cortex via subdural or depth ­electrodes.11,​16,​20,​21 This technique has been in use for more than 40 years, and two effects commonly observed have been described in the extensive literature. Cortical stimulation may activate cerebral function producing positive phenomena such as tonic or clonic movements, and special sensations, or cortical stimulation may inhibit function producing negative phenomena such as speech arrest or arrest of motor function. However, in the pediatric population, mapping of cortical function with direct stimulation may be less reliable because of limited patient cooperation and the absence of or inconsistent cortical responses of the immature cortex at lower stimulation thresholds compared with adults.11,​18,​19 Once interictal discharges and sufficient seizures have been recorded from intracranial electrodes, cortical stimulation is performed in a systematic fashion, usually over several days, depending on the number of electrodes, the area that needs to be mapped, and the patient’s degree of cooperation. Because there is a potential to induce seizures, ECoG monitoring is essential to detect afterdischarges that may herald increased epileptogenicity under the stimulated cortex. To reduce the risk of stimulation-induced seizures, anticonvulsants that may have been reduced or stopped to activate seizures for video-EEG recording and analysis should be restarted before initiating the cortical stimulation studies. In addition, temporary benzodiazepine coverage during the procedure may be useful.

General Principles and Techniques A widely used standard stimulation paradigm for extraoperative cortical stimulation via SDE grid uses biphasic rectangular pulses delivered at a rate of 50 per second in trains lasting 3 to 5 seconds.11,​18,​21 The pulse duration is held constant at 0.3 ms, whereas the stimulus intensity is increased in a stepwise fashion to a maximum of 15 mA or to a stimulus intensity that elicits a clinical response or afterdischarge less than the maximal threshold of 15 mA. The following stimulation parameters are used routinely in adult patients at the Cleveland Clinic Epilepsy Center: a stimulus is applied for a duration of 2 to 5 seconds to an active

135

136

IIb  Preoperative Electrophysiological Assessment electrode with a frequency of 50 Hz as a biphasic square wave, a constant current of 300 µs duration with incremental steps of 1 to 2 mA over a range of 1 to 15 mA.16 As the stimulus intensity is gradually increased to 15 mA, either positive responses are elicited or afterdischarges occur. Positive motor responses are elicited at the primary and supplementary motor area; sensory responses are elicited at the primary and secondary sensory area; negative motor responses are elicited at the primary and secondary supplementary negative motor areas; and language dysfunction is elicited over Broca’s, Wernicke’s, and the left basal temporal language area.20,​21,​22,​23,​24 Special symptoms resulting from cortical stimulation of the dominant parietal cortex may include agraphia, acalculia, finger agnosia, and left–right confusion seen in Gerstmann syndrome.25 A variety of auras have been reproduced by cortical stimulation.26 A distant reference electrode over a non-eloquent region of the cortex serves as a nonactive current sink. The active electrode is switched systematically from electrode to electrode across the entire grid allowing the function of the cortical area underlying each electrode to be investigated.16 There is a paucity of studies and reports about stimulation parameters in the pediatric population. However, conventional stimulation paradigms based on fixed pulse duration as described previously for adults rarely elicit responses in infants and young children.11,​18,​19 Various pediatric centers have developed and published stimulation paradigms that rely on increments in both stimulus intensity and pulse duration. These will be discussed below in more detail. Once eloquent cortex has been identified, a resection map based on the interictal and ictal epileptiform activities, which together define the epileptogenic zone and the geography of the surrounding eloquent cortex, is designed to allow maximal resection of the epileptic and associated lesional cortex with sparing of surrounding eloquent regions. In general, resection of primary eloquent cortex, which includes the primary motor, sensory, language, visual, and memory areas, results in neurological deficits. However, some secondary or accessory eloquent regions may be resected without significant permanent neurological deficits. These include the basal temporal language area, negative and second sensorimotor areas, and even the primary motor face area, which is bilaterally innervated.

Safety Issues and Complications in Pediatric Patients The use of subdural or depth electrodes is an invasive procedure that may be complicated by infection, hemorrhage, edema, mass effect, or infarction. The probability of complications associated with intracranial electrodes has been reported by various centers at 2 to 4%. Patients who have undergone prior high-dose brain irradiation may be at particular risk for developing reactive cerebral edema necessitating emergent removal of SDEs. This has only been reported in adult patients.27 In the pediatric age group, the complication rate of SDE implantation is also reportedly low and comparable to the adult experience.11 The energy applied to the surface of the brain via electrical stimulation per se may add an additional risk to damaging the cortex surface. Microscopic studies, however, have not shown gross structural damage, but mild inflammatory responses have been demonstrated in the pathology of the resected tissue.28,​29

The use of MRI-compatible platinum electrodes over stainless steel electrodes allows for coregistration imaging of the electrodes with MRI.30 There is, however, no reported safety data on higher field strength MRI and SDE compatibility, and practitioners need to proceed with caution at this age. There is no evidence that cortical stimulation produces kindling because the afterdischarge thresholds do not progressively decrease with repeated stimulation over a cortical region, although the thresholds may be variable with repeated stimulation.

Special Considerations in Children Electrical cortical stimulation results that are elicited are highly subject to ongoing developmental and maturational processes involving myelination and neuronal connectivity in the pediatric brain across the ages. An efficient and safe paradigm for eliciting responses from immature cortex based on physiological principles of stimulation and neural maturation has been elegantly studied and described by Jayakar et al.9,​18 The response characteristics of neural tissue in relation to stimulus parameters is best described by a strength–duration (SD) curve plotting the current intensity needed to produce a response as a function of pulse duration. The minimum intensity required to elicit a response at a very long pulse duration is termed the r­ heobase, and the pulse duration required to elicit a response using stimuli at twice the intensity of the rheobase is defined as the ­chronaxie. The chronaxie represents the safest point on the SD curve for eliciting a response and is significantly affected by myelination being considerably longer in unmyelinated fibers. The SD curve thus shifts to the left as axons myelinate and the chronaxie shortens. By increasing the stimulus intensity and the pulse width from the usual 0.3 ms used in adults to 1 ms, the longer chronaxies in children can be more effectively stimulated, and positive responses or afterdischarges can be elicited.9,​18 A graphic computation of the energies delivered at all values of current intensity between 1 and 15 mA and pulse duration between 0.3 and 1.0 ms has been published, and the points on the graph corresponding to progressively higher levels of energy sequentially traced.18 A stimulation paradigm following this tracing would have the least energy increment at each step but would be extremely cumbersome for routine clinical use. The authors therefore selected three sequences of alternating increase and decrease of intensity and pulse duration that approximates the outline of the tracing. This dual-increment paradigm starts with an intensity of 1 mA and pulse duration of 0.3 ms, and with each subsequent trial, the intensity and pulse duration are adjusted by 1 to 2 mA or 0.1 to 0.2 ms, respectively, until clinical responses or afterdischarges are obtained. In three patients aged 1, 3, and 4.5 years, positive cortical responses were elicited using the dual-increment paradigm after the standard fixed pulse duration paradigm failed to elicit clinical r­ esponses or afterdischarges. Thus, the technique of ­dual-­increment stimulation rather than the standard fixed-duration paradigm should be used in young children to accurately define c­ ritical cortex in the immature brain, facilitating safe excision of adjacent epileptic tissue. There are few studies defining detailed data of sensorimotor functional maps of the developing brain in children.11,​18,​31 This is partly because of the higher medical risks associated with implantation of SDE, difficulties with patient cooperation, and increased necessity for anesthesia. However, there is good evidence that the

16  Extra- and Intraoperative Electrocortical Stimulation stimulation threshold to activate normal and functionally abnormal cortex is higher in infants and young children. Clinical motor responses are frequently obtained at or above the afterdischarge threshold so that the stimulation paradigm may need to “override” the afterdischarge to obtain eloquent responses. Other authors have reported that they were able to elicit responses in young patients younger than 5 years of age by increasing the stimulus duration after failing to obtain any responses at the maximal fixed duration stimulation.32 Motor responses with tongue movement are difficult to achieve in children younger than the age of 2, and the motor responses from the lower face tend to be bilateral rather than contra- and unilateral when the lower rolandic cortex is electrically activated.19 Individual finger movements are usually first noted after the age of 3 years, and clonic movements appear subsequent to tonic movements in response to electrical stimulation of the central cortical hand area. This reflects maturational processes involving motor neuronal pathways in the cortical areas 4 and 6. Besides the effect of maturation on cortical stimulation results, there is ample evidence for atypical functional networks in children and adults with developmental and early lesions. This will be discussed in more detail in the sections on stimulation results in relation to pathologies. Besides the impact of maturation of cortical electrophysiology on ECS results, the language, motor, and other tasks must be adapted to the patient’s age and neurodevelopmental status to include considerations of the individual child’s ability to cooperate and understand performance limitations caused by attention and comprehension. This is generally more time consuming, and several sessions may be needed to obtain workable and reliable stimulation results.11,​19

In an earlier study, intraoperative brain-mapping techniques were found to be reliable, effective, and safe in children with brain tumors.34 Sensorimotor pathways could be reliably localized with intraoperative methods by these authors in their pediatric population with brain tumors. Language mapping results showed variability and some anatomical unpredictability in peritumoral cortex, aiding with the operative resection of the tumors adjacent to eloquent brain regions.34

Brain Malformations Newer studies with somatosensory-evoked potentials and electrical stimulation of the sensorimotor cortex in pediatric patients with malformations of cortical development suggest that the overlapping of sensory and motor functions across the central sulcus is more complex and extensive.35 This abnormal somatotopic organization in patients with cortical dysplasia supports the concept of abnormal, widespread cerebral organization in the dysgenetic cortex. This may be secondary to mechanisms leading to compensatory reorganization involving as yet unknown processes underlying brain plasticity.35 This type of somatotopic reorganization is demonstrated in the case seen in Fig. 16.1, Fig. 16.2, Fig. 16.3, and Fig. 16.4. A recent case report of an adolescent female further supports the notion of plastic reorganization in the proximity of dysplastic cortex lesions.36 The reported patient had intractable

The Effect of Pathology on Electrocortical Stimulation Mapping Specific pathologies may alter cortical stimulation thresholds, resulting in a reduction of eloquent responses and false– negative results that may put the patient at risk for deficits from surgical removal of potentially eloquent cortex. Several studies report the effect of lesions on cortical stimulation results in the pediatric population.

Tumors Intraoperative electrical stimulation was recently reported in 17 children with tumors of the central region in close relation to the motor pathways.33 Using 0.5-ms pulse width, 5-second stimulus duration, and 50-Hz frequency, these authors successfully identified motor eloquent cortex in 15 patients and in all patients younger than 5 years of age with current densities between 8.5 to 12.5 mA. These authors found intraoperative stimulation effective in mapping eloquent cortex in all patients with preexisting motor deficits, even in patients 5 years of age and younger. However, they reported failure to evoke motor responses in two cases of retrorolandic, low-grade tumors. Anatomical displacement of normally organized cortex, reorganized functional connectivity, and altered threshold responses caused by the developmental tumor lesions need to be considered when interpreting these stimulation results.

Fig. 16.1  Axial flair MRI showing focal cortical dysplasia involving the left central rolandic region.

137

138

IIb  Preoperative Electrophysiological Assessment focal epilepsy caused by a mild type 1 cortical dysplasia involving eloquent hand motor cortex defined by extra- and intraoperative stimulation. The lesion was resected, followed by complete paresis, which recovered substantially after several months, leaving the patient seizure free with minimal hand weakness. Eight pediatric and adult patients with frontal lobe cortical dysplasia involving eloquent cortex were operated on after extraoperative stimulation for medically intractable epilepsy.37 Functional language and motor regions and epileptogenic areas were assessed by extraoperative electrical cortical mapping and ECoG recording and found to colocalize with epileptogenic regions and

Fig. 16.2  Skull radiograph demonstrating placement of a lateral 8 × 8 cm subdural grid.

balloon cell negative dysplastic regions without fluid attenuated inversion recovery signal abnormalities on MRI. In a further study, right-sided language localization was demonstrated with extraoperative subdural stimulation in four of six patients with known bilateral language representation. 38 The etiologies included nondominant, right-sided dysplasias, tumors, and nonlesional MRIs. The language maps were located in the analogous and classic frontal and temporal language regions of the dominant hemisphere. One patient had a silent language map in this report, and another patient had a wide distribution of single language error sites over the right temporal lobe (Fig. 16.1, Fig. 16.2, Fig. 16.3, and Fig. 16.4).38

Fig. 16.3  A three-dimensional MRI reconstruction of the brain with coregistration of the subdural electrode grids demonstrates coverage of the lesion and identifies the overlying electrodes.

Fig. 16.4  This schematized map shows the perilesional epileptogenic zone that was defined by recording interictal and ictal discharges. Electrocortical stimulation demonstrates an abnormal somatopic homunculus with displacement of the motor hand area and redundancy of the sensory cortex of the upper extremity. A lesion-centered resection with close margins to eloquent cortex was performed with good results. (This image is provided courtesy of the Cleveland Clinic Foundation.)

16  Extra- and Intraoperative Electrocortical Stimulation

Acquired Pathology Cortical functional reorganization has also been reported in an 18-year-old patient with chronic epilepsy caused by an acquired perirolandic lesion (cavernous hemiangioma) who demonstrated expansion and redundancy of the perilesional hand and finger sensorimotor regions.39 It has been suggested that ongoing epileptiform activity may suppress normal cortical function overlying a lesion.39 A further case studied by serial cortical stimulation mapping showed that neighboring regions may take over function.40

Electrocortical Stimulation with Depth Electrodes in Children Stereotactically placed intracerebral electrodes are used to define a stable, unique epileptogenic zone by several centers in France and Italy. After recording of spontaneous seizures, all patients undergo intracerebral electrical stimulation, with the goal of better defining the epileptogenic zone and providing functional mapping of eloquent brain areas.41,​42 Stimulations are usually applied between contiguous contacts of electrodes at low frequency (1 Hz, single stimulus duration 2–3 ms, current intensity 0.4–3.0 mA) or high frequency (50Hz, single stimulus duration 1 ms, current intensity 1–3 mA) depending on the presumed excitability of a given structure and on the type of clinical signs that are to be elicited. High-frequency stimulation is preferred as reported by these authors to assess the organization of the epileptogenic network by inducing the patient’s electroclinical seizure and analyzing different ictal phenomena to gain anatomical functional correlates. Intracerebral electrical stimulation has been used to map cortical eloquent areas as well as subcortical critical bundles such as descending fibers of the corticospinal tracts. The French school has reported on the utility of mapping the somatosensory, motor, visual, and language areas in children and has found it valuable in cases of cortical dysplasia involving the central region where the normal anatomy of the gyri is disrupted and the pathology may be embedded within eloquent cortex and bundles. The authors report that the information provided by intracranial functional mapping strongly contributes to the absence of postoperative motor and linguistic deficits in their series of children undergoing resections after stereoelectroencephalography (SEEG) evaluation. There are no comparative studies looking at the sensitivity and specificity of ECS with SDE and SEEG.

Electrocortical Stimulation-Induced Responses Several robust responses have been described that define eloquent cortex with ECS. To infer that eloquent cortex has been defined, the stimulation needs to result in a reproducibly demonstrable change in neurological function that may either be a positive or negative phenomenon. Although the positive phenomena are easily observable or can be reported by older pediatric patients, the negative phenomena may go unnoticed and need to be specifically tested for, which may not be possible in the younger patient.

Primary Motor Area A primary motor area that has been termed area 4 according to the Brodmann’s cytoarchitectural map encompasses the

Video 16.1 Motor mapping: primary motor hand area. (This video is provided courtesy of Ingrid Tuxhorn.) ht t ps://www.thieme.de/de/q.ht m?p=opn/ tp/255910102/9781626238176_c016_v001&t=video

Video 16.2 Speech mapping. (This video is provided courtesy of Ingrid Tuxhorn.) h t t p s : / / w w w. t h i e m e . d e / d e / q . h t m ? p = o pn/ t p/255910102/9781626238176_c016_ v002&t=video ­ nterior bank of the precentral sulcus; a second premotor area, a termed area 6 encompasses the more anterior precentral gyrus and posterior portion of the superior frontal gyrus. The work of Penfield and colleagues has expanded our understanding of the somatotopic map of the body as depicted by Penfield’s figurine homunculus.12,​13 The larger extension of motor responses to areas as much as 4 cm anterior and 2 cm posterior to the central sulcus and not limited to the precentral gyrus led to the appreciation of a more extensive motor representation in the central region. In addition, it has become clear that the cytoarchitectural areas 4 and 6 cannot be delineated with ECS. Stimulation parameters, including frequency, duration, and intensity, may affect the type of motor response elicited that usually first involves clonic movements of distal muscle groups (Video 16.1). Maximizing the resections without incurring motor deficits in lesional and nonlesional frontal lobe epilepsy that encroach on or distort the precentral area may be best achieved using detailed motor mapping (Video 16.2). Localized reorganization of function may modify the traditional homuncular ­ representation and put patients at increased risk for ­postoperative deficits on the one hand or, on the other hand, dislocation of function may allow safe and more generous ­ resective procedures, as described in detail elsewhere in this chapter and illustrated in Fig. 16.1, Fig. 16.2, Fig. 16.3, Fig. 16.4.

Supplementary Sensorimotor Area Animal studies performed more than a century ago demonstrated that motor responses could be elicited by stimulation of the mesial aspect of the superior frontal gyrus just anterior to the primary motor leg area. Once implantable subdural grid electrodes and depth electrodes were developed, systematic studies of this interhemispheric mesial cortical region became possible in the 20th century.43,​44 This region has been named the supplementary sensorimotor area (SSMA) because both sensory and motor functions are represented. The motor pattern of responses is quite distinctive from that elicited by stimulating the primary motor area. The SSMA type motor responses are characterized by predominantly tonic responses of proximal muscle groups (which are frequently bilateral), asymmetric movements of the lower and upper extremities, and head and eye deviations and vocalizations. A somatotopic organization has also been defined that is anterior–posterior in orientation with the head and eye region lying anterior and the leg region posterior. Although resection

139

140

IIb  Preoperative Electrophysiological Assessment of the SSMA is generally safe, temporary contralateral weakness, difficulty in the initiation of movements, and mutism with intact comprehension has been reported. In addition, the proximity of the caudal end of the SSMA to primary motor and sensory control of the leg need to be appreciated. Resection of eloquent cortex in this region therefore needs a careful risk–benefit analysis and informed consent standards involving patient guidance with medical ethics experts.

Sensory Areas Stimulation studies have defined three distinct areas from which somatosensory sensation can be elicited: the primary sensory cortex (SI) in the postcentral gyrus, the secondary somatosensory cortex (S2) in the frontal and parietal operculum, and the supplementary sensorimotor cortical area in the mesial surface of the frontal and parietal cortex (S3).12,​13 Each region has a unique somatotopic organization: SI, located in the postcentral parietal region and consisting of Brodmann areas 3a, 3b, 2, and 1 has a clear somatotopy mirroring that of the motor strip. S2 is located on the superior bank of the sylvian fissure in proximity to the planum infraparietale of the frontal operculum and sensory responses to stimulation characteristically are somatotopically “discontiguous”—affecting the opposite whole body but also the ipsilateral side, especially simultaneous upper and lower limb involvement.23 The S3 responses are frequently admixed with tonic motor features and may localize to the body bilaterally, ipsilaterally, and contralaterally to the brain side that is stimulated. The semiology of somatosensory auras may also give a good indication of which sensory area is the symptomatogenic zone.45 Resections in the S1 region will lead to permanent changes in contralateral sensory perception with deficits affecting primarily position sense and fine touch, but vibration and pain sensation are usually not affected by S1 resections.

Language Mapping Lesional studies in patients first defined the anterior (Broca’s) and posterior (Wernicke’s) language areas in the left inferior frontal lobe and first temporal convolution, respectively. These areas were first mapped intraoperatively by Penfield and Roberts who induced speech arrest, alexia, agraphia, anomia, paraphasia, and occasional positive grunting noises with electrical interference.12,​13,​46 A superior language area was also defined by them located anterior to the rolandic motor foot area in the mesial frontal lobe.46 A third speech area has been defined in the basal temporal region of the dominant temporal lobe by Luders and Lesser et al in 1986.22 Reading aloud is a reliable screening task for mapping language area, and arrest of speech is the typical feature to look for (Video 16.3). Importantly, negative or positive motor responses and diminished responsiveness as a cause of speech arrest need to be ruled out by checking tongue movements. If slowing of speech occurs, then additional testing, including naming of objects, auditory word repetition, reading comprehension, and spontaneous speech, may be warranted. This will require good cooperation and an adequate developmental level in children to obtain reliable results. ECS usually produces interruption of verbal fluency at Broca’s area and evokes comprehension

Video 16.3 Motor mapping. (This video is provided courtesy of Ingrid Tuxhorn.) ht t ps://www.thieme.de/de/q.ht m?p=opn/ tp/255910102/9781626238176_c016_v003 &t=video ­ eficits when stimulating Wernicke’s area, but there may be a d significant overlap of symptoms. A robust negative motor response is seen when performing ECS in the inferior frontal gyrus just anterior to the primary facial motor representation, and this has been termed the primary negative motor area.12,​13,​47 Further studies have shown a more extensive distribution of negative motor areas of the upper extremities, extending over the lateral premotor cortex.47 Selective removal of the primary negative motor area is possible without producing persisting speech and language or motor deficits.

Studying More Specific Brain Anatomy with Stimulation Angular Gyrus The feasibility of intraoperative stimulation of the angular gyrus (AG) was recently studied and reported in five adult patients with circumscribed lesions (all had primary or metastatic tumors) in this region.48 Based on human and animal studies, it is known that the AG is a higher-order supraregional center that is integrated in a neuronal network mediating movement with complex projections to the pulvinar of the thalamus and ipsi- and contralateral cortical association areas in the prefrontal, temporal, and occipital lobes. Damage to the dominant AG may result in agraphia and alexia. A previous article reported a functional Gerstmann syndrome during electrical stimulation of the dominant perisylvian cortex.25 In this newer study,48 bipolar and monopolar cortical stimulation techniques were applied to the AG cortex and compound muscle action potentials were recorded in the contralateral arm. The study shows that selective electrical stimulation of the AG elicits a motor response in the ­contralateral upper extremity. The data reported show that the technique is feasible in the intraoperative setting and that the AG cortex plays a role in bimanual motor function, which deserves further study with this technique. This technique may be of value in studying patients with Landau–Kleffner syndrome undergoing surgical epilepsy treatment.

Interhemispheric Connections of Motor Areas Functional connectivity of the brain via various white matter tracts, which consist of the commissural fibers to the contralateral cortex, projection fibers to subcortical nuclei, and arcuate fibers to the ipsilateral cortical structures, have been studied in vitro and recently in vivo with newer MRI, especially with diffusion tensor imaging techniques as well as transcranial magnetic stimulation. In addition, newer neurophysiological techniques have been developed for the in vivo evaluation of neural fibers and their projections. A recent study to clarify

16  Extra- and Intraoperative Electrocortical Stimulation interhemispheric connections of motor cortex investigated corticocortical evoked potentials in vivo using subdural grid stimulation in patients undergoing presurgical evaluation for epilepsy treatment by delivering a bipolar pulsed stimulus to two electrodes overlying one motor cortex and recording the evoked potential contralaterally from the averaged ECoG.49 Contralateral evoked responses with stimulation of the motor cortex only (no responses were elicited form other stimulation sites) were recorded with interhemispheric latencies ranging from approximately 9 to 24 ms for the initial positive peak and 25 to 39 ms for the second negative peak. These results were felt to suggest that bilateral motor coordination is at least partially controlled at the level of the motor cortex. In pediatric cases, this technique may shed light on the neural mechanism of associated and mirror movements seen in children with epilepsy.

Stimulation-Induced Alien Limb Phenomenon Alien limb phenomenon was reported in a 14-year-old patient who had parietal lobe epilepsy caused by a malformation of cortical development in the left rolandic cortex.50 Stimulation of the central cortex in proximity to the frontal operculum induced involuntary grabbing right hand movements accompanied by the perception of alienness; the patient reported as if the arm belonged to someone else. The authors speculate that stimulation may have induced a functional disconnection or inhibition of primary sensory areas and activation of motor areas of the hand resulting in involuntary movements that were experienced as alien.50

Insular Cortex The insular cortex is not easily accessible for ECS because it is covered by the frontal, temporal, and parietal opercular cortex. Recent studies with intracerebral depth electrodes implanted transopercularly into the insular cortex have yielded more consistent responses compared with previous studies and include somatosensory responses, including painful sensations of the contralateral face, neck, hand, and upper limb (posterior insula); viscero sensitive responses (anterior insula) of the abdomen and thorax typically seen as the initial symptom in mesial temporal lobe epilepsy and a sense of pharyngeal constriction; and less frequently, simple acoustic hallucinations, experiential phenomena, olfactory, gustatory, or vegetative responses have been evoked.51

Laughter Nonictal laughter has been elicited by stimulation of the mesial frontal cortex. The laughter has been reported to be involuntary and not associated with mirth or emotion. Similarly, stimulation of the cingulate cortex, orbitofrontal cortex, and mesial, basal temporal structures has been reported to produce nonepileptic laughter.52

Visual Cortex ECS of Brodmann’s areas 17 (primary visual cortex; also defined as the striate cortex with the lines of Gennari), 18, and 19 (visual association cortical areas) may produce either well-defined visual symptoms correlating well with epileptic visual auras, simple visual hallucinations, or visual illusions

localized in the upper or lower (delineated by the calcarine fissure) contralateral quadrant.53 Most patients with occipital lobe epilepsy will have a visual field deficits but resection of occipital cortex in the face of normal visual fields will result in a new deficit that is usually somewhat compensated in the younger pediatric age group.

Auditory Cortex The primary auditory cortex, Brodmann’s area 41, is located in the posterior medial aspect of the gyrus of Heschl, whereas secondary auditory areas have been demonstrated in contiguous areas extending into the planum temporale and superior temporal sulcus (areas 42, 52, and 22). Because the stimulation response is subjective, and patient cooperation is essential. Elementary crude auditory sensations, hallucinations, and illusions have been described. Unilateral lesions in this region do not appear to lead to auditory deficits.

Negative Functional Effects Several negative responses that are stable and reproducible have been reported over the language areas (anterior, posterior, and basal temporal), primary and secondary negative motor areas, and other negative responses with stimulation of heteromodal associative cortex (supramarginal gyrus area 40, area 7, area 39), producing deficits of higher cortical functions, including a combination of deficits including alexia, anomia, apraxia, and Gerstmann syndrome).24,​25

Extraoperative Electrocortical Stimulation Compared with Other Noninvasive Functional Mapping Techniques In the last two decades, the Wada test, neuropsychological evaluation, and 2-deoxy-2[18F] fluoro-d-glucose positron emission tomography, which allows determination of preoperative language lateralization and assessment of memory adequacy, have been supplemented with additional ­ noninvasive mapping methods that offer a high spatial and temporal accuracy to localize sensory, motor, and language function. Recent noninvasive techniques such as functional MRI and MEG can aid in mapping cortical function as an adjunctive method when planning invasive mapping with chronically implanted SDEs. There is relatively good correlation between intraoperative ECoG and MEG, although direct measures of differences are influenced by the MEG source mapping of sulcal generators versus gyral surface maps with intracranial electrodes. The temporal course of neuronal language processing can be imaged noninvasively with millisecond resolution using MEG; however, at this stage, there is a paucity of pediatric data.54 Numerous studies document that functional MRI is a reliable technique for lateralizing hemispheric language dominance. A recent study compared the results of functional MRI language mapping with intraoperative ECS in patients with temporal lobe epilepsy.55 The sensitivity was 100%, and there was a high spatial accuracy with functional MRI, indicating that areas not activated could be safely resected.55 The authors emphasize that a combination of three language tasks including verb generation, picture naming, and sentence processing was needed to ensure the high sensitivity as no single task was

141

142

IIb  Preoperative Electrophysiological Assessment sufficient for this purpose. However, the specificity of functional MRI was low, and only 51% of functional MRI activations were confirmed on ECS.55 Motor cortex localization with functional MRI is generally highly concordant with intraoperative ECS mapping.56 Several additional electrophysiological techniques besides high frequency ECS, which we described in detail previously, have been studied to map motor subareas.57 These include slow cortical potentials termed Bereitschaftspotential that arise from the motor cortices and occur –1.5 seconds before the onset of selfpaced voluntary movements and are recorded with a long time constant amplifier. The Bereitschaftspotential reflect excitatory postsynaptic potentials in the superficial layer of the motor cortex, occurring in the apical dendrites of the pyramidal neurons. This technique may differentiate the M1, SSMA, and aid in the functional mapping of nonprimary and association cortices. No correlative studies exist with noninvasive techniques, and there is, to date, unfortunately no body of pediatric data. Noninvasive techniques for mapping brain function as described previously may obviate the need for invasive mapping in some cases of well-defined, single epileptogenic lesions and assist in the decision making to pursue invasive studies and potential surgery in complex cases of malformations in or near eloquent cortex.

„„ Intraoperative Electrocortical Stimulation Intraoperative mapping of cortical function by electrical stimulation has been used extensively in neurosurgery since it was first introduced in 1874 by Bartholow. Subsequently, Sir Victor Horsely and in 1909 Harvey Cushing used this technique to define the sensorimotor cortex surrounding a tumor. Ideally, the patient should be awake and responsive but comfortable to perform intraoperative mapping of language cortex, which is technically challenging in pediatric patients. Anesthesia techniques are therefore of prime importance to optimize the success of intraoperative stimulation—premedication with barbiturates, benzodiazepines, and antihistamines should be avoided if ECoG is planned because these agents significantly affect the EEG and may affect the seizure threshold. During the procedure, shortacting anesthetics such as propofol in combination with fentanyl are preferred because they permit rapid induction for the craniotomy procedure but an alert patient for subsequent functional mapping. Local analgesia with lidocaine may be used as a local field block for the scalp incision and dural incisions. The stimulation is performed with a hand-held stimulator using either uni- or bipolar stimuli with parameters quite similar to those used with chronic extraoperative stimulation.58 Closely spaced 5 mm and individualized stimulation points can be selected on the c­ ortex and then tagged to generate a stimulation map and verify the reproducibility of the responses by repeating the stimulation procedure.12,​13,​14 The ECoG will detect afterdischarges or seizures and permit stimulation within appropriate safety limits. The application of intraoperative monitoring is quite limited in the pediatric age group because of issues of feasibility in the awake craniotomy setting

and, more importantly, the s­ pecific challenges relating to higher stimulation thresholds of the immature cortex.58 However, the value of intraoperative ECoG in children with intractable neocortical epilepsy has been recently studied and reported.59

„„ Conclusion Cortical stimulation is a well-established method for defining cortical functional areas subserving language and sensorimotor function in children. In children, specific features need to be considered when performing stimulation and in the interpretation of the functional maps obtained: 1. There is a higher cortical threshold, which is age dependent and reduces inversely to age. Practically, this means that higher current densities (milliamperes) are needed to elicit responses. In addition, there is a greater variability of the stimulation threshold. Stimulation currents, therefore, need to be maximal at each site and afterdischarges may have to be “overridden” to obtain this. The dual-­stimulation paradigm varying the stimulus duration and intensity is valuable in defining critical cortex in young children. 2. Certain pathologies may further raise the stimulation threshold of tumors so that it may be difficult to obtain evoked motor responses. After lesion removal, it has been noted that the stimulation threshold may be lowered. 3. Pathology underlying the epileptic zone as an anatomical substrate is frequently developmental and may result in altered functional plasticity. This may lead to intra- and interhemispheric reorganization with atypical regions of functional mapping (e.g., bilateral language, displaced, or extended functional regions). 4. Because the immature cerebral cortex is relatively refractory to cortical stimulation with standard adult parameters, widened pulse widths (0.14–200 ms), higher frequency ranges (20–50 Hz), increased current densities (0.5–20 mA), and wider train duration ranges (3–25 seconds) may need to be used. 5. The invasive nature of stimulating the cortex directly should be balanced against the accuracy of mapping obtained in each patient as noninvasive imaging may be preferable in some cases. However, it still is the gold standard method for mapping eloquent cortex in proximity to the epileptogenic zone that needs to be resected to treat refractory focal epilepsy. 6. Intraoperative stimulation has limited application in the pediatric age group because of technical and feasibility issues relating to the awake craniotomy setting.

„„ Acknowledgment I thank Tim O’Connor, head EEG technologist, Cleveland Clinic, for his assistance with providing the case illustrations. I thank Dr. Asim Shahid, associate professor, Case Western Reserve ­University, Rainbow Babies and Children’s Hospital—Pediatric Neurologist and Epileptologist, for his assistance with selecting the illustrative video materials.

16  Extra- and Intraoperative Electrocortical Stimulation

References 1. National Institutes of Health Consensus Conference. Surgery for epilepsy. JAMA 1990;264(6):729–733 2. Aicardi J. Pediatric epilepsy surgery: how the view has changed. In: Tuxhorn I, Holthausen H, Boenigk H, eds. Pediatric Epilepsy Syndromes and Their Surgical Treatment. London: John Libby; 1997:3–7

24. Lüders HO, Dinner DS, Morris HH, Wyllie E, Comair YG. Cortical electrical stimulation in humans. The negative motor areas. Adv Neurol 1995;67:115–129 25. Morris HH, Lüders H, Lesser RP, Dinner DS, Hahn J. Transient neuropsychological abnormalities (including Gerstmann’s syndrome) during cortical stimulation. Neurology 1984;34(7):877–883

3. Cross JH, Jayakar P, Nordli D, et al; International League against Epilepsy, Subcommission for Paediatric Epilepsy Surgery. Commissions of Neurosurgery and Paediatrics. Proposed criteria for referral and evaluation of children for epilepsy surgery: recommendations of the subcommission for pediatric epilepsy surgery. Epilepsia 2006;47(6):952–959

26. Schulz R, Lüders HO, Tuxhorn I, et al. Localization of epileptic auras induced on stimulation by subdural electrodes. Epilepsia 1997;38(12):1321–1329

4. Cross JH. Epilepsy surgery in childhood. Epilepsia 2002;43 (Suppl 3):65–70

28. Gordon B, Lesser RP, Rance NE, et al. Parameters for direct cortical electrical stimulation in the human: histopathologic confirmation. Electroencephalogr Clin Neurophysiol 1990;75(5):371–377

5. Tuxhorn I, Kotagal P. Classification. Semin Neurol 2008;28(3): 277–288

27. Jobst BC, Williamson PD, Coughlin CT, Thadani VM, Roberts DW. An unusual complication of intracranial electrodes. Epilepsia 2000;41(7):898–902

6. Lueders H. Textbook of Epilepsy Surgery. London: Informa Ltd.; 2008

29. Wyler AR, Walker G, Somes G. The morbidity of long-term seizure monitoring using subdural strip electrodes. J Neurosurg 1991;74(5):734–737

7. Engel J, Van Ness PC, Rasmussen TB, Ojeman LM. Outcome with respect to epileptic seizures. In: Engel J, ed. Surgical Treatment of the Epilepsies. 2nd ed. New York, NY: Raven Press; 2003:609–621

30. Hamer HM, Morris HH, Mascha EJ, et al. Complications of invasive video-EEG monitoring with subdural grid electrodes. Neurology 2002;58(1):97–103

8. Babb TL, Brown WJ. Pathological findings in epilepsy. In: Epel J Jr, ed. Surgical Treatment of Epilepsies. New York, NY: Raven Press; 1987: 511–540

31. Chitoku S, Otsubo H, Harada Y, et al. Extraoperative cortical stimulation of motor function in children. Pediatr Neurol 2001;24(5):344–350

9. Jayakar P, Duchowny M, Resnick TJ. Subdural monitoring in the evaluation of children for epilepsy surgery. J Child Neurol 1994;9(Suppl 2):61–66

32. Schuele S, McIntyre C, Lueders H. General principles of cortical mapping by electrical stimulation. In: Luders H, ed. Textbook of Epilepsy Surgery. London: Informa Ltd.; 2008:963–977

10. Fogarasi A, Tuxhorn I, Janszky J, et al. Age-dependent seizure semiology in temporal lobe epilepsy. Epilepsia 2007;48(9):1697–1702

33. Signorelli F, Guyotat J, Mottolese C, Schneider F, D’Acunzi G, Isnard J. Intraoperative electrical stimulation mapping as an aid for surgery of intracranial lesions involving motor areas in children. Childs Nerv Syst 2004;20(6):420–426

11. Wyllie E, Lüders H, Morris HH III, et al. Subdural electrodes in the evaluation for epilepsy surgery in children and adults. Neuropediatrics 1988;19(2):80–86 12. Penfield W, Jasper H. Epilepsy and the Functional Anatomy of the Human Brain. Boston, MA: Little Brown; 1954 13. Penfield W, Rasmussen T. The Cerebral Cortex of Man. A Clinical Study of Localization of Function. New York, NY: Macmillian; 1957 14. Gallentine WB, Mikati MA. Intraoperative electrocorticography and cortical stimulation in children. J Clin Neurophysiol 2009;26(2):95–108 15. Ranck JB Jr. Which elements are excited in electrical stimulation of mammalian central nervous system: a review. Brain Res 1975;98(3):417–440 16. Nair DR, Burgess R, McIntyre CC, Lüders H. Chronic subdural electrodes in the management of epilepsy. Clin Neurophysiol 2008;119(1):11–28 17. Manola L, Roelofsen BH, Holsheimer J, Marani E, Geelen J. Modelling motor cortex stimulation for chronic pain control: electrical potential field, activating functions and responses of simple nerve fibre models. Med Biol Eng Comput 2005;43(3):335–343 18. Jayakar P, Alvarez LA, Duchowny MS, Resnick TJ. A safe and effective paradigm to functionally map the cortex in childhood. J Clin Neurophysiol 1992;9(2):288–293 19. Lachhwani D, Dinner D. Cortical Stimulation in the Definition of Eloquent Areas. Amsterdam: Elsevier; 2004 20. Luders H. Symptomatic Areas and Electrical Cortical Stimulation. New York, NY: Churchill Livingstone; 2000 21. Lesser RP, Lüders H, Klem G, et al. Extraoperative cortical functional localization in patients with epilepsy. J Clin Neurophysiol 1987;4(1):27–53 22. Lüders H, Lesser RP, Hahn J, et al. Basal temporal language area demonstrated by electrical stimulation. Neurology 1986;36(4):505–510 23. Lüders H, Lesser RP, Dinner DS, Hahn JF, Salanga V, Morris HH. The second sensory area in humans: evoked potential and electrical stimulation studies. Ann Neurol 1985;17(2):177–184

34. Berger MS, Kincaid J, Ojemann GA, Lettich E. Brain mapping techniques to maximize resection, safety, and seizure control in children with brain tumors. Neurosurgery 1989;25(5):786–792 35. Akai T, Otsubo H, Pang EW, et al. Complex central cortex in pediatric patients with malformations of cortical development. J Child Neurol 2002;17(5):347–352 36. Chamoun RB, Mikati MA, Comair YG. Functional recovery following resection of an epileptogenic focus in the motor hand area. Epilepsy Behav 2007;11(3):384–388 37. Marusic P, Najm IM, Ying Z, et al. Focal cortical dysplasias in eloquent cortex: functional characteristics and correlation with MRI and histopathologic changes. Epilepsia 2002;43(1):27–32 38. Jabbour RA, Hempel A, Gates JR, Zhang W, Risse GL. Right hemisphere language mapping in patients with bilateral language. Epilepsy Behav 2005;6(4):587–592 39. Kirsch HE, Sepkuty JP, Crone NE. Multimodal functional mapping of sensorimotor cortex prior to resection of an epileptogenic perirolandic lesion. Epilepsy Behav 2004;5(3): 407–410 40. Lado FA, Legatt AD, LaSala PA, Shinnar S. Alteration of the cortical motor map in a patient with intractable focal seizures. J Neurol Neurosurg Psychiatry 2002;72(6):812–815 41. Cossu M, Cardinale F, Colombo N, et al. Stereoelectroencephalography in the presurgical evaluation of children with drug-­ resistant focal epilepsy. J Neurosurg 2005;103(4, Suppl):333–343 42. Cossu M, Cardinale F, Castana L, Nobili L, Sartori I, Lo Russo G. Stereo-EEG in children. Childs Nerv Syst 2006; 22(8):766–778 43. Lim SH, Dinner DS, Pillay PK, et al. Functional anatomy of the human supplementary sensorimotor area: results of extraoperative electrical stimulation. Electroencephalogr Clin Neurophysiol 1994;91(3):179–193 44. Fried I, Katz A, McCarthy G, et al. Functional organization of human supplementary motor cortex studied by electrical stimulation. J Neurosci 1991;11(11):3656–3666

143

144

IIb  Preoperative Electrophysiological Assessment 45. Tuxhorn IE. Somatosensory auras in focal epilepsy: a clinical, video EEG and MRI study. Seizure 2005;14(4):262–268 46. Penfield W, Roberts L. Speech and Brain Mechanisms. Princeton, NJ: Princeton University Press; 1959 47. Lüders HO, Lesser RP, Dinner DS, et al. A negative motor response elicited by electrical stimulation of the human frontal cortex. Adv Neurol 1992;57:149–157 48. Kombos T, Picht T, Suess O. Electrical excitability of the angular gyrus. J Clin Neurophysiol 2008;25(6):340–345 49. Terada K, Usui N, Umeoka S, et al. Interhemispheric connection of motor areas in humans. J Clin Neurophysiol 2008;25(6):351–356 50. Boesebeck F, Ebner A. Paroxysmal alien limb phenomena due to epileptic seizures and electrical cortical stimulation. Neurology 2004;63(9):1725–1727 51. Isnard J, Mauguière F. [The insula in partial epilepsy] Rev Neurol (Paris) 2005;161(1):17–26 52. Hoppe M. Cortical mapping by electrical stimulation: other eloquent areas. In: Luders H, ed. Textbook of Epilepsy Surgery. London: Informa Ltd.; 2008 53. Murphey DK, Maunsell JH, Beauchamp MS, Yoshor D. Perceiving electrical stimulation of identified human visual areas. Proc Natl Acad Sci USA 2009;106(13):5389–5393

54. Roberts TP, Zusman E, McDermott M, Barbaro N, Rowley HA. Correlation of functional magnetic source imaging with intraoperative cortical stimulation in neurosurgical patients. J Image Guid Surg 1995;1(6):339–347 55. Rutten GJ, Ramsey NF, van Rijen PC, Noordmans HJ, van Veelen CW. Development of a functional magnetic resonance imaging protocol for intraoperative localization of critical temporoparietal language areas. Ann Neurol 2002;51(3):350–360 56. Chapman PH, Buchbinder BR, Cosgrove GR, Jiang HJ. Functional magnetic resonance imaging for cortical mapping in pediatric neurosurgery. Pediatr Neurosurg 1995;23(3):122–126 57. Ikeda A, Miyamoto S, Shibasaki H. Cortical motor mapping in epilepsy patients: information from subdural electrodes in presurgical evaluation. Epilepsia 2002;43(Suppl 9):56–60 58. Çataltepe O, Comair Y. Intrasurgical cortical stimulation. In: Luders H, Noachtar S, eds. Epileptic Seizures: Pathophysiology and Clinical Semiology. New York, NY: Churchill Livingstone; 2000:172–176 59. Asano E, Benedek K, Shah A, et al. Is intraoperative electrocorticography reliable in children with intractable neocortical epilepsy? Epilepsia 2004;45(9):1091–1099

17

  Stereoelectroencephalography Ika Noviawaty and Patrick Chauvel

Summary Stereoelectroencephalography (SEEG) provides critical information to determine the epileptogenic zone and feasibility of surgical resection using multiple intracerebral depth electrodes in a group of patients with nonconcordant noninvasive diagnostic data. SEEG provides direct access to three-dimensional recording of cortical and subcortical brain regions by stereotactic implantation of multiple and multi-contact intracerebral electrodes. The goal of SEEG exploration is to delineate epileptogenic zone based on anatomo-electro-clinical analysis to tailor resection in achieving seizure freedom outcome. The implantation strategy is carefully defined, based on precise hypotheses generated by the presurgical evaluation result about the most probable cerebral regions involved in seizure origin and propagation. Some of the advantages of utilizing SEEG exploration are in its ability to give access to deep cortical structures and its capability in mapping three-­ dimensional aspect of the epileptogenic network with relatively lowcomplication rate. Keywords:  stereoelectroencephalograpy, anatomo-­ electroclinical analysis, spatiotemporal, intracerebral e ­ lectrodes

„„ Introduction Epilepsy is a disorder of the brain characterized by an enduring predisposition to generate epileptic seizures and by the neurobiological, cognitive, psychological, and social consequences of this condition.1,​2 Recurrent epileptic seizures are transient occurrence of signs and/or symptoms due to abnormal, excessive or synchronous neuronal activity in the brain.3 Although medical management with antiepileptic drugs is the first line of treatment in epilepsy patients, approximately one-third of them fail to respond to medical management. Surgical management of epilepsy is a well-established treatment option for this patient group. These patients undergo extensive presurgical assessment to determine if they are good surgical candidates and what would be the most appropriate surgical approach. In a group of patients with ­nonconcordant noninvasive diagnostic data, monitoring with ­intracranial electrodes provides critical information to determine the epileptogenic zone and feasibility of surgical resection.4 SEEG provides this information using intracerebral electrodes for diagnostic purposes.

„„ Historical Evolution of Stereoelectroencephalography Epilepsy surgery in France started in the beginning of the 1950s, which was also a time of rapid advances in diagnostic electroencephalography (EEG) techniques and its application in epilepsy surgery cases. At that time, Henri Hécaen, a neurologist/neuropsychologist, who visited the ­ Montreal Neurological Institute in Canada when Wilder Penfield was fully active, encouraged neurosurgeons G. Mazars and J. Guillaume to establish an epilepsy surgery program at Ste Anne Hospital in Paris. They initially performed surgeries assisted by electrocorticography. A few years later, Jean Talairach, another neurosurgeon who worked in a distinct neurosurgical department at Ste Anne Hospital, started to collaborate with Jean Bancaud, a clinical neurophysiologist, who had previously worked at La Salpêtrière Hospital in P ­ aris. Bancaud was familiar with cortical mapping and clinical semiology-based epilepsy localization technique developed by Penfield, but he realized that the surface electrophysiological assessment methods were not sufficient to localize seizure onset on a rational basis. Talairach was well versed with stereotaxy and functional anatomy and published the first stereotactic atlas of deep brain nuclei in 1958.5 The ­collaboration between Talairach and Bancaud pioneered the stereotactic implantation of depth electrodes for epilepsy. The first stereotactic surgery operating room was opened in Ste Anne in 1959 and Talairach and Bancaud introduced the term “stereoelectroencephalography” or “SEEG” in 1962. Although the first chronic depth electrode cerebral recording was performed by Bickford and Kairns at Oxford in 1944, the modern approach to stereotactically placed depth electrode recording for direct exploration of the brain was pioneered by Talairach and Bancaud.6,​7,​8,​9 The Talairach’s stereotaxic system based on anterior–posterior commissure line and the superimposition of angiography and encephalographies r­ evolutionized the field of multimodal image-guided epilepsy surgery. SEEG leads to three-dimensional localization of the lesional, irritative, and epileptogenic zones as the basic constituents of an epileptogenic lesion. The method also allowed invasive EEG and individualized tailored resective surgery to be done at a ­different time. SEEG was born and developed in Europe, but over the last decade, many centers in the North America also started to adopt the technique. This is clearly in relation with an increased recognition of the complexity of temporal epilepsies which extend beyond the MRI visible hippocampal sclerosis, the diversity of

146

IIb  Preoperative Electrophysiological Assessment the frontal and parietal epilepsies, and the number of MRI negative epilepsy patients whom can benefit from epilepsy surgery.

„„ Basic Principles of Stereoelectroencephalography SEEG provides direct access to three-dimensional recording of cortical and subcortical brain regions by stereotactic implantation of multiple and multi-contact intracerebral electrodes. The implantation strategy is carefully defined, based on precise hypotheses about the most probable cerebral regions involved in seizure origin and propagation. These hypotheses are formed based on the detailed a ­ nalysis of the noninvasive evaluation results. Then spatiotemporal seizure dynamics in correlation with the emergence of seizure semiology (anatomo-­ electroclinical correlations) is essential to plan an optimal ­surgical strategy. The aim of optimal epilepsy surgery strategy is to achieve seizure freedom while avoiding undesirable ­neurological postoperative deficit. The basic principles of SEEG, which Bancaud and Talairach developed and defined in 1965, have been validated over half a century.10,​11,​12,​13 The rapid development of neuroimaging during this time had also contributed to the evolution of SEEG that demonstrated epilepsy as a dynamic disease through its unique three-dimensional recording access. SEEG had provided evidence that epileptic seizures are not always originated in one restricted “focus.” Instead, seizures are resulted from abnormal synchronous changes in cortical networks (of variable sizes) disrupting their normal functions and producing epileptic manifestations with propagation.13,​14,​15,​16

Anatomo-Electro-Clinical Correlations and Epileptogenic Network Concepts Clinical semiology is the manifestation of epilepsy and is what patients experience from their disease. Historically, direct observation of the seizures and postmortem brain examination were the only two available methods used to advance the comprehension of epileptic diseases in the 19th century. Decades later, the development of EEG contributed to the further advancement of neurological observation by providing an objective marker of neuronal dysfunction. Emergence of semiology in correlation with EEG modification further refined the hypothesis forming process of localization of seizure onset and spread. Bancaud and Talairach developed the anatomoelectroclinical correlations concept in the preimaging era and yet this concept remains fundamental in SEEG interpretation, especially in cases without MRI visible lesion. SEEG allows the analysis of temporo-spatial relationship in seizure generation and spread across distant structures. There is a time lag of variable duration between electrical onset and the appearance of the first clinical signs. What happens within this time lag when ictal discharge precedes clinical onset is a key window analysis to ensure the precise localization of the epileptogenic zone. The variable time lag duration depends on the brain regions involved. The time window between clinical onset and full expression of the clinical semiology is also variable.

In medial temporal seizures, signs and symptoms emerge slowly and progressively while in frontal seizures, signs and symptoms emerge quite rapidly. This is a critical time window to analyze the relationship between clinical semiology and anatomical localization. The local electrical activity modification at the onset of a seizure is insufficient to explain the emergence of semiology. Emergence of semiology is dependent on coactivation or deactivation of cortical areas organized as a paroxysmal network.17 This activation/deactivation can often involve distant cortical areas. This phenomenon is well documented in functional neuroimaging literature.18,​19,​20,​21 In epileptic seizures occurring within neural networks, the transition between interictal and ictal state maybe more or less abrupt. This occurs when a form of “tuning” in which an epileptic discharge within a certain brain region induces activities in the same frequency or its harmonic in another structure which is presumably functionally connected. The whole regions in the epileptic network may also be synchronized in this fashion, contributing to the progression of the clinical semiology (Fig. 17.1).22 The Jacksonian concept of hierarchical levels of organization in the central nervous system suggests that upper levels control the lower level ones. The possible application of such concept in epilepsy disorder is disturbance in the cortical function causing release phenomena (loss of control) that may manifest as disinhibition of motor signs and behavior, which is not the expected functional manifestation of the areas involved. In animal model temporal lobe seizures, the role of synchronized thalamocortical oscillations has been shown to act as seizure amplifier. The result in human study showed increased correlation between thalamus and temporal lobe structures, particularly at the end of the seizures, which also negatively correlates with seizure duration. The proposed theory is, the hyper synchronization can also serve as seizure terminating factor.23

Propagation Theory The classical model of seizure propagation was the corticocortical spread of ictal discharges to the neighboring areas (“march”). An example of this propagation pattern is ­Jacksonian march where clonic ictal discharges in neighboring areas of the somatotopic primary motor cortex produce muscle clonic jerks in the corresponding peripheral territories.24 In fact, the relationship between semiology and the brain network where the epileptic seizures develop is complex. Seizure propagation in SEEG presents as spatiotemporal dynamic depending on the epilepsy organization and the connectivity of the involved regions. Brain regions involved in the production and early propagation of epileptic seizures is called epileptogenic zone or network (Fig. 17.2). This is the one to remove to get a patient seizure free. SEEG had brought us to an understanding that epileptogenic zones are rarely focal, possibly involving a multi-­areal network as a whole at seizure onset. This concept has become increasingly accepted as a reality in the epilepsy community. Moreover, the network concept offers a framework to illustrate the relation between dynamics of brain electrical activity and emergence of clinical semiology.

17 Stereoelectroencephalography

Fig. 17.1  Insulo-opercular cortex generates oroalimentary automatisms in temporal seizures.

„„ Stereoelectroencephalography Indications SEEG is indicated in difficult to localize medically refractory epilepsy where presurgical noninvasive evaluation results point to more than one hypothesis on localization. Some of the advantages of utilizing SEEG exploration are in its ability to give access to deep cortical structures and its capability in mapping three-dimensional aspect of the epileptogenicnetwork in addition to the amenability of bihemispheric exploration with relatively low complication rate. In difficult to localize medically refractory epilepsy, SEEG is indicated to prove a primary hypothesis and to rule out several less likely alternative hypotheses including multifocal seizure onset. Even in cases where the hypothetical epileptogenic zone involves eloquent cortex, oftentimes, SEEG is still preferred over subdural grids. SEEG evaluation is preferred to prioritize epileptogenic zone localization based on the analysis of anatomo-electro-clinical correlations and detailed assessment of particular cerebral function of interest, such as ictal language function in dominant hemisphere epilepsies. A largely distributed implantation without any strong hypothesis (“fishing expedition”) is strongly discouraged. Typical scenarios where SEEG is indicated include the following: 1. Nonlesional epilepsy (MRI negative): The lack of visible anatomical lesion means that preimplantation hypothesis

is dependent on other noninvasive evaluation results, especially the video-EEG and single-photon emission CT. 2. Lesional cases: Lesional cases typically can forego invasive EEG evaluation. This is usually the case if there is a clearly visible lesion and scalp EEG evaluation shows evidence of epileptogenic zone arising from the lesion along other concordant noninvasive workup. However, in cases with discordant evidence of noninvasive workup, such as in dual pathology cases or solitary lesional cases, further investigation to document the anatomical lesion and epileptogenic zone relation is needed and SEEG is the preferred method of exploration in these cases. 3. Underlying need for bihemipsheric explorations: This scenario is commonly found if probable epileptogenic zone is in the mesial cerebral structure. In these cases, the primary hypotheses formed based on the noninvasive workup are often localizing but not very well lateralizing. Bihemispheric explorations can also be indicated in suspected multifocal epileptogenic zone. 4. Surgical failure cases: These failures can be classified as failure to achieve seizure freedom after one-stage resective surgery or failure of subdural grid evaluation in identifying the epileptogenic zone.25 The latter is usually the case when EEG seizure reflects a pattern of propagation from ictal discharges generated elsewhere. This emphasizes the importance of anatomo-electro-clinical concept in analyzing the correlation between EEG changes and the emergence of clinical semiology.

147

148

IIb  Preoperative Electrophysiological Assessment

Fig. 17.2 (a–c)  Epileptogenic zone and semiology: The early spread network.

„„ Stereoelectroencephalography Implantation Strategy SEEG implantation strategy is formulated based on the hypothesis generated by the presurgical evaluation result. The number of electrodes is variable between 5 and 16 electrodes depending on the individual case c­ haracteristics. The electrodes are implanted orthogonally or oblique under general anesthesia through drill holes that was performed by a stereotactically guided robotic arm or s­ urgical frame and secured with bolts. This allows simultaneous recording from mesial and lateral regions. Each electrode contains multiple contacts, which is t­ ypically 8 to 16 contacts and are chosen based on the length of implantation trajectory. The diameters of each contact are generally 0.8 and 1.5 mm apart. Brain imaging such as thin sliced CT is obtained after implantation to confirm placement and to check for potential complications. Postoperative thin sliced CT scan is fused with the preoperative MRI through a dedicated software. A letter of alphabet is assigned for each electrode. When electrodes are located on the left hemisphere, an apostrophe following the letter is used to identify it. The goal of SEEG implantation strategy is to place electrodes in certain anatomical structures and areas determined based on

the hypotheses of probable epileptogenic zones in a safest way possible. In addition, obtaining undersampling information and oversampling cerebral tissue should be avoided. Undersampling can potentially lead to obtaining inadequate information that may lead to failure to define epileptogenic zone and propagation areas while oversampling can lead to untoward complications. The number of electrodes that can be implanted in one patient is limited. Consequently, the accuracy of hypothetic framework and precision in electrodes placement are critical in determining the success or failure of the SEEG exploration.26,​27 Details of surgical implantation techniques and methodology are discussed in Chapter 34 of this book.

„„ Stereoelectroencephalography Recording There are some technical features which are essential to obtain the critical information to define the epileptogenic zone. High-definition video recording performed in a timely manner during ictal events are necessary for describing and interpreting the clinical semiology. The EEG recording software needs to be capable to sample a high frequency rate of at least 1,000 Hz to

17 Stereoelectroencephalography

Fig. 17.3  Interictal activity in focal cortical dysplasia (FCD).

acquire signals from a large number of channels with a reference electrode. The EEG activity is displayed using bipolar recordings between contiguous contacts and a referential montage should also be available. Furthermore, the EEG recording software should also have the capability of recording cortical stimulation and evoked potentials. A single lead of electrocardiogram lead should also be available during the monitoring period.

Stereoelectroencephalography Data Analysis and Interpretation Daily review of the SEEG recording during awake and sleep period will facilitate the recognition of the recording baseline characteristics and the dynamic changes of ictal and interictal patterns as anticonvulsant medications are weaned off or with alteration of other external factors. Montage used for the review should be organized based on anatomical logical grouping on related regions (i.e., anterior to posterior) to aid visual analysis. An epileptologist’s knowledge and experience is crucial in recognizing normal, abnormal, and nonspecific changes in the SEEG pattern. The fact that different anatomical structures might have different normal background rhythms should be taken into consideration.28,​29,​30,​31,​32

Interictal Activities Interictal activities can be influenced by multiple factors. Immediately after the implantation, interictal activities might be suppressed for 12 to 24 hours due to anesthesia effect. Attention to modification of interictal discharges during wakefulness or sleep, reduction of anticonvulsants and during pre- and postictal period is also important in SEEG recording analysis. The presence of normal and abnormal interictal activities, its dynamic changes, abundance, temporospatial distribution,

­relation with each other should be documented. SEEG recording will give overall assessment of the most pathological activities. Determining the most pathological regions can aid to promote further interpretation and analysis of the structure linked in production of pathological activities. Interictal spikes may appear synchronous between close or distant but closely connected structures. They are the clear illustrations of interictal networks. Interictal spikes or spikes networks can also appear independently or concurrently in different regions.33,​34,​35 Another crucial aspect of the interictal analysis is the presence of paroxysmal fast activities. Fast or very fast activity of 100 Hz and above (fast ripples) has been suggested as a marker of epileptogenicity and different patterns of fast ripples may also provide prognostication information. A specific pattern of interictal spiking in combination of fast activities in the same region is highly concerning as pathological region. This has been particularly described in focal cortical dysplasia (FCD; see Fig. 17.3).36,​37,​38 The regions of abnormal interictal activities constitute the irritative zone(s), which can be in the same location as the seizure onset (primary irritative zone) or spatially distinct from the seizure onset (secondary irritative zone).39,​40,​41 The relationship between spikes and fast activities with epileptogenic network remains highly interesting topic of research in the epilepsy community.42

Ictal Activities EEG seizure onset is characterized by a distinct change in the baseline interictal brain rhythm which can occur in a sudden or gradual manner. In SEEG visual analysis, it is important to analyze preictal changes as much as analyzing the first clear ictal electrical changes. These electrical changes should precede onset of clinical signs with variable time lag depending

149

150

IIb  Preoperative Electrophysiological Assessment

Fig. 17.4  Mesial temporal seizure onset.

on the seizure localization. Considering the limited spatial sampling for SEEG, sampling is considered suboptimal if there are no discernable electrical changes preceding the clinical onset. Several seizures need to be captured and analyzed for reproducibility or variability of electroclinical patterns. Semiology analysis with its correlation of delay in electrical changes in the regions involved by the discharges in the early part of clinical semiology emergence is a quintessential part of SEEG interpretation. The emergence and evolution of semiology reflects dynamic activation or deactivation of widespread epileptogenic network that involved various cortical and subcortical structures. The type of electrical discharges and degree of synchrony across structures also affects clinical semiology expression. Sub clinical seizures where there are EEG changes without correlating clinical symptoms can also occur.43 High-frequency discharges seem to be the most specific with seizure onset and epileptogenic network. In mesial temporal structures, beta and low gamma frequency (15–30 Hz) are more often seen, while in neocortical seizures, it is more often associated with higher frequency (30–100 Hz or more) (Fig. 17.4). Seizures related to FCD, the preictal to ictal ­evolution is characterized by increased preictal spikes rhythmicity which then abruptly ceases and concurrently, a fast low-voltage ­discharge appears.36,​44,​45

Postictal Activities Early postictal analysis should be focused on seizure termination, attenuation of the background activities and the earliest recovery of certain regions. In general, the more pathological regions recover from the postictal slowing or attenuation slower. Synchronization reaches its peak during

the final stages of the seizure. Increasing synchronization toward the end of seizures has not only been noted in animal models, but also in several studies of intracranial recorded seizures in humans. There are relatively few documentations in the literature regarding characteristics of postictal changes.46,​47,​48

Stimulation Cortical stimulation is an essential element in SEEG interpretation. Stimulations are performed after capturing spontaneous seizures. When electrical stimulations are applied to the brain regions involved in the epileptogenic network, they may generate full ictal presentation, after discharges with or without clinical semiology or ictal symptoms with or without associated EEG changes.49,​50,​51

Defining the Epileptogenic Zone The goal of SEEG interpretation is to delineate the epileptogenic zone. Epileptogenic zone is used to determine the amount of cerebral structure responsible for seizure generation to be removed during tailored resection to achieve seizure freedom. Epileptogenic zone is not merely a seizure onset zone which is mainly based on the absolute latency of the ictal discharge onset. Epileptogenic zone also considers early propagation zone which is called “early spread network” which is closely linked to clinical semiology emergence. Epileptogenic zone is defined as the region(s) of “primary organization” of ictal discharge which constitutes a pattern which is reproducible among seizures and may be triggered by electrical stimulation. If either of the condition is not met, then there may be a p ­ ossibility of undersampling or there is more than a single epileptogenic zone.

17 Stereoelectroencephalography

„„ Special Considerations for Pediatric Patients There is growing evidence that early surgical intervention in children with medically refractory focal epilepsy can be successful in controlling the seizures and prevent further detrimental consequences in psychomotor development and brain maturation. Selecting surgical candidates in children can be challenging as seizure semiology and EEG seizures may not be as localizing as their older counterparts. Incomplete myelination may also interfere with the chance of identifying cortical dysplasia regions in MRI brain. Therefore, this necessitates invasive EEG evaluation to accurately delineate epileptogenic zone. There is not one ideal invasive EEG evaluation. SEEG exploration is superior in complex cases for mapping the epileptogenic

zone based on the anatomo-electro-clinical analysis (see ­Chapter 34). SEEG has been increasingly used in children now and SEEG surgical complication rates in children is similar with adults.52,​53,​54,​55,​56,​57,​58,​59

„„ Conclusion In summary, SEEG implantation is guided by hypotheses about the brain regions involved in seizure generation. These hypotheses are formed based on the scalp EEG, clinical semiology, and other noninvasive workup results. The goal of SEEG exploration is to delineate epileptogenic zone based on anatomo-electro-clinical analysis to tailor resection in achieving seizure freedom outcome. SEEG is a safe procedure with a low rate of surgical complications.

References 1. Fisher RS, Acevedo C, Arzimanoglou A, et al. ILAE official report: a practical clinical definition of epilepsy. Epilepsia 2014;55(4):475–482 2. Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE commission on therapeutic strategies. Epilepsia 2010;51(6):1069–1077 3. Bickford RG, Dodge HW Jr, Sem-Jacobsen CW, Petersen MC. ­Studies on the electrical structure and activation of an epileptogenic focus. Proc Staff Meet Mayo Clin 1953;28(6):175–180 4. K ovac S, Vakharia VN, Scott C, Diehl B. Invasive epilepsy surgery evaluation. Seizure 2017;44:125–136 5. Talairach J, David M, Tournoux P. L’exploration Chirugicale ­Stereotaxique du Lobe Temporale dans L’epi`lepsie Temporale. Paris: Masson et Cie; 1958 6. Schijns OE, Hoogland G, Kubben PL, Koehler PJ. The start and development of epilepsy surgery in Europe: a historical review. Neurosurg Rev 2015;38(3):447–461 7. Reif PS, Strzelczyk A, Rosenow F. The history of invasive EEG evaluation in epilepsy patients. Seizure 2016;41:191–195 8. Feindel W, Leblanc R, de Almeida AN. Epilepsy surgery: historical highlights 1909–2009. Epilepsia 2009;50(Suppl 3):131–151 9. Wilson SJ, Engel J Jr. Diverse perspectives on developments in epilepsy surgery. Seizure 2010;19(10):659–668 10. Bancaud J, Talairach J. La Stéréo-ÉlectroEncéphaloGraphie Dans l’Épilepsie. In: Bancaud J, Talairach J, eds. Paris: Masson et Cie; 1965 11. Bancaud J, Ribet MF, Chagot D. Origine comparée des paroxysmes de pointes “infra-clinique” et des crises électro-cliniques spontanées dans l’épilepsie. Rev Electroencephalogr Neurophysiol Clin 1975;5(1):63–66 12. Kahane P, Landré E, Minotti L, Francione S, Ryvlin P. The Bancaud and Talairach view on the epileptogenic zone: a working hypothesis. Epileptic Disord 2006;8(Suppl 2):S16–S26 13. Isnard J, Taussig D, Bartolomei F, et al. French guidelines on stereoelectroencephalography (SEEG). Neurophysiol Clin 2018;48(1)5–13 14. Chauvel P, Vignal J, Biraben A, et al. Stereoelectroencephalography. Multimethodological Assessment of the Epileptic Focus. New York, NY: Springer Verlag; 1996:80–108 15. Munari C, Talairach J, Musolino A, et al. Stereotactic methodology of functional neurosurgery in tumoral epileptic patients. Ital J Neurol Sci 1983;2(suppl):69–82 16. Chauvel P, Buser P, Badier JM, Liegeois-Chauvel C, Marquis P, ­Bancaud J. [The “epileptogenic zone” in humans: ­representation

of intercritical events by spatio-temporal maps.] Rev Neurol (Paris) 1987;143(5):443–450 17. Chauvel P, McGonigal A. Emergence of semiology in epileptic seizures. Epilepsy Behav 2014;38:94–103 18. Friston KJ. Functional and effective connectivity: a review. Brain Connect 2011;1(1):13–36 19. Stephan KE, Friston KJ. Analyzing effective connectivity with functional magnetic resonance imaging. Wiley Interdiscip Rev Cogn Sci 2010;1(3):446–459 20. Campo P, Garrido MI, Moran RJ, et al. Network reconfiguration and working memory impairment in mesial temporal lobe epilepsy. Neuroimage 2013;72:48–54 21. Pinotsis DA, Hansen E, Friston KJ, Jirsa VK. Anatomical connectivity and the resting state activity of large cortical networks. Neuroimage 2013;65:127–138 22. Aupy J, Noviawaty I, Krishnan B, et al. Insulo-opercular cortex generates oroalimentary automatisms in temporal seizures. Epilepsia 2018;59(3):583–594 23. Guye M, Régis J, Tamura M, et al. The role of corticothalamic coupling in human temporal lobe epilepsy. Brain 2006;129(Pt 7):1917–1928 24. Chauvel P, Trottier S, Vignal JP, Bancaud J. Somatomotor seizures of frontal lobe origin. Adv Neurol 1992;57:185–232 25. Vadera S, Mullin J, Bulacio J, Najm I, Bingaman W, Gonzalez-­ Martinez J. Stereoelectroencephalography following subdural grid placement for difficult to localize epilepsy. Neurosurgery 2013;72(5):723–729, discussion 729 26. Cardinale F, Casaceli G, Raneri F, Miller J, Lo Russo G. Implantation of stereoelectroencephalography electrodes: a systematic review. J Clin Neurophysiol 2016;33(6):490–502 27. Kalamangalam GP, Tandon N. Stereo-EEG implantation strategy. J Clin Neurophysiol 2016;33(6):483–489 28. Bulacio JC, Chauvel P, McGonigal A. Stereoelectroencephalography: Interpretation. J Clin Neurophysiol 2016;33(6):503–510 29. Serletis D, Bulacio J, Bingaman W, Najm I, González-Martínez J. The stereotactic approach for mapping epileptic networks: a prospective study of 200 patients. J Neurosurg 2014;121(5): 1239–1246 30. Gonzalez-Martinez JA. The stereo-electroencephalography: the epileptogenic zone. J Clin Neurophysiol 2016;33(6): 522–529 31. Bartolomei F, Lagarde S, Wendling F, et al. Defining epileptogenic networks: contribution of SEEG and signal analysis. Epilepsia 2017;58(7):1131–1147

151

152

IIb  Preoperative Electrophysiological Assessment 32. Lopes da Silva F, Blanes W, Kalitzin SN, Parra J, Suffczynski P, Velis DN. Epilepsies as dynamical diseases of brain systems: basic models of the transition between normal and epileptic activity. Epilepsia 2003;44(Suppl 12):72–83 33. Bourien J, Bartolomei F, Bellanger JJ, Gavaret M, Chauvel P, Wendling F. A method to identify reproducible subsets of co-activated structures during interictal spikes. Application to intracerebral EEG in temporal lobe epilepsy. Clin Neurophysiol 2005;116(2):443–455 34. Bartolomei F, Wendling F, Régis J, Gavaret M, Guye M, Chauvel P. Pre-ictal synchronicity in limbic networks of mesial temporal lobe epilepsy. Epilepsy Res 2004;61(1–3):89–104 35. Spencer SS, Goncharova II, Duckrow RB, Novotny EJ, Zaveri HP. Interictal spikes on intracranial recording: behavior, physiology, and implications. Epilepsia 2008;49(11):1881–1892 36. Wendling F, Bartolomei F, Bellanger JJ, Bourien J, Chauvel P. Epileptic fast intracerebral EEG activity: evidence for spatial decorrelation at seizure onset. Brain 2003;126(Pt 6):1449–1459 37. Kerber K, Dümpelmann M, Schelter B, et al. Differentiation of specific ripple patterns helps to identify epileptogenic areas for surgical procedures. Clin Neurophysiol 2014;125(7): 1339–1345 38. Chassoux F, Devaux B, Landré E, et al. Stereoelectroencephalography in focal cortical dysplasia: a 3D approach to delineating the dysplastic cortex. Brain 2000;123(Pt 8):1733–1751 39. Badier JM, Chauvel P. Spatio-temporal characteristics of paroxysmal interictal events in human temporal lobe epilepsy. J Physiol Paris 1995;89(4)(–)(6):255–264 40. Malinowska U, Bergey GK, Harezlak J, Jouny CC. Identification of seizure onset zone and preictal state based on characteristics of high frequency oscillations. Clin Neurophysiol 2015;126(8):1505–1513 41. Malinowska U, Badier JM, Gavaret M, Bartolomei F, Chauvel P, Bénar CG. Interictal networks in magnetoencephalography. Hum Brain Mapp 2014;35(6):2789–2805 42. Roehri N, Pizzo F, Lagarde S, et al. High-frequency oscillations are not better biomarkers of epileptogenic tissues than spikes. Ann Neurol 2018;83(1):84-97 43. Perucca P, Dubeau F, Gotman J. Intracranial electroencephalographic seizure-onset patterns: effect of underlying pathology. Brain 2014;137(Pt 1):183–196 44. Fisher RS, Webber WR, Lesser RP, Arroyo S, Uematsu S. High-­ ­ frequency EEG activity at the start of seizures. J Clin ­Neurophysiol 1992;9(3):441–448 45. Gnatkovsky V, de Curtis M, Pastori C, et al. Biomarkers of epileptogenic zone defined by quantified stereo-EEG analysis. Epilepsia 2014;55(2):296–305

46. Toussaint D, Moura M, Allouche L, et al. Can Early Post-Ictal Activities Help to Better Localise and Lateralise the Epileptogenic Zone. Epilepsia. Oxon: Blackwell; 2005:324 47. Schiff SJ, Sauer T, Kumar R, Weinstein SL. Neuronal spatiotemporal pattern discrimination: the dynamical evolution of seizures. Neuroimage 2005;28(4):1043–1055 48. 48. Kramer MA, Eden UT, Kolaczyk ED, Zepeda R, E ­ skandar EN, Cash SS. Coalescence and fragmentation of cortical networks during focal seizures. J Neurosci 2010;30(30): 10076–10085 49. Trébuchon-Da Fonseca A, Bénar CG, Bartoloméi F, et al. Electrophysiological study of the basal temporal language area: a convergence zone between language perception and production networks. Clin Neurophysiol 2009;120(3):539–550 50. Chauvel P, Landré E, Trottier S, et al. Electrical stimulation with intracerebral electrodes to evoke seizures. Adv Neurol 1993;63:115–121 51. Kovac S, Kahane P, Diehl B. Seizures induced by direct electrical cortical stimulation—mechanisms and clinical considerations. Clin Neurophysiol 2016;127(1):31–39 52. Cossu M, Schiariti M, Francione S, et al. Stereoelectroencephalography in the presurgical evaluation of focal epilepsy in infancy and early childhood. J Neurosurg Pediatr 2012;9(3): 290–300 53. Taussig D, Dorfmüller G, Fohlen M, et al. Invasive explorations in children younger than 3 years. Seizure 2012;21(8): 631–638 54. Taussig D, Chipaux M, Lebas A, et al. Stereo-electroencephalography (SEEG) in 65 children: an effective and safe diagnostic method for pre-surgical diagnosis, independent of age. Epileptic Disord 2014;16(3):280–295 55. Freri E, Matricardi S, Gozzo F, Cossu M, Granata T, Tassi L. P ­ erisylvian, including insular, childhood epilepsy: presurgical workup and surgical outcome. Epilepsia 2017;58(8): 1360–1369 56. Gonzalez-Martinez J, Mullin J, Bulacio J, et al. Stereoelectroencephalography in children and adolescents with d ­ifficultto-­ localize refractory focal epilepsy. Neurosurgery 2014;75 (3):258–268, discussion 267–268 57. Gonzalez-Martinez J, Lachhwani D. Stereoelectroencephalography in children with cortical dysplasia: technique and results. Childs Nerv Syst 2014;30(11):1853–1857 58. Gonzalez-Martinez J, Najm IM. Indications and selection criteria for invasive monitoring in children with cortical dysplasia. Childs Nerv Syst 2014;30(11):1823–1829 59. Cossu M, Cardinale F, Castana L, Nobili L, Sartori I, Lo Russo G. Stereo-EEG in children. Childs Nerv Syst 2006;22(8):766–778

18

  Magnetoencephalography Hiroshi Otsubo, Kota Kagawa, and O. Carter Snead III

Summary Localization-related epilepsy is more often associated with an extratemporal epileptogenic focus in children than that seen in adults. The neocortical epileptic zones frequently are adjacent to eloquent cortex, and the surgical treatment requires accurate delineation of both epileptogenic and functional zones. Magnetoencephalography (MEG) has been reported to localize both epileptogenic and eloquent cortices in children. Here we discuss basic principles of MEG, magnetoencephalography spike sources (MEGSSs), and the clinical applications of MEG in: (1) lesional epilepsy such as focal cortical dysplasia (FCD) and tuberous sclerosis complex (TSC); (2) extratemporal lobe epilepsy; (3) insular epilepsy; (4) temporal lobe epilepsy; (5) nonlesional epilepsy; and (6) recurrent/residual seizures. Keywords:  epileptogenic zone, epileptic network, FCD at the bottom of sulcus, bottom of sulcus plus, FCD at the brain surface

„„ Introduction Localization-related, or focal, epilepsy refractory to antiepileptic drugs is more often associated with an extratemporal epileptogenic focus in children than that seen in adults. Thus invasive intracranial electroencephalography (EEG) with extraoperative subdural electrode recordings to localize the epileptogenic zone often is necessary in children. These neocortical epileptic zones frequently are adjacent to eloquent cortex, and the surgical treatment requires accurate delineation of both epileptogenic and functional zones. MEG has been reported to be a valuable noninvasive technique that can be used to localize both epileptogenic and eloquent cortices in children with medically refractory focal epilepsy who are undergoing evaluation for surgical treatment of their seizure disorder.1,​2,​3,​4,​5,​6

„„ Basic Principles of Magnetoencephalography and Magnetic Source Imaging MEG is a technique for measuring the magnetic fields associated with the intracellular current flows within neurons. Source localization of epileptic spikes and evoked responses as

determined by MEG are coregistered with MRI as magnetic source imaging (MSI).7 MEG is based on the physical phenomenon that intracellular neuronal electrical currents generate accompanying magnetic fields. The orientation of the magnetic field relative to the electrical current is described as Orsted’s “right-hand rule,” which states that when the thumb of the right hand is pointed in the direction of the electrical current, the surrounding magnetic flux is aligned in the direction of the right fingers. MEG uses highly sensitive biomagnetometers to detect extracranial magnetic fields produced by intracellular neuronal currents. On the basis of the right-hand rule, MEG is primarily sensitive to signals arising from regions in which the apical dendrites are tangentially oriented to the skull and scalp surface. The source localization has to solve the inverse problem that calculates the three-dimensional intracranial location, orientation, and strength of the neuronal sources backward from a measured extracranial magnetic field pattern. The accuracy of a solution of the inverse problem depends on numerous factors, including the forward problem. The forward problem uses an iterative algorithm to determine the location, orientation, and strength of the equivalent current dipole that best account for the m ­ easured magnetic field pattern. The accuracy of the forward problem is critically determined by the shape and conductivity of the volume conductor of head model. MEG forward solution is more robust than that of EEG because of homogeneous conductivity in a magnetic field. Therefore, the localization of both MEGSS and evoked responses on MSI is quite reliable for presurgical evaluation in pediatric localization-related epilepsy.2 In short, MEG is an extremely valuable and reliable technique with which to localize the source of interictal epileptiform discharges.8

„„ Magnetoencephalography Spike Sources The Hospital for Sick Children in Toronto, Canada, has pioneered the use of MEG for clinical application in pediatric epilepsy. From August 2000 to December 2016, MEG was utilized in more than 1,300 patients with focal epilepsy as part of a presurgical protocol that also includes careful definition of seizure semiology based on clinical features and prolonged scalp video-EEG, MRI, and neuropsychological testing.2 More

154

IIb  Preoperative Electrophysiological Assessment than 800 of these children have progressed to epilepsy surgery based on these data. We have defined the distribution of MEGSS by number and density.6 An MEG spike cluster is six or more spike sources with 1 cm or less between adjacent sources. An MEG spike scatters is fewer than six spike sources regardless of the distance between sources or spike sources with more than 1 cm between sources regardless of the number of sources in a group. The zone of clustered MEGSS correlates with the ictal onset zone and the prominent interictal zone as determined by extraoperative intracranial video-EEG as recorded from subdural electrodes.6,​9,​10 MEG spike scatters alone should be examined by intracranial video-EEG, because an epileptic zone may exist within the scatter distribution of MEGSS. We have shown that complete resection of MEG clusters is correlated with postsurgical seizure freedom.11,​12 For presurgical evaluation, concordant MEG localization to the region of resection, invasive electrocorticography, and MRI abnormality is associated with improved seizure outcome.13 Discordant lateralization of EEG spike sources and MEGSSs indicate an undetermined epileptogenic hemisphere and contraindicate surgery without further testing.14

„„ Lesional Epilepsy

Fig. 18.1  Schematic presentation of three types of focal cortical dysplasia (FCD) type II and magnetoencephalography (MEG) spike dipoles. (a) The remote cluster of MEG spike dipoles and the MRI-visible lesion of FCD at the bottom of sulcus. (b) The partially overlapping cluster of MEG spike dipoles and FCDs indicate visible lesion on MRI with consecutively extending epileptogenic zone (the bottom of sulcus plus). (c) The overlapping cluster of MEG spike dipoles and FCD at the brain surface. Note that MEG spike dipoles represent the center of gravity within the extensive FCD at the brain surface.23

Surgical treatment of seizure disorders secondary to a lesion requires that the lesion be removed and epileptogenic tissue removed or disconnected. MSI provides accurate data on the spatial relations of lesion, epileptogenic zone, and functional cortex in children with lesional extratemporal epilepsy.4 MEG delineates asymmetric epileptogenic zones surrounding lesions and eloquent cortex.15 When the focal seizures are secondary to a neoplasm, complete tumor resection with resection of MEGSS marginal to the tumor is associated with favorable outcomes despite residual postexcisional ECOG spikes and extramarginal MEGSS.4 When the focal seizures are secondary to dysplastic brain, the cortical dysplasia as characterized by clusters of MEGSS within and extending from MRI lesion should be removed completely, including both the anatomic lesion and MEGSS, to achieve seizure freedom.16,​17,​18

of a deep sulcus.22 Nakajima et al investigated spatial congruence between MEG spikes and MRI lesion at the bottom of sulcus (Fig. 18.1). MEGSS were partially overlapping MRI lesion in four of eight patients and discordant in another four patients. Lesionectomy of MRI lesion can control seizures for patients with discordant MEGSS. MRI lesions partially overlapped with MEGSS may have an extending epileptogenic zone consecutive to the MRI visible lesion. In these cases, the localization of MEGSS can be discordant to the MRI lesion because of: (1) closed-field effect; (2) lower neuronal density; and (3) small size.23

„„ Focal Cortical Dysplasia

„„ Tuberous Sclerosis Complex

FCD is a migrational abnormality of the brain that is intrinsically epileptogenic and causes refractory epilepsy that is surgically curable.17 MEG is particularly useful in the presurgical evaluation of patients with FCD because the lesions at cortical sulci generate radial biomagnetic fields that are ideally detectable by MEG sensors.19 Routine MRI studies may fail to identify lesions in a subset of FCD. However, a high resolution, thin slice special MRI study targeting MEGSS localization can reveal MRI negative FCD.20 Presurgical delineation of FCD is essential for surgical planning and resection of the epileptogenic lesion. Widjaja et al has shown that 11 (85%) of 13 patients with complete resection of clustered MEGSS and 15 (88%) of 17 patients with complete removal of MRI lesions achieved postoperative seizure freedom.21 In contrast to the epileptogenic zone in cases of FCD at the brain surface, many small FCD lesions are detected at the bottom

MEG has proven useful in identifying children with TSC who may be candidates for epilepsy surgery (Fig. 18.2). In 2006, Wu et al studied six children with focal seizures secondary to TSC.24 In six TSC patients with focal seizures secondary to bilateral multilobar cortical tubers, ictal video-EEG predicted the region of resection with 56% sensitivity, 80% specificity, and 77% accuracy. Interictal MEG, however, fared better, with 100% sensitivity, 94% specificity, and 95% accuracy. In TSC, MEGSS tend to localize around visible tubers. Okanishi et al have reported that the maximal possible resection of scattered MEGSS in addition to the clustered MEGSS correlated with improved seizure outcome in children with TSC.25 MEG enabled precise localization of the epileptic foci and provided crucial information of surgical treatment in children with localization-related epilepsy secondary to TSC.26,​27,​28,​29

18 Magnetoencephalography

Fig. 18.2  MEG in the presurgical evaluation of a child with tuberous sclerosis. Axial T2 MRI shows multiple cortical tubers, magnetoencephalography spike sources (MEGSSs), and auditory-evoked field. This 17-year-old, right-handed boy presents intractable epilepsy secondary to tuberous sclerosis complex. His seizures consist of gagging followed by clonic movements of face and left upper extremity. MEG shows a total of 70 MEGSSs consisting of two clusters over the right hemisphere. (a) Axial T2 MRI shows 4 of 46 clustered MEGSSs over the right temporooccipital region around the occipital cortical tuber. (Closed triangles represent the location of MEGSSs and tails indicate the orientation of the MEGSS.) A closed square represents auditory-evoked field. (b) Axial T2 MRI shows 3 of 24 clustered MEGSSs in the right inferior frontal to superior Rolandic region, superior and posterior to the frontal cortical tuber. He underwent intracranial video-electroencephalography monitoring using 103 subdural grid and depth electrodes. Cortical resection was performed over the right occipitotemporal region and inferior frontal region, which correlated to the two MEGSS clusters. He has been seizure free for 9 months with medications.

„„ Extratemporal Lobe Epilepsy In infants and young children, the occipital lobe frequently generates focal onset seizures and even infantile spasms.30 More occipital spikes migrate anteriorly than frontal spikes migrate posteriorly in children.31 Ibrahim has reported that overall, 68% of 41 patients with refractory occipital lobe epilepsy had a satisfactory seizure outcome following surgery.32 They identified younger age at the seizure onset as being associated with a greater number of seizure types and more extensive resection needed to yield seizure freedom. Therefore, in younger patients with extratemporal localization-related epilepsy, multiple clustered MEGSS are often seen in temporal/parietal/occipital lobes, whereas in older patients, single clusters are observed frequently with an ictal onset zone in the frontal lobe.9 These data suggest that posteriorly dominant epilepsy can extend anteriorly to expand the epileptic network through anatomic and functional connections in developing brains, whereas frontal lobe epilepsy less frequently migrates to other lobes. Therefore, multiple clustered MEGSS associated with the posterior epileptic network may require extensive resection, especially in young children. Conversely, the single cluster that correlates with a discrete anterior epileptic region in relatively old patients may predict a successful focal resection.

The diagnosis of frontal lobe epilepsy may be compounded by poor electroclinical localization on scalp EEG, caused by deep, distributed, or rapidly propagating epileptiform activities over the bilateral hemispheres. The yield of MEGSS in terms of localization of the epileptogenic zone in frontal lobe epilepsy is superior to that of EEG because of high spatial and temporal resolution with the former (Fig. 18.3).33 Mu et al have reported that 58.1% of 31 patients with a single MEG spike cluster and 0% of 7 patients with multiple clusters in frontal lobe epilepsy became seizure free following surgery.34 When interictal epileptiform discharges on scalp EEGs show a diffuse hemispheric distribution, or bilateral synchronous spike-waves, analysis of MEGSS at the earliest time point or dynamic statistical parametric maps can lateralize and localize the epileptogenic zone.35,​36 As MEG is noninvasive and can monitor the entire brain, it is superior to intracranial video-EEG in terms of investigating the propagation of epileptic activity.37 Shirozu et al. applied gradient magnetic-field topography (GMFT) to visualize the extent and dynamic changes of epileptic activity on a three-dimensional brain-surface image.38 The distribution of GMFT at the onset of MEG spike was found to overlap with the ictal onset zone observed during intracranial video-EEG and to correlate with the surgical area removed that led to postoperative seizure freedom. The d ­ istribution

155

156

IIb  Preoperative Electrophysiological Assessment

Fig. 18.3  MEG in presurgical evaluation of a child with extratemporal, localization-related epilepsy. T1 MRI shows magnetoencephalography spike sources (MEGSSs) and somatosensory-evoked field (SEF). This 17-year-old right-handed boy presents with sensory aura with or without secondarily generalized tonic–clonic seizures. MRI showed small nonspecific high fluid-attenuation inversion recovery signal seen in the right peri-Rolandic region. (a) Axial T1 MRI shows 6 of clustered 75 MEGSSs over the postcentral gyrus. (Closed triangles represent the location of MEGSSs and tails indicate the orientation of the MEGSS.) Open black circle represents SEF by left median nerve stimulation. (b) Sagittal T1 MRI shows seven MEGSSs around the central sulcus with predominant postcentral gyrus spreading to supramarginal gyrus. The open black circle represents SEF. He underwent intracranial videoelectroencephalography monitoring using 120 subdural electrodes over the right frontoparietal region. Corticectomy of right postcentral gyrus, including clustered MEGSSs, was performed. The pathology was reported cortical dysplasia type IIB. He has been seizure free for 5 months with medications.

of GMFT at the peak of the MEG spike was always larger than that at the onset. In age-related epilepsy, benign Rolandic epilepsy (BRE) and Landau–Kleffner syndrome (LKS) are forms of childhood epilepsy that share particular characteristics and can be controlled with medication. Both BRE and LKS have identical orientation of MEGSSs which direct vertical to the Rolandic fissure39 and Sylvian fissure,40 respectively. However, a subgroup of patients who manifest some of the characteristic of both BRE and LKS, but who do not fulfill all criteria for these epilepsy syndromes have been designated atypical BRE and LKS variant. We have introduced the term malignant Rolandic–Sylvian epilepsy to describe this subgroup, which is characterized by frontocentrotemporal spikes on EEG, absence of lesions on MRI, MEGSS with random orientations around Rolandic and Sylvian fissures, intractable sensorimotor partial seizures that progress to secondary generalization, and neurocognitive problems.3

„„ Insular Epilepsy The insula is located deep below the opercular cortices and can have extensive networks with other potentially ­epileptogenic areas. Insular seizures mimic or exist alongside ­seizures originating from adjacent frontal, temporal, or parietal lobe.41,​42 Presurgical noninvasive studies have significant ­limitations in this brain region. They cannot reliably rule out insular epilepsy.43 Ictal or interictal scalp EEG has insufficient spatial resolution to differentiate insular from overlying ­frontal, ­parietal, or temporal lobe epilepsy.44 Positron emission

­tomography (PET) or single-photon emission (SPE) CT may provide misleading results or multifocal abnormalities.45 Mohamed et al has reported that MEG is superior to PET and ictal SPECT in detecting insular lobe epilepsy.46 These authors investigated 14 patients with insular epilepsy, and found three patterns of MEGSS: (1) an anterior operculoinsular cluster (7 patients); (2) posterior operculoinsular cluster (2 patients); and (3) a ­diffuse perisylvian distribution (4 patients); no spikes were detected in one patient. Six patients with anterior operculoinsular cluster, 1 with posterior operculoinsular cluster, and 2 with diffuse perisylvian distribution underwent insular epilepsy surgery with a favorable postsurgical outcome (Engel 1 or 2).47 Among patients with anterior operculoinsular cluster who proceeded to have surgery, MEG provided superior information when compared to ictal SPECT in four of six patients and to interictal PET in five of six patients.

„„ Temporal Lobe Epilepsy Temporal lobectomy in children for temporal lobe epilepsy has a seizure-free outcome similar to that reported in adults.48 Unlike extratemporal localization-related epilepsy, MEGSS in temporal lobe epilepsy do not represent the exact location of the source of interictal epileptiform discharges.8 There are five reasons for the failure of MEG in this regard: 1. Mesial temporal areas are farther from MEG sensors.49 Because magnetic fields attenuate in square proportion to the distance from the source,50 there are less prominent MEG spikes with mesial temporal discharges.

18 Magnetoencephalography 2. The cylindrical architecture of hippocampal neurons cancels the generated excitatory postsynaptic potentials (closed circuit), in contrast to the linear and laminar architecture of neocortical neurons (open circuit).49 3. Insufficient coverage of the subtemporal magnetic fields by a whole-head MEG sensor array increases errors for dipole estimation. 4. The propagation of epileptiform discharges to surrounding temporal structures through the limbic network is not suitable for application of single dipole analysis.51 5. Magnetic fields from lateral and superior temporal cortices overwhelm those from mesial temporal structures.

likely to occur in children with bilateral MEG clusters or only scatters, multiple seizure types, and incomplete resection of the proposed epileptogenic zone. MEGSS can significantly guide the surgical strategy of invasive EEG monitoring in patients with nonlesional epilepsy.11,​12,​58 Clustered MEGSS can be factored into the surgical planning for intracranial video-EEG monitoring by utilization of surgical navigation techniques to guide the placement of tailored subdural grids and the targeting of depth electrodes via stereotactic EEG.59,​60,​61,​62,​63 These complementary approaches have been demonstrated to lead to successful seizure control.

MEG is more valuable in those cases where the temporal lobe is part of a wider circuitry in localization-related epilepsy. For example, in a child with intractable epilepsy secondary to a temporo-parieto-occipital porencephalic cyst after encephalitis, vertically oriented MEGSS were obtained without superolateral temporal cortices. The absence of superolateral temporal cortices, prominent temporal EEG spikes, less prominent MEG spikes, and mesiobasal synthetic aperture magnetometry spikes using spatial filtering method all indicated that the vertically oriented MEGSS were projected directly from the mesiobasal temporal region.52 In another example, a 9-year-old boy with coexisting benign Rolandic epileptiform discharges with intractable mesial temporal lobe epilepsy secondary to hippocampal sclerosis on MRI was studied with, scalp video-EEG. The video-EEG showed left temporal rhythmic sharp waves after the clinical onset of epigastric aura, followed by staring.53 An MEG identified Rolandic MEGSS, which were prominent on the scalp EEG as well. However, MEG was unable to localize the epileptogenic sources in the temporal lobe because the higher amplitude signals of the Rolandic spikes masked the lower amplitude spikes from the mesial temporal network in this case. The benign form of Rolandic MEGSS in which orientation of dipoles are identical and vertical to central sulcus similar to those of BRE are often seen in children as an age-related phenomenon and occasionally seen in adults with temporal lobe epilepsy.

„„ Recurrent or Residual Seizures after Surgery

„„ Nonlesional Epilepsy Nonlesional epilepsy represents a challenge for epilepsy surgical evaluation in children. In a cohort of 75 children younger than 12 years of age who underwent resective surgery for intractable epilepsy at a pediatric epilepsy center, 35 had no identifiable focal lesion on MRI.54 In addition, some researchers have shown that MRI does not aid in the presurgical evaluation in nearly 29% of patients in whom it is normal or shows ­nonspecific findings.55 The outcome after epilepsy surgery in patients with normal brain MRI depends on the case selection criteria and expertise of the epilepsy center. Surgery for intractable epilepsy in children with normal MRI findings but clustered MEGSS provided good postsurgical outcomes in the majority.19,​56,​57 Restricted ictal onset zone predicted postoperative seizure freedom. Seizure freedom was most likely to occur when there was concordance between EEG and MEG localization and least likely to occur when these results were divergent. Postoperative seizure freedom was less

The success rate of surgery in extratemporal epilepsies, which are particularly common in children, continues to be disappointing, with a 27 to 46% seizure-free rate in long term follow-up by meta-analysis.64 Standard MRI techniques used postoperatively in patients after epilepsy surgery may miss the extent of the residual lesion. Similarly, postoperative ictal scalp EEG findings are misleading because of skull defects, dural scarring, cerebrospinal fluid-filled intracranial cavities, and alterations or distortions of brain structures from a previous surgery. In patients who have a second epilepsy surgery after the initial one failed to control seizures, the interpretation of invasive EEG as recorded from subdural electrodes becomes complicated by differences in amplitude between normal and gliotic cortical surfaces at the site of previous surgery. Specific MEGSS patterns delineated the epileptogenic zone in 17 children with recurrent seizures after previous epilepsy surgery.65 The clustered MEGSS occurred at the margins of previous resections within two contiguous gyri in 10 patients (group A), extended spatially from a margin by 3 cm or less in 3 patients (group B), and were remote from the resection margin by more than 3 cm in 6 patients (group C). Two patients had concomitant group A and C clusters. Eleven of 13 children who underwent repeat surgeries that included resection of the area of clustered MEGSSs obtained favorable surgical outcomes. MEG is particularly advantageous in those children in whom a second epilepsy surgery is being contemplated, because in children who have had previous surgery, MEG signals are far less distorted by postoperative skull defects, subdural scarring, arachnoid adhesions, and shifting of the normal brain into resection cavities than the scalp EEG. Thus, MEG can identify the recurrent epileptogenic zone for the subgroup of patients with late recurrent seizures after epilepsy surgery.

„„ Functional Mapping A successful outcome from epilepsy surgery is generally defined as a seizure-free state with no imposition of neurological ­deficit.66 To achieve these twin goals, two criteria must be ­fulfilled. First, precise localization of the epileptogenic zone in the brain is necessary. Second, one must determine the anatomic localization of eloquent cortex that subserves sensory, motor, language, and memory function. Therefore, the neurosurgeon

157

158

IIb  Preoperative Electrophysiological Assessment requires the precise anatomic correlation between the epileptogenic zone and eloquent cortex before surgery. Noninvasive MEG studies are now used routinely in some centers to localize eloquent cortex in patients undergoing epilepsy surgery. The somatosensory-evoked magnetic field (SEF) for m ­ edian nerve stimulation is now widely accepted a most reliable ­method for identifying the primary somatosensory cortex and localization of the central sulcus.67 Because the N20m component of SEF reflects the direct neuronal activity of primary ­sensory cortex, the SEF is generated from the posterior bank of the central sulcus. Pihko et al68 successfully measured the SEF in normal newborns during their frequent postprandial sleep. However, sleep recordings are less feasible in older infants because of their shortened sleep cycles. Therefore, total intravenous anesthesia using propofol has been applied for MEG and MRI studies in uncooperative children.69 We analyzed 26 infants younger than 4 years of age under total intravenous anesthesia using propofol and showed that SEFs can still be detected and reliably observed under these conditions.70 MEG source localization of evoked fields can address whether functional reorganization of

primary sensory modalities exists in malformations of cortical development.71 MEG can also identify motor cortex. Movement-related cerebral magnetic fields following voluntary finger movement have demonstrated a unique area of motor control in children.72 The auditory-evoked magnetic field is used to identify the primary auditory cortex. The prominent components of N100m around 100 ms after contralateral audio stimulation represent the auditory-evoked magnetic field in the Heschl’s gyrus and the planum temporale.73,​74 Similarly, the visual-evoked magnetic field is used to localize the primary visual cortex. P100m at around 100 ms after visual stimulation produces visual-evoked magnetic fields in the mesial occipital region.75 MEG has been reported to be useful in the lateralization and localization of language in seizure patients.76,​77,​78 During MEG recordings, patients engaged in a word recognition task have been shown to activate language areas. Excellent agreement has been reported between MEG data and those obtained from Wada test.79 In addition, there is good correlation between MEG and intraoperative direct cortical mapping in terms of localization of receptive language areas.80

References 1. Wheless JW, Willmore LJ, Breier JI, et al. A comparison of magnetoencephalography, MRI, and V-EEG in patients evaluated for epilepsy surgery. Epilepsia 1999;40(7):931–941

13. Englot DJ, Nagarajan SS, Imber BS, et al. Epileptogenic zone localization using magnetoencephalography predicts seizure freedom in epilepsy surgery. Epilepsia 2015;56(6):949–958

2. Minassian BA, Otsubo H, Weiss S, Elliott I, Rutka JT, Snead OC III. Magnetoencephalographic localization in pediatric epilepsy surgery: comparison with invasive intracranial electroencephalography. Ann Neurol 1999;46(4):627–633

14. Ochi A, Otsubo H, Iida K, et al. Identifying the primary epileptogenic hemisphere from electroencephalographic (EEG) and magnetoencephalographic dipole lateralizations in children with intractable epilepsy. J Child Neurol 2005; 20(11):885–892

3. Otsubo H, Chitoku S, Ochi A, et al. Malignant Rolandic–Sylvian epilepsy in children: diagnosis, treatment, and outcomes. Neurology 2001;57(4):590–596 4. Otsubo H, Ochi A, Elliott I, et al. MEG predicts epileptic zone in lesional extrahippocampal epilepsy: 12 pediatric surgery cases. Epilepsia 2001;42(12):1523–1530 5. Pataraia E, Simos PG, Castillo EM, et al. Does magnetoencephalography add to scalp video-EEG as a diagnostic tool in epilepsy surgery? Neurology 2004;62(6):943–948 6. Iida K, Otsubo H, Matsumoto Y, et al. Characterizing magnetic spike sources by using magnetoencephalography-guided neuronavigation in epilepsy surgery in pediatric patients. J Neurosurg 2005;102(2, Suppl):187–196 7. Holowka SA, Otsubo H, Iida K, et al. Three-dimensionally reconstructed magnetic source imaging and neuronavigation in pediatric epilepsy: technical note. Neurosurgery 2004;55(5):1226 8. Ebersole JS. Defining epileptogenic foci: past, present, future. J Clin Neurophysiol 1997;14(6):470–483 9. Oishi M, Kameyama S, Masuda H, et al. Single and multiple clusters of magnetoencephalographic dipoles in neocortical epilepsy: significance in characterizing the epileptogenic zone. Epilepsia 2006;47(2):355–364 10. Agirre-Arrizubieta Z, Huiskamp GJM, Ferrier CH, van Huffelen AC, Leijten FS. Interictal magnetoencephalography and the irritative zone in the electrocorticogram. Brain 2009;132(Pt 11):3060–3071 11. Ochi A, Otsubo H. Magnetoencephalography-guided epilepsy surgery for children with intractable focal epilepsy: SickKids experience. Int J Psychophysiol 2008;68(2):104–110 12. Albert GW, Ibrahim GM, Otsubo H, et al. Magnetoencephalography-guided resection of epileptogenic foci in children. J Neurosurg Pediatr 2014;14(5):532–537

15. Bennett-Back O, Ochi A, Widjaja E, et al. Magnetoencephalography helps delineate the extent of the epileptogenic zone for surgical planning in children with intractable epilepsy due to porencephalic cyst/encephalomalacia. J Neurosurg Pediatr 2014;14(3):271–278 16. Bast T, Oezkan O, Rona S, et al. EEG and MEG source analysis of single and averaged interictal spikes reveals intrinsic epileptogenicity in focal cortical dysplasia. Epilepsia 2004;45(6):621–631 17. Otsubo H, Iida K, Oishi M, et al. Neurophysiologic findings of neuronal migration disorders: intrinsic epileptogenicity of focal cortical dysplasia on electroencephalography, electrocorticography, and magnetoencephalography. J Child Neurol 2005;20(4):357–363 18. Kagawa K, Iida K, Kakita A, et al. Electrocorticographic-­ histopathologic correlations implying epileptogenicity of dysembryoplastic neuroepithelial tumor. Neurol Med Chir (Tokyo) 2013;53(10):676–687 19. Jung J, Bouet R, Delpuech C, et al. The value of magnetoencephalography for seizure-onset zone localization in magnetic resonance imaging-negative partial epilepsy. Brain 2013;136(Pt 10):3176–3186 20. Wilenius J, Medvedovsky M, Gaily E, et al. Interictal MEG reveals focal cortical dysplasias: special focus on patients with no visible MRI lesions. Epilepsy Res 2013;105(3):337–348 21. Widjaja E, Otsubo H, Raybaud C, et al. Characteristics of MEG and MRI between Taylor’s focal cortical dysplasia (type II) and other cortical dysplasia: surgical outcome after complete resection of MEG spike source and MR lesion in pediatric cortical dysplasia. Epilepsy Res 2008;82(2–3):147–155 22. Besson P, Andermann F, Dubeau F, Bernasconi A. Small focal cortical dysplasia lesions are located at the bottom of a deep sulcus. Brain 2008;131(Pt 12):3246–3255

18 Magnetoencephalography 23. Nakajima M, Widjaja E, Baba S, et al. Remote MEG dipoles in focal cortical dysplasia at bottom of sulcus. Epilepsia 2016;57(7):1169–1178

45. Ryvlin P, Minotti L, Demarquay G, et al. Nocturnal hypermotor seizures, suggesting frontal lobe epilepsy, can originate in the insula. Epilepsia 2006;47(4):755–765

24. Wu JY, Sutherling WW, Koh S, et al. Magnetic source imaging localizes epileptogenic zone in children with tuberous sclerosis complex. Neurology 2006;66(8):1270–1272

46. Mohamed IS, Gibbs SA, Robert M, Bouthillier A, Leroux JM, Khoa Nguyen D. The utility of magnetoencephalography in the presurgical evaluation of refractory insular epilepsy. Epilepsia 2013;54(11):1950–1959

25. Okanishi T, Akiyama T, Mayo E, et al. Magnetoencephalography spike sources interrelate the extensive epileptogenic zone of tuberous sclerosis complex. Epilepsy Res 2016;127:302–310 26. Iida K, Otsubo H, Mohamed IS, et al. Characterizing magnetoencephalographic spike sources in children with tuberous sclerosis complex. Epilepsia 2005;46(9):1510–1517 27. Xiao Z, Xiang J, Holowka S, et al. Volumetric localization of epileptic activities in tuberous sclerosis using synthetic aperture magnetometry. Pediatr Radiol 2006;36(1):16–21 28. Kamimura T, Tohyama J, Oishi M, et al. Magnetoencephalography in patients with tuberous sclerosis and localization-related epilepsy. Epilepsia 2006;47(6):991–997 29. Sugiyama I, Imai K, Yamaguchi Y, et al. Localization of epileptic foci in children with intractable epilepsy secondary to multiple cortical tubers by using synthetic aperture magnetometry kurtosis. J Neurosurg Pediatr 2009;4(6):515–522 30. Koo B, Hwang P. Localization of focal cortical lesions influences age of onset of infantile spasms. Epilepsia 1996;37(11):1068–1071 31. Oguni H, Hayashi K, Osawa M. Migration of epileptic foci in children. Adv Neurol 1999;81:131–143 32. Ibrahim GM, Fallah A, Albert GW, et al. Occipital lobe epilepsy in children: characterization, evaluation and surgical outcomes. Epilepsy Res 2012;99(3):335–345 33. Ossenblok P, de Munck JC, Colon A, Drolsbach W, Boon P. Magnetoencephalography is more successful for screening and localizing frontal lobe epilepsy than electroencephalography. Epilepsia 2007;48(11):2139–2149 34. Mu J, Rampp S, Carrette E, et al. Clinical relevance of source location in frontal lobe epilepsy and prediction of postoperative long-term outcome. Seizure 2014;23(7):553–559 35. Hara K, Lin FH, Camposano S, et al. Magnetoencephalographic mapping of interictal spike propagation: a technical and clinical report. AJNR Am J Neuroradiol 2007;28(8):1486–1488 36. Shiraishi H, Ahlfors SP, Stufflebeam SM, et al. Application of magnetoencephalography in epilepsy patients with widespread spike or slow-wave activity. Epilepsia 2005;46(8):1264–1272 37. Shibata S, Matsuhashi M, Kunieda T, et al. Magnetoencephalography with temporal spread imaging to visualize propagation of epileptic activity. Clin Neurophysiol 2017;128(5):734–743 38. Shirozu H, Iida K, Hashizume A, et al. Gradient magnetic-field topography reflecting cortical activities of neocortical epilepsy spikes. Epilepsy Res 2010;90(1–2):121–131 39. Ishitobi M, Nakasato N, Yamamoto K, Iinuma K. Opercular to interhemispheric source distribution of benign Rolandic spikes of childhood. Neuroimage 2005;25(2):417–423 40. Sobel DF, Aung M, Otsubo H, Smith MC. Magnetoencephalography in children with Landau-Kleffner syndrome and acquired epileptic aphasia. AJNR Am J Neuroradiol 2000;21(2):301–307 41. Ryvlin P, Kahane P. The hidden causes of surgery-resistant ­ temporal lobe epilepsy: extratemporal or temporal plus? Curr Opin Neurol 2005;18(2):125–127 42. Barba C, Barbati G, Minotti L, Hoffmann D, Kahane P. Ictal clinical and scalp-EEG findings differentiating temporal lobe epilepsies from temporal ‘plus’ epilepsies. Brain 2007;130(Pt 7):1957–1967 43. Weil AG, Fallah A, Lewis EC, Bhatia S. Medically resistant pediatric insular-opercular/perisylvian epilepsy. Part 1: invasive monitoring using the parasagittal transinsular apex depth electrode. J Neurosurg Pediatr 2016;18(5):511–522 44. Nguyen DK, Nguyen DB, Malak R, et al. Revisiting the role of the insula in refractory partial epilepsy. Epilepsia 2009;50(3):510–520

47. Engel JJ, Van Ness PC, Rasmussen TB, et al. Outcome with respect to epileptic seizures. In: Engel J Jr., ed. Surgical Treatment of the Epilepsies. New York, NY: Raven Press; 1993:367–373 48. Benifla M, Otsubo H, Ochi A, et al. Temporal lobe surgery for intractable epilepsy in children: an analysis of outcomes in 126 children. Neurosurgery 2006;59(6):1203–1213, discussion 1213–1214 49. Mikuni N, Nagamine T, Ikeda A, et al. Simultaneous recording of epileptiform discharges by MEG and subdural electrodes in temporal lobe epilepsy. Neuroimage 1997;5(4 Pt 1):298–306 50. Sato S, Balish M, Muratore R. Principles of magnetoencephalography. J Clin Neurophysiol 1991;8(2):144–156 51. Alarcon G, Guy CN, Binnie CD, Walker SR, Elwes RD, Polkey CE. Intracerebral propagation of interictal activity in partial epilepsy: implications for source localisation. J Neurol Neurosurg Psychiatry 1994;57(4):435–449 52. Imai K, Otsubo H, Sell E, et al. MEG source estimation from mesio-basal temporal areas in a child with a porencephalic cyst. Acta Neurol Scand 2007;116(4):263–267 53. RamachandranNair R, Ochi A, Benifla M, Rutka JT, Snead OC III, Otsubo H. Benign epileptiform discharges in Rolandic region with mesial temporal lobe epilepsy: MEG, scalp and intracranial EEG features. Acta Neurol Scand 2007;116(1):59–64 54. Paolicchi JM, Jayakar P, Dean P, et al. Predictors of outcome in pediatric epilepsy surgery. Neurology 2000;54(3):642–647 55. Semah F, Picot MC, Adam C, et al. Is the underlying cause of epilepsy a major prognostic factor for recurrence? Neurology 1998;51(5):1256–1262 56. RamachandranNair R, Otsubo H, Shroff MM, et al. MEG predicts outcome following surgery for intractable epilepsy in children with normal or nonfocal MRI findings. Epilepsia 2007;48(1):149–157 57. Widjaja E, Shammas A, Vali R, et al. FDG-PET and magnetoencephalography in presurgical workup of children with localizationrelated nonlesional epilepsy. Epilepsia 2013;54(4):691–699 58. Schneider F, Irene Wang Z, Alexopoulos AV, et al. Magnetic source imaging and ictal SPECT in MRI-negative neocortical epilepsies: additional value and comparison with intracranial EEG. Epilepsia 2013;54(2):359–369 59. Sutherling WW, Mamelak AN, Thyerlei D, et al. Influence of magnetic source imaging for planning intracranial EEG in epilepsy. Neurology 2008;71(13):990–996 60. Knowlton RC, Razdan SN, Limdi N, et al. Effect of epilepsy magnetic source imaging on intracranial electrode placement. Ann Neurol 2009;65(6):716–723 61. Ahmed R, Rutka JT. The role of MEG in pre-surgical evaluation of epilepsy: current use and future directions. Expert Rev Neurother 2016;16(7):795–801 62. Murakami H, Wang ZI, Marashly A, et al. Correlating magnetoencephalography to stereo-electroencephalography in patients undergoing epilepsy surgery. Brain 2016;139(11):2935–2947 63. Iida K, Otsubo H. Stereoelectroencephalography: indication and efficacy. Neurol Med Chir (Tokyo) 2017;57(8):375–385 64. Téllez-Zenteno JF, Dhar R, Wiebe S. Long-term seizure outcomes following epilepsy surgery: a systematic review and meta-­ analysis. Brain 2005;128(Pt 5):1188–1198 65. Mohamed IS, Otsubo H, Ochi A, et al. Utility of magnetoencephalography in the evaluation of recurrent seizures after epilepsy surgery. Epilepsia 2007;48(11):2150–2159

159

160

IIb  Preoperative Electrophysiological Assessment 66. Snead OC III. Surgical treatment of medically refractory epilepsy in childhood. Brain Dev 2001;23(4):199–207 67. Kawamura T, Nakasato N, Seki K, et al. Neuromagnetic evidence of pre- and post-central cortical sources of somatosensory evoked responses. Electroencephalogr Clin Neurophysiol 1996;100(1):44–50 68. Pihko E, Lauronen L, Wikström H, et al. Somatosensory evoked potentials and magnetic fields elicited by tactile stimulation of the hand during active and quiet sleep in newborns. Clin Neurophysiol 2004;115(2):448–455 69. Sharma R, Pang EW, Mohamed I, et al. Magnetoencephalography in children: routine clinical protocol for intractable ­epilepsy at the Hospital for Sick Children. In: Cheyne D, Ross B, Stroink G, Weinberg H, eds. New Frontiers in Biomagnetism. International Congress Series 2007;1300. Amsterdam: Elsevier; 2007:685–688 70. Bercovici E, Pang EW, Sharma R, et al. Somatosensory-evoked fields on magnetoencephalography for epilepsy infants younger than 4 years with total intravenous anesthesia. Clin Neurophysiol 2008;119(6):1328–1334 71. Burneo JG, Kuzniecky RI, Bebin M, Knowlton RC. Cortical reorganization in malformations of cortical development: a magnetoencephalographic study. Neurology 2004;63(10):1818–1824 72. Gaetz W, Cheyne D. Localization of sensorimotor cortical rhythms induced by tactile stimulation using spatially filtered MEG. Neuroimage 2006;30(3):899–908

73. Nakasato N, Kumabe T, Kanno A, Ohtomo S, Mizoi K, Yoshimoto T. Neuromagnetic evaluation of cortical auditory function in patients with temporal lobe tumors. J Neurosurg 1997;86(4):610–618 74. Pang EW, Gaetz W, Otsubo H, Chuang S, Cheyne D. Localization of auditory N1 in children using MEG: source modeling issues. Int J Psychophysiol 2003;51(1):27–35 75. Nakasato N, Yoshimoto T. Somatosensory, auditory, and visual evoked magnetic fields in patients with brain diseases. J Clin Neurophysiol 2000;17(2):201–211 76. Pataraia E, Simos PG, Castillo EM, et al. Reorganization of languagespecific cortex in patients with lesions or mesial temporal epilepsy. Neurology 2004;63(10):1825–1832 77. Breier JI, Castillo EM, Simos PG, et al. Atypical language representation in patients with chronic seizure disorder and achievement deficits with magnetoencephalography. Epilepsia 2005;46(4):540–548 78. Lee D, Sawrie SM, Simos PG, Killen J, Knowlton RC. Reliability of language mapping with magnetic source imaging in epilepsy surgery candidates. Epilepsy Behav 2006;8(4):742–749 79. Papanicolaou AC, Simos PG, Castillo EM, et al. Magnetocephalography: a noninvasive alternative to the Wada procedure. J Neurosurg 2004;100(5):867–876 80. Papanicolaou AC, Simos PG, Breier JI, et al. Magnetoencephalographic mapping of the language-specific cortex. J Neurosurg 1999;90(1):85–93

19

  Structural Brain Imaging in Pediatric Epilepsy Charles Raybaud and Elysa Widjaja

Summary Neuroimaging, specifically MRI (oriented by the clinical, functional, and electroencephalographic [EEG]/magnetoencephalographic [MEG] data), plays a central role in the assessment of epilepsy, as it is the most efficient way to demonstrate a lesion that is causally associated with the disease and to make possible a surgical treatment of it. Such a lesion may be obvious, but it may commonly be quite subtle, and for that reason, the best possible quality of imaging is mandatory: multiplanar, multisequence high-definition imaging, as much as possible with a “high field” magnet (3+ teslas) and a multiple phased-­array coil. Diffusion tensor imaging (DTI) also is needed for the surgical planning. The best indications for epilepsy surgery are the lesions that are typically associated with severe, drug-­resistant epilepsy. A common example in children are the low-grade epilepsy-associated tumors (LEATs; ganglioglioma and variants, dysembryoplastic neuroepithelial tumor [DNET], and pleomorphic xanthoastrocytoma [PAX] mostly, but also more rarely, angiocentric glioma, papillary glioneuronal tumor (PGNT), as well as any glial tumor involving the cortex). Focal cortical dysplasias (FCDs) are epileptogenic developmental lesions that may be very subtle. They are classified histologically in three types: most typical, FCD 2 (Taylor’s) is defined by the presence of monstrous cells; FCD 1, more difficult to identify radiologically, is defined by an abnormal cortical layering; and FCD 3 are defined by their association with a principal lesion (hippocampal sclerosis, LEATs, vascular malformation, destructive lesions). Other surgical indications include the hemimegalencephalies (HMEs; hemispheric FCD) and sometimes ­tuberous sclerosis TSC (syndromic FCD), as well as the hypothalamic hamartomas, typically sessile and attached to the mammillary bodies. Cortex-associated cavernomas are obvious indications. Palliative surgery may be needed in Sturge– Weber’s or Rasmussen’s diseases, as well as in some cases of chronic encephalomalacic lesions. Obviously, like in adults but much less commonly, mesial temporal sclerosis is an important cause of refractory epilepsy in children, and therefore a surgical indication. In exceptional cases, major malformations of cortical development (MCD) such as a heterotopia, a polymicrogyria (PMG), or a schizencephaly may also benefit from surgery and can be diagnosed by MRI only. Keywords:  epilepsy, imaging, low-grade epilepsy associated tumors (LEATs), focal cortical dysplasia (FCD), ­hemimegalencephaly (HME), tuberous sclerosis (TSC), ­hypothalamic h ­ amartoma, Sturge–Weber, Rasmussen’s, mesial temporal sclerosis

„„ Introduction From an imaging point of view, a distinction should be made between a first seizure (which may reflect any acute brain disease) and a true, chronic epilepsy; but obviously a new seizure may mark the onset of a chronic epilepsy. Even in the event of a first seizure, a distinction should be made in children, between a febrile seizure, for which the diagnostic yield of imaging is very low,1 and an a-febrile seizure, in which the diagnostic yield becomes significant.2 Any epilepsy that is not a benign, idiopathic form of epilepsy (which can be ­recognized on clinical and ­electrographic features) requires neuroimaging. This includes lesional focal epilepsy, specific epileptic syndromes pointing to structural brain abnormalities such as Ohtahara or West syndromes, new status epilepticus, and the so-called catastrophic epilepsies that are associated with a progressive deterioration. The main imaging modality for detecting the pathologic substrate responsible for epilepsy is MRI; a normal-appearing CT does not rule out a subtle lesion like an FCD; even if the CT is abnormal, it lacks the precision and the specificity of magnetic resonance (MR), and the use of ionizing r­ adiation in developing children is not recommended. In children with newly diagnosed epilepsy, MR detected an ­abnormality in 62/388 (13%).3 In those with intractable ­epilepsy, MR detected an abnormality in 82 to 86% of cases.4,​5 As compared with CT, MRI has excellent soft tissue contrast, better spatial resolution, multiplanar capability, and higher sensitivity, and it is therefore the imaging modality of choice. Even in patients with “normal” images, a-posteriori review of the images, or repeat imaging, especially with advances in techniques, can help previously undetected structural abnormalities be identified.6,​7 It is crucial to correlate MR-identified ­substrate with clinical and electrophysiologic data (EEG, MEG) to avoid false localization. MRI also has prognostic implications: failure to detect a lesion on MRI leads to a worse surgical outcome compared to when a lesion is identified.

„„ Magnetic Resonance Techniques The sensitivity of MR for detecting abnormalities depends on the MR techniques used, the pathologic substrate, and the experience of the interpreting physician.8 An optimal MR technique for assessing the pathological substrate should include a variety of imaging sequences, including T1- and T2-weighted

162

IIc  Preoperative Neuroimaging imaging, proton density, and FLAIR sequences. These need to be acquired in at least two orthogonal planes covering the whole brain, using the minimum slice thickness. In those with temporal lobe epilepsy, the coronal plane should be perpendicular to the long axis of the hippocampus to optimize visualization of the mesial temporal structures. For extratemporal epilepsy, the bicommissural plane (parallel to the anterior–posterior commissure line) is the standard method used. A three-dimensional T1 volume sequence with slice thickness of 1 mm provides excellent gray/white matter contrast, can be reformatted into any orthogonal or nonorthogonal planes and may be subjected to additional post-processing without the penalty of additional imaging time. Double inversion recovery sequences have been recommended as well.9,​10 Gadolinium does not improve the sensitivity of MR in patients with epilepsy and should only be used to characterize selective intracerebral lesions such as vascular malformations or tumors. A systematic approach should be used in order to optimize detection of subtle lesions and double pathologies. It should be remembered that in young infants, the cortical–subcortical MR-contrast is reversed as compared to the mature brain because of the lack of myelination. This means that during the maturation process, there is a period of months during which this contrast is attenuated, and the diagnosis of subtle cortical dysplasia becomes difficult. On the other hand, repeated focal seizures induce an early myelination which can be used to identify the location of the epileptogenic zone in some cases. MR spectroscopy may be used to better characterize a lesion (mostly between tumors, or tumor versus dysplasia according to the spectral profile), or, by showing a decreased N-acetylaspartate (NAA), to locate the epileptogenic area. However, the changes may reflect the structural abnormalities as well as ipsi- and contralateral metabolic alterations related to the seizures, making the interpretation of the results difficult.11,​12 Diffusion imaging may show striking abnormalities during status epilepticus, reflecting cytotoxic edema both locally and remotely in the ipsilateral pulvinar, ipsilateral hippocampus, contralateral cerebellum (diaschisis-like response) as well as in the corpus callosum.13 This brain response to repeated/ prolonged seizures is due to a mismatch between the energy consumption and supply; it may be useful to locate a focus. In addition to conventional MR, DTI better depicts the structure of the white matter. Although it is not currently used for diagnosis, it may help in understanding the abnormalities of white matter associated with cortical malformations.13,​14 DTI tractography also helps in preparing the surgical approach to a lesion.13,​15 Perfusion imaging in the setting of epilepsy is still in an experimental stage. Structural image analyses describe various computer-assisted methods devised to improve the rate of detection of subtle brain abnormalities in epilepsy. Most use segmentation methods and post-processing algorithms aimed at quantifying the ­volume of various compartments of the brain such as gray and white matter or specific lobes or structures such as the hippocampus, or to evaluate the thickness of the cortex or blurring of the cortical–subcortical junction. Unfortunately, these methods are not fully automated and are time consuming and difficult to use in clinical practice. Multimodal integration of the anatomic data from MRI, functional data from interictal PET or ictal/ interictal SPECT, and electrophysiological data from ­ cortical

or stereo-EEG, or from MEG may also help in ­ identifying the ­ epileptogenic focus with its underlying s­ tructural ­abnormalities.16 The presurgical anatomic assessment includes functional MRI and tractography mostly. Functional MRI aims at identifying the location of the main cortical functions, which in children with chronic structural and functional abnormalities may not match the classic anatomic distribution. Tractography provides the location of the principal white matter bundles. Studies of the functional connectivity are useful to understand how epilepsy might impact the brain function but are not yet part of the presurgical assessment.17

„„ Epileptogenic Substrates While the range of pathologic substrates responsible for intractable partial epilepsy in children is similar to those of adults, MCD and developmental tumors are more commonly detected in surgical specimens of pediatric epilepsy patients. In contrast, hippocampal sclerosis is less common in pediatric patients compared to adults.

Low-Grade Epilepsy-Associated Tumors LEATs may account for up to two thirds of the surgical pathologic substrate.18 They originate in and develop from the cortex and therefore clinically present with seizures. They are slow growing tumors with well-defined margin, typically without associated edema or necrosis. Complete removal results in good seizure control or renders the patient seizure free. These tumors often are associated with adjacent FCD, which is now assumed to be secondary to the epileptogenic activity of the tumor,19 although it has also been hypothesized that the tumor could derive from the same precursor cells from the dysplastic tissue.20,​21 The epilepsy-associated tumors include ­ganglioglioma and its variants (gangliocytoma, extraventricular neurocytoma, desmoplastic infantile ganglioglioma), DNET, PXA, and low-grade astrocytoma or oligodendroglioma. More glioneuronal tumors have been described (angiocentric glioma, PGNT) but are uncommon.22 Gangliogliomas are slightly more common in males and much more common in children than in adults. They are macroscopically observed to be up to eight times larger in children than in adults.23 They are associated with chronic epilepsy in 85% of cases, and are located mostly in the temporomesial (50%) or temporolateral (29%) location.24 Histologically, they are comprised of two cellular populations, one neuronal and one glial. The neuronal component does not expand; the glial component rarely can be malignant. Gangliogliomas may present as a solid mass in 43%, a cyst in 5%, and a mixed lesion in 52%. They involve the cortex, usually broaden the gyri and may cause remodeling of the adjacent bone. They usually form discrete, well-demarcated masses in the neocortex, but they are more infiltrative mass in the mesial temporal region.25 On MR, the tumor may appear hypo- or isointense to gray matter on T1-weighted imaging and hyper- or isointense on T2-weighted imaging/FLAIR. No diffusion restriction. Some may demonstrate intrinsic high T1 signal, due to calcification that may be seen in about 30 to 50% of cases. Enhancement following gadolinium administration is common, seen in up to 60% of cases. It can be nodular, ring-like, or solid. Leptomeningeal involvement may be seen. Patients with

19  Structural Brain Imaging in Pediatric Epilepsy ganglioglioma had the best outcome compared to other types of LEAT (92% Engel class I and II).18,​24 Gangliocytoma is uncommon, affecting adolescent and young adults. It involves the cortex and consists of neurons without glial tissue. It usually has solid and cystic components. On MRI, gangliocytoma demonstrates low T1 and high T2/FLAIR signals and enhances on the postgadolinium scans. Desmoplastic infantile ganglioglioma is a rare, likely congenital tumor that develops in infants. Usually huge, more often suprasylvian in location, the tumor is partly cystic and partly solid. The solid portion incorporates the cortex and is diffusely attached to the dura; it is strongly desmoplastic and may be calcified.26 The cystic portion extends into the white matter. In about half the cases, the infant presents with macrocephaly, neurological deficits, and seizures. On MRI the solid portion is isointense to gray matter on T1-weighted imaging, hyper-, iso-, or hypointense on T2-weighted imaging, usually heterogeneously. Contrast enhancement of the solid portion is intense and extends to the dura. The wall of the cystic component does not enhance.26 In spite of the spectacular appearance of the tumor, the prognosis is good.26 DNET affects the temporal and frontal lobes predominantly, and in children is less common (14%) than ganglioglioma (43%).18 It may have a well-demarcated margin (50%) or the margins may be slightly blurred; it may be multinodular; it may cause broadening of the gyri, effacement of the sulci and distortion of the ventricles27 and result in remodeling of the overlying skull vault in 44% of cases.28 Calcification has been reported.18,​28 On MRI, the tumor is of low T1/high T2 signal, often with multicystic, multinodular changes and “bubbly” appearance and it may appear dark with a thin rim of high signal on FLAIR sequence.27 It appears wedge-shaped and extends to the ventricle in 30% of cases. One third of cases show faint punctate or ring enhancement. Spontaneous hemorrhage has been reported. The tumor usually is stable over the years but a significant increase in size has been documented in a few cases.29 DNET has less favorable outcome compared to ganglioglioma, which may be attributed to incomplete resection of the tumor or to the presence of the associated cortical dysplasia, which is not visible on MRI. The differential diagnoses of DNET include other LEATs and FCD. FCD does not enhance and usually does not have mass effect; the cortex may be blurred but is never completely effaced as it is in the case of a tumor. Gangliogliomas usually have more mass effect. MR spectroscopy is reported to be normal in DNET,30 while high choline and low NAA are observed in gliomas and gangliogliomas. Another tumor described as a nonspecific DNET or a cortical oligodendroglioma (WHO grade 2) is an intracortical hemispheric tumor that presents with isolated epilepsy without neurological deficit or increased intracranial pressure. It has a DNET-like appearance on imaging: triangular cortical lesion with septa, low T1/high T2 signals, without surrounding edema or mass effect and no enhancement. In contrast to the deep oligodendroglioma, this peripheral epilepsy-associated tumor has a good prognosis.31 PXA is a rare tumor affecting children and young adults. Like the other LEATs, PXA is slow growing, located in the cortex, and highly epileptogenic. It is located in the cortex and extends across the parenchyma, supratentorial in 98% of cases, mostly temporal (49%). Both cystic and solid it may demonstrate

c­ontinuity with the dura. Calcification is rare; the tumor is well circumscribed without peritumoral edema.32 Hemorrhage has been reported. It contains predominantly glial component, but may also contain neuronal elements.33 On MR, the tumor is hypo- to isointense to gray matter on T1-weighted imaging and hyper- to isointense on T2-weighted imaging/FLAIR with no diffusion restriction, and the cystic portion is isointense to cerebrospinal fluid. Post gadolinium, the nodular solid component as well as the adjacent meninges enhances (dural “tail”). Like DNET, PXA is associated with cortical dysplasia.34 The prognosis is generally good, but it may recur and become malignant. Angiocentric glioma form a homogeneous but ill demarcated cortical-based mass with infiltration of the gyral white matter, sometimes with a transcerebral extension (“stalk”) toward the ventricular wall. The gyrus is bulky with effacement of the adjacent sulci but without edema or significant mass effect. On MR the cortical component is described as bright on T1, and bright on T2/FLAIR, the white matter component as dark on T1 and bright on T2/FLAIR, with no diffusion restriction and no contrast enhancement.25 The surrounding FCD is not apparent on MR. PGNT (WHO grade 1) typically presents with a history of recent seizures, or incidentally. It is juxtaventricular more often than peripherally located, usually frontal. It is well demarcated, solid and cystic/necrotic, sometimes cystic with a mural nodule, hypointense on T1, hyperintense on T2 and FLAIR, with no restricted diffusion. Both the solid portion and the cyst wall enhance. Calcification is possible. Edema and mass effect, if any, are mild.25 Diffuse, low-grade (fibrillary) astrocytoma (DA, grade 2) is located predominantly in the frontal or temporal lobes and invades the cortex. It does not contain ganglionic cells. MR demonstrates a homogenous T1-hypointense and T2/FLAIR-­ hyperintense mass with no restricted diffusion that expands and infiltrates the adjacent cortex; it does not enhance but may show prominent traversing vessels. Calcification, cysts, ­hemorrhage, or surrounding edema are uncommon. In contradistinction with the histologically similar (but g ­ enetically different) tumor in the adults, DA in children almost never demonstrates malignant progression.

Common Malformations of Cortical Development Of the different subtypes of MCD, the lesions that more commonly undergo surgical management of the epilepsy are FCD, HME, and TSC. FCD is intrinsically epileptogenic and is a frequent cause of epilepsy in children.35,​36 The mechanism of the epilepsy is still unclear: abnormal firing from the dysplastic neural cells, dysfunction of synaptic circuits with abnormal synchronization of the neuronal population, or abnormal organization of the inhibitory interneurons. Originally coined by Taylor et al37 to describe a specific abnormality made of a disorganized cortical architecture with “bizarre neurons” and giant dysmorphic “balloon cells,” the term FCD has become used extensively in the literature to refer to a wide range of derangements of the cortex, so that various classifications have been proposed. The most recent ILAE consensus classification distinguishes three groups of dysplasia:19

163

164

IIc  Preoperative Neuroimaging yy FCD type 1: °° FCD 1a: retention of the cortical microcolumnar radial organization. °° FCD 1b: absence of the six-layered tangential organization. °° FCD 1c: associated changes of FCD 1a and 1b. yy FCD type 2 (Taylor’s): °° FCD 2a: disruption of cortical layering with dysmorphic neurons. °° FCD 2b: disruption of cortical layering with dysmorphic neurons and balloon cells. yy FCD type 3 (associated with a principal lesion): °° FCD 3a: temporal cortical lamination abnormality associated with hippocampal sclerosis. °° FCD 3b: cortical lamination abnormality adjacent to a glial or glioneuronal tumor. °° FCD 3c: cortical lamination abnormality adjacent to a vascular malformation. °° FCD 3d: cortical lamination abnormality adjacent to any lesion acquired during the early life. In all, there is a disorder of the cortical architecture which may be isolated (FCD type 1), or associated with cellular monstrosities (FCD 2, Taylor’s FCD), or adjacent to a specific principal lesion (FCD type 3). FCDs 1 and 2 are different lesions thought to relate to the stage at which abnormal development occurred.38 FCD type 2 are felt to result from an early disturbance of the glial–neuronal differentiation.39 FCD 2a is characterized by the presence of dysmorphic neurons, FCD 2b by the presence of dysmorphic neurons and balloon-cells. The dysmorphic neurons may either be pyramidal or interneuronal, but the balloon cells are of glial origin rather. FCD 2 is more likely to demonstrate abnormal signal and heterotopic neurons in the white matter, extending from the cortex to the ventricles (transmantle dysplasia);40 as this reflects the migration path from the periventricular zone to the cortex, it supports the suggestion that they occur early during the development.40,​41 In addition, the presence of numerous neurons in layer 1 of the cortex and the white matter, and their abnormal morphology,42 suggests an excessive late neurogenesis with possible retention of radial glia and subplate neurons, which would indicate that the process may still go on in the late stage of cortical development. In any case, the abnormal or misplaced cells may not be correctly connected, and the gyration, which depends on the connectivity, may therefore be abnormal.43 On MRI, the main diagnostic features are seen in 90% of FCD 2b, but only 50% of FCD 2a.44 The cortex typically is focally thick, usually bright on T2 and FLAIR, but also, sometimes, on T1.45 The blurring of the cortical–subcortical junction is more typical for FCD 2b. The subcortical white matter may be dark on T1 (similar to the cortex), and is often bright on T2 and FLAIR, less however than the cortex which can always be recognized: this permits the differentiation of FCD from LEATs (Fig. 19.1). The white matter signal changes may taper from the cortex to the ventricle, forming a transmantle dysplasia that is nearly specific of FCD 2b (Fig. 19.2).40 The blurring of the cortical–­ subcortical junction is more common in FCD 2b (Fig. 19.3). Gyration abnormalities are present but usually discreet. A particular form of FCD 2b is the bottom of sulcus dysplasia with an often-faint focal cortical signal abnormality in the depth of an ­exceedingly deep sulcus (Fig. 19.4).45,​46 FCD type 1 is often an extensive lesion associated with a poor lobar development and, clinically, with a developmental

Fig. 19.1  Right occipital focal cortical dysplasia (FCD) 2, T2-weighted imaging. The lesion is characterized by the bright signal of the cortex and still more, of the white matter, with the cortex being recognizable.

Fig. 19.2  Left frontal focal cortical dysplasia (FCD) 2b, FLAIR. The characteristic feature is the transmantle dysplasia: the signal abnormality of the white matter tapers from the dysplastic cortex to the ventricular wall. The cortex is thick and the cortical–subcortical junction is blurred.

delay. In contradistinction with FCD 2 it is classified as a late ­ alformation.38 FCD 1a is characterized by the retention of the m early radial columnar pattern, FCD 1b by the lack of the six-­ layer pattern, and FCD 3c by a combination of both. Radial cellular columns appear early in the cortex and reflect the radial glia-­guided migration of the neurons. The tangential six-layer pattern develops secondarily due to the developing intracortical

19  Structural Brain Imaging in Pediatric Epilepsy

Fig. 19.3  Left precentral focal cortical dysplasia (FCD) 2b, T1-weighted imaging. The cortical–subcortical limit of the dysplastic cortex appears blurred as compared with the surrounding normal cortex.

­connectivity.47 It may be absent if the connectivity fails to develop, or be altered if the cortex is injured later in development. The ­dysplastic pattern may be acquired postnatally as the disruption of the connectivity would favor the development of giant, ill oriented neurons with an abnormal circuitry. Such findings have been observed in infants following severe perinatal or early postnatal injuries.48—​​52 On MRI, the diagnosis of FCD 1 is difficult and probably often missed. The main features would be a poor development of a brain lobe and an attenuation of the normal contrast of the subjacent white matter, assumed to reflect a gliosis and demyelination secondary to the repeated seizure activity, rather than to represent the dysplasia itself. Minor MCD (formerly known as microdysgenesis) are, like the FCD type 1, considered to be due to late developmental ­disturbances.38 Morphologically they are poorly defined, characterized mostly by the presence of heterotopic neurons. Radiologically they are, like FCD 1, either missed or only suspected because of a lobar atrophy and faint white matter changes (Fig. 19.5). This group however also includes “variants” such as the cortical perivascular satellitosis,53 the hyaline astrocytic inclusions,54 or the oligodendroglial hyperplasia55 that may present with more striking, if not specific, MR signal changes. FCD type 3 is (by definition) associated with a principal lesion, and is always a FCD 1: if it is a FCD 2, this is considered a double lesion.19 FCD 3a is associated with a hippocampal sclerosis; the dysplasia of the temporal cortex may include any combination of temporal sclerosis, neuronal heterotopia and small lentiform heterotopia (not seen on MRI). It is not known whether, of hippocampal sclerosis and dysplasia, one causes the other or whether they develop together.56 FCD 3b is

Fig. 19.4  Left frontal bottom of sulcus focal cortical dysplasia (FCD) 2, T1-weighted imaging. The superior frontal sulcus (anterior portion, parasagittal) is deeper on the left than on the right. The FCD at the bottom of the sulcus appears brighter on T1 than the adjacent cortex.

Fig. 19.5  Right temporal focal cortical dysplasia (FCD) 1/minor malformations of cortical development (MCD), fluid attenuation inversion recovery (FLAIR). The normally dark signal of the right temporal white matter is lost as compared with the left. The temporal lobe, and to some degree, the right hemispheres are also smaller on the right than on the left.

associated with the epileptogenic cortical tumors (e.g., LEATs); the dysplasia may result from the epileptogenic activity of the tumor19 but alternatively the tumor might originate from the dysplastic cortex.20 FCD 3c is associated with, and likely secondary to, a v ­ ascular malformation. FCD 3d is secondary to an

165

166

IIc  Preoperative Neuroimaging ­encephaloclastic ­scarring process.48,​49,​50,​51,​52 On MRI, FCD 3a may be identified when hippocampal sclerosis is associated with a small temporal lobe with faint white matter signal change. In FCD 3b, 3c, and 3d, the diagnosis is assumed from the presence of the principal lesion. In the specific instance when the principal lesion is a meningioangiomatosis, the vascular lesion is usually not seen but the cortical dysplastic pattern may be quite characteristic with a T2-dark, thick cortex overlying an unmyelinated segment of white matter (Fig. 19.6); contrast administration may occasionally reveal some pial enhancement. The MR appearance of FCDs may change with brain maturation. In longitudinal MR studies the early study may be normal but repeat imaging at a later date may demonstrate high T2 signal in the white matter, high T1 signal in the cortex and blurring of the cortical/subcortical white matter junction.57 Also, in contrast to the high T2/FLAIR signal seen in mature children, the white matter adjacent to the dysplastic cortex may demonstrate low T2/high T1 signal in neonates and infants: this is postulated to be secondary to an early myelination induced by the repeated seizures. Experimental study in the mouse has shown that repeated neuronal electrical activity such as occurs in seizures induces myelination.58 HME may be sporadic or associated with neurocutaneous syndromes (TSC; epidermal nevus syndrome, linear nevus sebaceous syndrome, hypomelanosis of Ito, Proteus syndrome, Klippel–Trenaunay syndrome, and encephalo-cranio-cutaneous lipomatosis). The sporadic form is often considered a hemispheric variant of FCD; however, it is a heterogeneous entity and it may preferably be referred to as “hemispheric MCD.”59 Histologically, it presents an abnormal gyration, abnormal connectivity, dyslamination, poorly defined gray– white matter junction, dysmorphic neurons in both gray and white matter and balloon cells in 50% of cases. Clinically, it may present early with intractable epilepsy, hemiparesis, hemianopia, and mental retardation. On imaging, one hemisphere is large with an expanded calvarium, and in many but not all, an enlarged lateral ventricle of the affected side.60 The cortex is usually thick with broad gyri, the white

Fig. 19.6  Meningioangiomatosis with focal cortical dysplasia (FCD) 3c, T2-weighted imaging. The medial cortex on the right side is thick and dark (diffuse microcalcification), while the underlying white matter is bright.

matter has low T1, high T2/FLAIR signals with possible cystic changes and calcification. In infants, the white matter of the affected hemisphere may show high T1/low T2 signal, suggesting early myelination,61 likely due to seizure activity. HME may involve the cerebellum62 or may be limited to part of the hemisphere.63 With recurrent seizures or refractory status epilepticus, the enlarged hemisphere may secondarily become atrophic. Due to the intractable, poorly controlled seizures and progressive deterioration, functional or anatomical hemispherectomy may be required to control the seizures.59 Brain abnormalities in TSC may be considered a syndromic variant of FCD, with similar defects along the genetic–metabolic pathways. The “giant astrocytes” in cortical tubers correspond to the “balloon cells” present in Taylor’s FCD 2b. On imaging, TSC is characterized by cortical/subcortical tubers with broad gyri, thick cortex, abnormal signals in the cortex and subcortical white matter, common transmantle dysplasia, and occasionally calcification and cystic changes. The subependymal nodules are more commonly calcified, may enhance, are typically about the foramina of Monro, and may grow to form the giant cell astrocytomas (SEGA). The cerebellar cortex can be dysplastic as well. Surgery is occasionally possible if electrophysiology and PET or SPECT imaging demonstrate a single epileptogenic focus.

Other Malformations of Cortical Development Besides FCD, other MCD include disorders of cellular proliferation such as microcephaly, for which epilepsy, when it occurs, is only part of a globally severe condition; migration disorders such as the nodular and band heterotopia; organization disorders such as the PMG and the schizencephaly. Surgery is not commonly indicated in these conditions. Gray matter heterotopia is masses of apparently normal gray matter located in abnormal places. Their epileptogenicity is assumed to result from the abnormal connections they develop. The overlying cortex also is somewhat dysplastic, typically in proportion to the size of the heterotopia. On MRI, heterotopia has a similar signal as the central or cortical gray matter. Nodular heterotopia are designated as periventricular (isolated, multiple, or diffuse; never on basal ganglia or thalamus or corpus callosum) and subcortical (often huge, transcerebral, mixed with white matter). Band heterotopia is usually subcortical and corresponds to the lesser end of the agyria/pachygyria spectrum.38 Most cases of heterotopia are not indications for epilepsy surgery. PMG and schizencephaly are both classified as disorders of late organization of the cortex38 and may be sporadic, familial or acquired (often associated with CMV). In about 50% of cases, the patients present with neurological deficits and epilepsy. PMG is characterized by densely folded cortical layers below a smooth continuous molecular layer. Usually centered about the sylvian fissure, the malformation may extend variably over the hemispheric convexities, with a disorganized sulcal pattern; it may be uni- or bilateral, usually not symmetrical. The corresponding white matter and the brainstem are atrophic. Areas of high T2 signal in the white matter would suggest previous CMV infection. Typically, the abnormal cortex is still functional, and the surrounding normal-appearing cortex is epileptogenic, probably because of an abnormal connectivity. PMG of genetic origin are more often bilateral and symmetric. In general, PMG is a poor

19  Structural Brain Imaging in Pediatric Epilepsy

Fig. 19.7  Sessile posterior hypothalamic hamartoma, T2-weighted imaging. In this patient with severe gelastic seizures, a small hypothalamic mass is seen attached to the ventricular side of the mammillary body; the T2 signal is slightly brighter than that of the hemispheric cortex.

indication for epilepsy surgery, except for hemispherectomy, but in rare cases, it may be focal, and can be resected.64 Schizencephaly is characterized by a transcerebral cleft joining the ventricle and lined with polymicrogyric cortex. The cleft may be small or large, uni- or bilateral, not symmetrical. The septum pellucidum is usually absent. Surgical indications for schizencephaly-associated epilepsy are uncommon but there are reports of successful surgery of closed lip schizencephaly.65,​66,​67 The epileptogenic cortex surrounding the cleft was described as dysplastic with giant neurons.66

Hypothalamic Hamartoma Hypothalamic hamartomas are abnormal masses of normalappearing gray matter located in, or attached to the tuber cinereum. A distinction is made between the more anterior parahypothalamic pedunculated hamartomas clinically associated with a central precocious puberty; and the more posterior intrahypothalamic (sessile) hamartomas connected to the mammillary bodies, clinically associated with an early onset severe, typically gelastic epilepsy (which over time results in cognitive deterioration and behavioral problems).68,​69 The signal is similar to that of the normal gray matter, becoming slightly different with time (lower T1, brighter T2 and FLAIR) due to developing seizure-related gliosis (Fig. 19.7). There is no enhancement or calcification; a hypothalamic hamartoma may uncommonly be cystic. Uni- or bilateral, intra- or extraventricular, symmetric or not, the mass is of variable size, and grows in proportion to the brain only. It may extend anteriorly to the pituitary stalk, and posteriorly may splay the cerebral peduncles apart and displace the basilar artery. Epilepsy-associated hypothalamic hamartoma can be treated with surgical disconnection or radiosurgery.

Fig. 19.8  Hippocampal sclerosis, T2-weighted imaging. Classic appearance of the hippocampal sclerosis: bright T2 signal of the left hippocampal head, loss of the internal structure.

Hippocampal Sclerosis While hippocampal sclerosis is the most common epileptogenic substrate seen in adult surgical epilepsy series, it is less common in children. Typical hippocampal sclerosis is characterized by a neuronal loss and gliosis predominantly (but not only) in CA1 and CA4. This corresponds to hippocampal sclerosis ILAE type 1 of the recent ILAE consensus classification of hippocampal sclerosis, which is more often associated with an initial precipitating injury in early childhood, an early seizure onset and a favorable postsurgical seizure control.56 MR findings include high T2/FLAIR signal and atrophy of the hippocampus, loss of internal architecture patterning and of hippocampal head digitations (Fig. 19.8), and possibly atrophy of ipsilateral mammillary body and fornix. It may be associated with volume loss and white matter pallor of the temporal lobe; this reflects gliosis and demyelination secondary to seizures, and may, or not, be associated with a cortical dysplasia (classified as FCD 3a).56 With modern equipment, the sensitivity of MR in detecting hippocampal sclerosis by qualitative assessment is very high. Hippocampal volume reduction correlates with the severity of the neuronal cell loss.

Rasmussen’s Encephalitis Rasmussen’s encephalitis is a chronic progressive encephalitis of still unknown etiology.70 Anatomically, it corresponds to a sequential, multifocal unilateral inflammatory destructive process of the cortex of one hemisphere, with a significant loss of neurons and astrocytes. Clinically, the seizures begin abruptly in a previously normal child and include partial seizures and epilepsia partialis continua. With disease progression, the patients in months develop hemiparesis or hemiplegia and marked cognitive decline. Histologic findings are those of chronic nonspecific encephalitis with perivascular lymphocytic cuffing, gliosis, activated microglial nodules, and neuronal loss, resulting at a later stage in a nonspecific atrophy, gliosis, and minimal inflammatory

167

168

IIc  Preoperative Neuroimaging

Fig. 19.9  Rasmussen’s encephalitis, T2-weighted imaging. (a) Mild atrophy of the right frontal operculum with cortical thinning. Three years later (b), prominent parenchymal volume loss with extension of the cortical necrosis. The lateral ventricle is larger than before on the right side but also on the left side.

cellular infiltrates. Early in the course of the disease, MRI may be entirely normal or show only swelling of the cortex. With disease progression, it demonstrates progressive abnormal signal in the white matter and cortex and developing atrophy (Fig. 19.9).  Serial MRI of 10 patients revealed a step-like evolution, with different focal areas becoming involved successively.71 Frontal and frontotemporal lobe involvement is common, possibly ­associated with ipsilateral atrophy of striatum and hippocampus. Proton MR spectroscopy reveals decreased NAA concentration which correlate with the neuronal loss. Medical treatments are not really efficient, and surgical hemispherectomy may be the best option to stop the progression of the disease.

Other Causes for Partial or Catastrophic Epilepsies in Children The intellectual and neurological developments of children with Sturge–Weber disease depend on the occurrence of repeated seizures. If the medical treatment fails, hemispherectomy may be required on the condition that the other hemisphere is not involved. Characteristic features of Sturge–Weber disease are a diffuse enhancement of the surface of one hemisphere, typically over its posterior portion, reflecting the pial angiomatosis on T1 or FLAIR contrast-enhanced sequences. Using fat saturation, it may show associated abnormalities of the ocular choroid and in the calvarium as well. Other usual findings are a large choroid plexus in the ipsilateral ventricle and prominent, DVAlike transmedullary veins. Acute ischemia with focal edema, bleed or diffuse hemispheric swelling from prolonged seizure activity may be demonstrated. Hemispheric atrophy may result

from the seizure activity and/or from the perfusion defect. In infants, the white matter may present with the low T2 signal of a seizure-induced early myelination. Calcification is unusual in infants but develops over the years. Arteriovenous malformations are usually not epileptogenic in children, except for large ones. Cavernomas are often located at the cortical–subcortical junction and are often epileptogenic, assumedly because of the hemosiderin deposition, but a ­cortical dysplasia seems to be commonly associated (FCD 3c).19 They may also clot, or bleed. They are best depicted by their blooming artifacts on T2*GE/susceptibility imaging. Surgery is the treatment of choice. The rare sporadic meningioangiomatosis is characterized by a meningovascular proliferation and ­calcification;72 as mentioned above, the underlying cortex may be dysplastic and intrinsically epileptogenic (FCD 3c).19 The treatment then is surgical. Refractory epilepsy and epileptic encephalopathies are common complications of early occurring destructive brain lesions because of the aberrant reorganization of the neuronal circuitry in the vicinity of the scar (FCD 3d, see above). They typically relate to late gestational, perinatal, or early infantile events such as hemorrhages, ischemia, infection, trauma, HIE, and hypoglycemia. In severe cases, surgery for removal of the epileptogenic cortex may be an option.

„„ Conclusion In addition to the substrates mentioned above, any kind of brain malformation (including Chiari II malformation, or the classical commissural agenesis) may present with focal epilepsy. Over

19  Structural Brain Imaging in Pediatric Epilepsy the last decade, thanks to better signal-to-noise (high field, better coils) and better imaging sequence design, the efficacy of MRI in detecting and assessing the epilepsy-associated focal lesions and preparing the surgical strategy has consid-

erably improved. Changes may be subtle; however, to improve the diagnostic yield, MRI must be associated with the clinical assessment, EEG (and MEG where available) data, as well as functional studies with SPECT or PET.

References 1. Kimia AA, Ben-Joseph E, Prabhu S, et al. Yield of emergent neuroimaging among children presenting with a first complex febrile seizure. Pediatr Emerg Care 2012;28(4):316–321 2. Al-Shami R, Khair AM, Elseid M, et al. Neuro-imaging evaluation after the first afebrile seizure in children: a retrospective observational study. Seizure 2016;43:26–31 3. Berg AT, Testa FM, Levy SR, Shinnar S. Neuroimaging in children with newly diagnosed epilepsy: a community-based study. Pediatrics 2000;106(3):527–532 4. Bronen RA, Fulbright RK, Spencer DD, et al. Refractory epilepsy: comparison of MR imaging, CT, and histopathologic findings in 117 patients. Radiology 1996;201(1):97–105 5. Scott CA, Fish DR, Smith SJ, et al. Presurgical evaluation of patients with epilepsy and normal MRI: role of scalp video-EEG telemetry. J Neurol Neurosurg Psychiatry 1999;66(1):69–71 6. Wang ZI, Suwanpakdee P, Jones SE, et al. Re-review of MRI with post-processing in nonlesional patients in whom epilepsy surgery has failed. J Neurol 2016;263(9):1736–1745 7. Jeon TY, Kim JH, Lee J, Yoo SY, Hwang SM, Lee M. Value of repeat brain MRI in children with focal epilepsy and negative findings on initial MRI. Korean J Radiol 2017;18(4):729–738 8. Daghistani R, Widjaja E. Role of MRI in patient selection for surgical treatment of intractable epilepsy in infancy. Brain Dev 2013;35(8):697–705 9. Morimoto E, Kanagaki M, Okada T, et al. Anterior temporal lobe white matter abnormal signal (ATLAS) as an indicator of seizure focus laterality in temporal lobe epilepsy: comparison of double inversion recovery, FLAIR and T2W MR imaging. Eur Radiol 2013;23(1):3–11 10. Soares BP, Porter SG, Saindane AM, Dehkharghani S, Desai NK. Utility of double inversion recovery MRI in paediatric epilepsy. Br J Radiol 2016;89(1057):20150325 11. Guye M, Le Fur Y, Confort-Gouny S, et al. Metabolic and electrophysiological alterations in subtypes of temporal lobe epilepsy: a combined proton magnetic resonance spectroscopic imaging and depth electrodes study. Epilepsia 2002;43(10):1197–1209

19. Blümcke I, Thom M, Aronica E, et al. The clinicopathologic ­spectrum of focal cortical dysplasias: a consensus classification proposed by an ad hoc Task Force of the ILAE diagnostic methods commission. Epilepsia 2011;52(1):158–174 20. Blümcke I, Löbach M, Wolf HK, Wiestler OD. Evidence for developmental precursor lesions in epilepsy-associated glioneuronal tumors. Microsc Res Tech 1999;46(1):53–58 21. Pasquier B, Péoc’H M, Fabre-Bocquentin B, et al. Surgical pathology of drug-resistant partial epilepsy. A 10-year-experience with a series of 327 consecutive resections. Epileptic Disord 2002;4(2):99–119 22. Blümcke I, Aronica E, Becker A, et al. Low-grade epilepsy-­ associated neuroepithelial tumours—the 2016 WHO classification. Nat Rev Neurol 2016;12(12):732–740 23. Provenzale JM, Ali U, Barboriak DP, Kallmes DF, Delong DM, McLendon RE. Comparison of patient age with MR imaging features of gangliogliomas. AJR Am J Roentgenol 2000;174(3):859–862 24. Luyken C, Blümcke I, Fimmers R, Urbach H, Wiestler OD, Schramm J. Supratentorial gangliogliomas: histopathologic grading and tumor recurrence in 184 patients with a median follow-up of 8 years. Cancer 2004;101(1):146–155 25. Raybaud C. Cerebral hemispheric low-grade glial tumors in children: preoperative anatomic assessment with MRI and DTI. Childs Nerv Syst 2016;32(10):1799–1811 26. Tamburrini G, Colosimo C Jr, Giangaspero F, Riccardi R, Di Rocco C. Desmoplastic infantile ganglioglioma. Childs Nerv Syst 2003;19(5–6):292–297 27. Ostertun B, Wolf HK, Campos MG, et al. Dysembryoplastic neuroepithelial tumors: MR and CT evaluation. AJNR Am J ­ ­Neuroradiol 1996;17(3):419–430 28. Stanescu Cosson R, Varlet P, Beuvon F, et al. Dysembryoplastic neuroepithelial tumors: CT, MR findings and imaging follow-up— a study of 53 cases. J Neuroradiol 2001;28(4):230–240 29. Daghistani R, Miller E, Kulkarni AV, Widjaja E. Atypical characteristics and behavior of dysembryoplastic neuroepithelial tumors. Neuroradiology 2013;55(2):217–224

12. Wu WC, Huang CC, Chung HW, et al. Hippocampal alterations in children with temporal lobe epilepsy with or without a history of febrile convulsions: evaluations with MR volumetry and proton MR spectroscopy. AJNR Am J Neuroradiol 2005;26(5):1270–1275

30. Bulakbasi N, Kocaoglu M, Ors F, Tayfun C, Uçöz T. Combination of single-voxel proton MR spectroscopy and apparent diffusion coefficient calculation in the evaluation of common brain tumors. AJNR Am J Neuroradiol 2003;24(2):225–233

13. Widjaja E, Blaser S, Miller E, et al. Evaluation of subcortical white matter and deep white matter tracts in malformations of cortical development. Epilepsia 2007;48(8):1460–1469

31. Peters O, Gnekow AK, Rating D, Wolff JEA. Impact of location on outcome in children with low-grade oligodendroglioma. Pediatr Blood Cancer 2004;43(3):250–256

14. Szmuda M, Szmuda T, Springer J, et al. Diffusion tensor tractography imaging in pediatric epilepsy - A systematic review. Neurol Neurochir Pol 2016;50(1):1–6

32. Lipper MH, Eberhard DA, Phillips CD, Vezina LG, Cail WS. Pleomorphic xanthoastrocytoma, a distinctive astroglial tumor: neuroradiologic and pathologic features. AJNR Am J Neuroradiol 1993;14(6):1397–1404

15. Nilsson D, Starck G, Ljungberg M, et al. Intersubject variability in the anterior extent of the optic radiation assessed by tractography. Epilepsy Res 2007;77(1):11–16 16. Dorfer C, Widjaja E, Ochi A, Carter Snead Iii O, Rutka JT. Epilepsy surgery: recent advances in brain mapping, neuroimaging and surgical procedures. J Neurosurg Sci 2015;59(2):141–155 17. Widjaja E, Zamyadi M, Raybaud C, Snead OC, Doesburg SM, Smith ML. Disrupted global and regional structural network and subnetworks in children with localization-related epilepsy. AJNR Am J Neuroradiol 2015;36(7):1362–1368 18. Luyken C, Blümcke I, Fimmers R, et al. The spectrum of longterm epilepsy-associated tumors: long-term seizure and tumor outcome and neurosurgical aspects. Epilepsia 2003;44(6): 822–830

33. Im SH, Chung CK, Kim SK, Cho BK, Kim MK, Chi JG. Pleomorphic xanthoastrocytoma: a developmental glioneuronal tumor with prominent glioproliferative changes. J Neurooncol 2004;66(1–2):17–27 34. Lach B, Duggal N, DaSilva VF, Benoit BG. Association of pleomorphic xanthoastrocytoma with cortical dysplasia and neuronal tumors. A report of three cases. Cancer 1996;78(12):2551–2563 35. Palmini A, Gambardella A, Andermann F, et al. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 1995;37(4):476–487 36. Otsubo H, Ochi A, Elliott I, et al. MEG predicts epileptic zone in lesional extrahippocampal epilepsy: 12 pediatric surgery cases. Epilepsia 2001;42(12):1523–1530

169

170

IIc  Preoperative Neuroimaging 37. Taylor DC, Falconer MA, Bruton CJ, Corsellis JA. Focal dysplasia of the cerebral cortex in epilepsy. J Neurol Neurosurg Psychiatry 1971;34(4):369–387 38. Barkovich AJ, Guerrini R, Kuzniecky RI, Jackson GD, Dobyns WB. A developmental and genetic classification for malformations of cortical development: update 2012. Brain 2012;135(Pt 5):1348–1369 39. Englund C, Folkerth RD, Born D, Lacy JM, Hevner RF. Aberrant neuronal-glial differentiation in Taylor-type focal cortical dysplasia (type IIA/B). Acta Neuropathol 2005;109(5):519–533 40. Colombo N, Tassi L, Galli C, et al. Focal cortical dysplasias: MR imaging, histopathologic, and clinical correlations in surgically treated patients with epilepsy. AJNR Am J Neuroradiol 2003;24(4):724–733 41. Barkovich AJ, Kuzniecky RI, Bollen AW, Grant PE. Focal transmantle dysplasia: a specific malformation of cortical development. Neurology 1997;49(4):1148–1152 42. Andres M, Andre VM, Nguyen S, et al. Human cortical dysplasia and epilepsy: an ontogenetic hypothesis based on volumetric MRI and NeuN neuronal density and size measurements. Cereb Cortex 2005;15(2):194–210 43. Mellerio C, Roca P, Chassoux F, et al. The power button sign: a newly described central sulcal pattern on surface rendering MR images of type 2 focal cortical dysplasia. Radiology 2015;274(2):500–507 44. Colombo N, Tassi L, Deleo F, et al. Focal cortical dysplasia type IIa and IIb: MRI aspects in 118 cases proven by histopathology. Neuroradiology 2012;54(10):1065–1077 45. Grant PE, Barkovich AJ, Wald LL, Dillon WP, Laxer KD, Vigneron DB. High-resolution surface-coil MR of cortical lesions in medically refractory epilepsy: a prospective study. AJNR Am J Neuroradiol 1997;18(2):291–301 46. Besson P, Andermann F, Dubeau F, Bernasconi A. Small focal cortical dysplasia lesions are located at the bottom of a deep sulcus. Brain 2008;131(Pt 12):3246–3255 47. Marin-Padilla M. Prenatal and early postnatal ontogenesis of the human motor cortex: a golgi study. I. The sequential development of the cortical layers. Brain Res 1970;23(2):167–183 48. Lombroso CT. Can early postnatal closed head injury induce cortical dysplasia. Epilepsia 2000;41(2):245–253 49. Marín-Padilla M. Developmental neuropathology and impact of perinatal brain damage. I: Hemorrhagic lesions of neocortex. J Neuropathol Exp Neurol 1996;55(7):758–773 50. Marín-Padilla M. Developmental neuropathology and impact of perinatal brain damage. II: white matter lesions of the neocortex. J Neuropathol Exp Neurol 1997;56(3):219–235 51. Marín-Padilla M. Developmental neuropathology and impact of perinatal brain damage. III: gray matter lesions of the neocortex. J Neuropathol Exp Neurol 1999;58(5):407–429 52. Marín-Padilla M, Parisi JE, Armstrong DL, Sargent SK, Kaplan JA. Shaken infant syndrome: developmental neuropathology, progressive cortical dysplasia, and epilepsy. Acta Neuropathol 2002;103(4):321–332 53. Komori T, Arai N, Shimizu H, Yagishita A, Mizutani T, Oda M. Cortical perivascular satellitosis in intractable epilepsy: a form of cortical dysplasia? Acta Neuropathol 2002;104(2):149–154 54. Hazrati LN, Kleinschmidt-DeMasters BK, Handler MH, et al. Astrocytic inclusions in epilepsy: expanding the s­pectrum

of filaminopathies. J Neuropathol Exp Neurol 2008;67(7): 669–676 55. Schurr J, Coras R, Rössler K, et al. Mild malformation of cortical development with oligodendroglial hyperplasia in frontal lobe epilepsy: a new clinic-pathological entity. Brain Pathol 2017;27(1):26–35 56. Blümcke I, Thom M, Aronica E, et al. International consensus classification of hippocampal sclerosis in temporal lobe epilepsy: a Task Force report from the ILAE commission on diagnostic methods. Epilepsia 2013;54(7):1315–1329 57. Yagishita A, Arai N, Maehara T, Shimizu H, Tokumaru AM, Oda M. Focal cortical dysplasia: appearance on MR images. Radiology 1997;203(2):553–559 58. Demerens C, Stankoff B, Logak M, et al. Induction of myelination in the central nervous system by electrical activity. Proc Natl Acad Sci U S A 1996;93(18):9887–9892 59. Lega B, Mullin J, Wyllie E, Bingaman W. Hemispheric malformations of cortical development: surgical indications and approach. Childs Nerv Syst 2014;30(11):1831–1837 60. Barkovich AJ, Chuang SH. Unilateral megalencephaly: correlation of MR imaging and pathologic characteristics. AJNR Am J Neuroradiol 1990;11(3):523–531 61. Yagishita A, Arai N, Tamagawa K, Oda M. Hemimegalencephaly: signal changes suggesting abnormal myelination on MRI. Neuroradiology 1998;40(11):734–738 62. Di Rocco F, Novegno F, Tamburrini G, Iannelli A. Hemimegalencephaly involving the cerebellum. Pediatr Neurosurg 2001;35(5):274–276 63. D’Agostino MD, Bastos A, Piras C, et al. Posterior quadrantic dysplasia or hemi-hemimegalencephaly: a characteristic brain malformation. Neurology 2004;62(12):2214–2220 64. Wang DD, Knox R, Rolston JD, et al. Surgical management of medically refractory epilepsy in patients with polymicrogyria. Epilepsia 2016;57(1):151–161 65. Leblanc R, Tampieri D, Robitaille Y, Feindel W, Andermann F. ­Surgicaltreatmentofintractableepilepsyassociatedwiths­ chizencephaly. Neurosurgery 1991;29(3):421–429 66. Maehara T, Shimizu H, Nakayama H, Oda M, Arai N. Surgical treatment of epilepsy from schizencephaly with fused lips. Surg Neurol 1997;48(5):507–510 67. Cascino GD, Buchhalter JR, Sirven JI, et al. Peri-ictal SPECT and surgical treatment for intractable epilepsy related to schizencephaly. Neurology 2004;63(12):2426–2428 68. Arita K, Kurisu K, Kiura Y, Iida K, Otsubo H. Hypothalamic hamartoma. Neurol Med Chir (Tokyo) 2005;45(5):221–231 69. Freeman JL, Coleman LT, Wellard RM, et al. MR imaging and spectroscopic study of epileptogenic hypothalamic hamartomas: analysis of 72 cases. AJNR Am J Neuroradiol 2004;25(3):450–462 70. Pardo CA, Nabbout R, Galanopoulou AS. Mechanisms of epileptogenesis in pediatric epileptic syndromes: Rasmussen encephalitis, infantile spasms, and febrile infection-related epilepsy syndrome (FIRES). Neurotherapeutics 2014;11(2):297–310 71. Bien CG, Urbach H, Deckert M, et al. Diagnosis and staging of Rasmussen’s encephalitis by serial MRI and histopathology. Neurology 2002;58(2):250–257 72. Wiebe S, Munoz DG, Smith S, Lee DH. Meningioangiomatosis. A comprehensive analysis of clinical and laboratory features. Brain 1999;122(Pt 4):709–726

20

  Functional Magnetic Resonance Imaging in Pediatric Epilepsy Surgery Torsten Baldeweg and Frédérique Liégeois

Summary Functional MRI (fMRI) has been proven to be an important noninvasive diagnostic tool for neurosurgical practice at four key stages: (1) assessing the feasibility of surgical resection and predicting its risk for cognition; (2) refining the extent and location of the resection; (3) selecting patients for invasive functional mapping procedures; and (4) visualizing functional areas intraoperatively. Here we review the latest evidence on the application of fMRI in pediatric surgical practice, from recent methodological improvements in fMRI scanning of children, assessment of language and memory lateralization, the localization of eloquent cortex, as well as localization of epileptic discharges. Keywords:  fMRI, epilepsy surgery, eloquent cortex, language lateralization, language dominance, language, memory, ­children, epileptic discharge

„„ Instrumentation and Methods Basic Principles of Functional Magnetic Resonance Imaging fMRI is based on the dependence of T2-weighted signal on the oxygenation status of hemoglobin, and the derived signal is called blood oxygen level dependent (BOLD) signal. The fMRI signal is delayed with respect to the onset of neuronal activity by approximately 6 seconds and shows a more temporally protracted time course, lasting approximately 20 seconds before returning to baseline. Neurophysiological studies in the monkey visual cortex have shown that neuronal excitation results in enhanced (positive) BOLD signal, whereas reduction in net neural excitation results in reduced (negative) BOLD signal.8 The spatial resolution is in the order of several mm (typical voxel size: 3 × 3 × 3 mm3), and its effective resolution will depend on the spatial filtering used during signal processing.

„„ Introduction

Experimental Designs

fMRI has rapidly entered the armory of imaging methods available for the management of patients with focal epilepsy who may be surgery candidates. This is particularly true for presurgical investigations of cognitive and motor functions. The factors that have contributed to its success are chiefly its noninvasiveness, reproducibility, and wide availability. fMRI can reveal clinically useful information about the functional status of cortical tissue in relation to eloquent functions and, more recently, about origin and spread of epileptic activity. In this chapter we will outline some of the basic principles of application of fMRI for complementing the neuropsychological evaluation of children and adolescents undergoing neurosurgical treatment. We will focus on two main issues that have received most attention so far: does fMRI provide a noninvasive means to reliably estimate hemispheric dominance of language and memory functions and localization of eloquent cortex? Finally, we will briefly mention developments in the mapping of blood flow changes associated with epileptic activity. The use of fMRI in mapping sensorimotor cortex before surgery has been reviewed previously1,​2 and will not be covered in this chapter. Several reviews cover different aspects relevant to fMRI application to epilepsy surgery evaluation3,​4,​5 and its use in pediatric populations.1,​6,​7

fMRI activation is detected using one of two experimental paradigms: block designs consisting of repeated activation-baseline state cycles and event-related designs, in which discrete events are analyzed separately, allowing for control of behavioral task performance during data analysis. The latter is used for studies of memory (e.g., comparing recalled versus forgotten/new items), whereas the former is commonly used with language and motor activation studies. Block designs have been used most consistently in children and adolescents, because their higher signal-to-noise ratio reduces the scanning time required to obtain robust fMRI activation. Stimulation tasks are tailored to the cognitive domains under investigation, using visual or auditory stimulus presentation. The most commonly used task to assess hemispheric dominance for expressive language is silent word generation to letters (fluency) or words (verb or synonym generation). Other language tasks include story comprehension, semantic d ­ ecision-making (based on an object description,9 or “gap f­illing” in heard ­stories).10 Memory tasks commonly use r­ ecognition of visually presented words, pictures, or human faces. In recent years, the use of resting-state fMRI (rs-fMRI) for the presurgical evaluation of children has also emerged.11 The major advantage of this method is that no task is per-

172

IIc  Preoperative Neuroimaging formed, with children awake or under sedation. Connectivity within “resting-state networks” is derived from the synchronized low frequency fluctuations in BOLD signal that spontaneously occur between distant brain regions. The ability of rs-fMRI methods to lateralize functions at an individual level will need replication and validation before they become part of presurgical planning.

Response Monitoring Earlier language fMRI studies have instructed patients to covertly generate verbal responses to avoid head movements caused by overt speech. Although this has generally worked well with most patients, it has the disadvantage of being prone to ambiguous results if activation patterns are atypical or in loss of activation in less cooperative patients. More recent studies have used experimental protocols that require patients to respond using button presses to indicate a choice between different stimulus categories or presence of certain target items.12 Overt speech tasks are also widespread, using either continuous or sparse fMRI acquisition to allow for overt verbal responses during silent gaps.13 Overt speech during continuous fMRI scanning is made possible using sensitive MRI-compatible microphones. The induced subtle movement artifacts can be reduced using modern motion- and artifact-suppression post-processing methods, such as motion fingerprint14 and functional image artefact correction heuristic.15 A direct comparison of covert and overt speech during language fMRI scanning in children with focal epilepsy showed greater sensitivity of overt speech for identifying activation in language cortex and generally improved the yield of presurgical fMRI.16

Assessment of Laterality of fMRI Activation The term fMRI lateralization is commonly used to refer to the hemispheric asymmetry of fMRI activation in a given region of interest and may refer to extent of activation, level of activation, or both. Some epilepsy centers successfully rely on a qualitative judgment of laterality, performed by fMRI experienced neuroradiologists,12,​17,​18 whereas others use a combination of qualitative and statistical measures19,​20 as converging evidence. A major practical advance here is the “LI-toolbox,” which derives robust, threshold-independent estimates of laterality in specific regions of interest (Broca’s and Wernicke’s regions, cerebellum) using a bootstrapping algorithm.19

„„ Applications for Presurgical Evaluation The majority of published studies have been conducted on adults, and only a minority of pediatric studies have been reported.20,​21,​22,​23,​24 We therefore review here the evidence from both adult and childhood epilepsy studies combined while attempting to point out issues that may have specific significance for pediatric practice.

Preoperative Assessment of Language Lateralization Language fMRI tasks are used in candidates for epilepsy ­surgery with a view to improve the prognosis for postsurgical speech and language deficits. The main questions are whether ­surgery is planned in the language-dominant hemisphere and ­whether language cortex is located near the planned resection. The increased frequency of atypical language lateralization (i.e., right-sided or bilateral) in patients with focal epilepsy or lesions of the left hemisphere has been known for more than 40 years.25 Early studies using the intracarotid amobarbital test (IAT, or Wada test), also suggested that mainly lesions within classic language cortex (Broca’s and Wernicke’s regions) are responsible for inducing a shift of language to the right hemisphere. More recent investigations, with the benefit of modern neuroimaging, have, nevertheless, shown that a considerable proportion of patients with early acquired or developmental left-sided perisylvian lesions showed evidence of intrahemispheric reorganization of language (i.e., retain typical left-­ sided lateralization; see example in Fig. 20.1a), often near the lesion.23,​24,​26 Furthermore, patients with epilepsy arising from pathology in regions remote from classic language cortex, especially in the mesial temporal cortex, often show atypical, often bilateral, language representation (Fig. 20.1b).23,​27,​28 A study in a large cohort of patients with a left hemisphere epileptogenic focus (including children) identified the following factors associated with atypical language lateralization: left-handedness, onset of epilepsy before age 6, and MRI lesion type.29 Regarding the latter pathology factor, it is notable that patients with stroke showed a very high rate of reorganization and that approximately 35% of patients with a normal MRI had atypical language. This latter finding points to the possibility that epileptic activity drives functional reorganization.30 In children with left-sided focal lesional epilepsy, factors predicting reorganization included lesion location—in anterior superior temporal gyrus and posterior inferior frontal (Broca’s) region— and left-handedness (mostly pathology-induced). This study also indicated that structural brain asymmetries, such as a short planum temporale in the right hemisphere, can constrain the likelihood of language reorganization to the right, contralesional hemisphere.31 It is noteworthy that there can be multiple patterns of language reorganization in patients with epilepsy, including a discrepancy between temporal and frontal lateralization within the same individuals (“crossed lateralization”; Fig. 20.1c).32 This indicates that language dominance should be assessed on a regional or lobar basis rather than using a single hemispheric lateralization index.

Assessment of Language Lateralization: fMRI and IAT Studies on Adults Although the IAT remains the gold standard for determining hemispheric dominance for speech and language, many epilepsy centers now routinely use fMRI.5,​33 Comparative studies using both methods17,​18,​34 in adults have found agreement in approximately 80 to 90% of cases. Notwithstanding the obvious

20  Functional Magnetic Resonance Imaging in Pediatric Epilepsy Surgery

Fig. 20.1  Examples of functional MRI language reorganization patterns in children with focal lesions of the left hemisphere using a silent verb generation task. (a) Intrahemispheric reorganization in two patients with extensive perisylvian developmental lesions (cases no. 1 and 2; for more details see Liégeois et al23). (b) Interhemispheric language reorganization in a child (case no. 3) with hippocampal sclerosis (circled). The bilateral fMRI language representation was confirmed by intracarotid amobarbital test. (c) “Crossed” lateralization in a child with a posterior temporal lesion. The expressive language task (verb generation) activates left inferior frontal and posterior temporal regions, while the receptive task (story comprehension) recruits mainly right temporal regions. (d) Contralesional language organization in a patient with a left hemisphere stroke lesion before and after hemispherectomy. Surgery resulted in no language deterioration. Note: the crosshair indicates local maximum activation. Left hemisphere is on the left.

difference between an observation-based ­method (fMRI) and an ­inhibition procedure (IAT),2 multiple other factors are likely to cause disparity,12,​17,​35 including differences between fMRI and IAT tasks18,​36 and variability in the regions of interest chosen for fMRI

lateralization analysis.35,​37 ­Finally, in contrast to the IAT, fMRI will also detect activation in regions that are not essential for the performance of a task (redundant activation). If these regions are localized in the nondominant hemisphere, this is

173

174

IIc  Preoperative Neuroimaging likely to render fMRI findings more bilateral (see examples in Woermann et al18). No method is currently available that could distinguish essential from redundant activation foci on fMRI. The contralesional activation detected on fMRI may also indicate the potential for postsurgical reorganization of function. Indeed, two studies that have used both fMRI and IAT have found fMRI to be a better predictor of postoperative cognitive outcome. The study by Sabsevitz et al38 in 24 patients who underwent left anterior lobe lobectomy (L-ATL) showed the fMRI lateralization index to be 100% sensitive and 73% predictive of significant visual naming decline. A study by Binder et al39 corroborates this finding for verbal memory outcome after L-ATL and will be discussed in more detail later. A large-scale study in 229 adult patients with epilepsy40 has demonstrated concordance of language fMRI with IAT, especially in left lateralized cases. Discordance between methods increased with increasing bilaterality of fMRI. Extreme discordance was rare (< 2%) and could be due to crossed language dominance between frontal and temporal regions, which was not assessed in this study. Although the direct comparison of fMRI with IAT was an obvious first step in validating this new method, the IAT itself is not free of potential problems,41 such as lack of standardization and reproducibility, agitation and obtundation in some patients, potential cross-flow between the hemispheres, focus on expressive tasks, and many others (see Chapter 26). Indeed, there are reports of erroneous lateralization using the IAT, as evidenced by electrocortical stimulation (ECS)42 or postsurgical dysphasia43 in which functional neuroimaging had indicated the correct lateralization. Therefore, only further postoperative outcome studies of the kind reported by Sabsevitz et al38 and Binder et al39 can demonstrate the true predictive value of fMRI-derived language lateralization. In summary, a review of all available studies44 concluded that that fMRI greatly increases the probability of correctly predicting language dominance in multiple subgroups of surgery patients with and without epilepsy.

Pediatric Studies Only a few studies have specifically investigated the role of fMRI in pediatric epilepsy surgery candidates,20,​21,​22,​24,​45 commonly comparing fMRI with a mixture of invasive investigations (IAT, ECS) or clinical observations. These studies have confirmed the feasibility and accuracy of fMRI in estimating language dominance in children, with the proviso that in some bilateral fMRI cases, only unilateral corroborating evidence was available.20,​24 A study involving a small series of children who also underwent ECS and IAT reported bilateral fMRI activation more often than suggested by IAT46; however, details of the procedure were not given. A recent series in 20 children showed excellent concordance with no case of discordant classification.45 The authors cautioned that in cases with bilateral language ESM or IAT are still performed, and that visual expert review of fMRI is superior in accurately predicting dominance. Longitudinal fMRI studies are particularly useful for investigating children with epilepsy caused by extensive left ­hemispheric injury or progressive neurodegenerative conditions, such as Rasmussen’s encephalitis, in revealing the gradual process of language reorganization,47 which can be used to ­optimize the timing of surgery and perhaps also for predicting

the level of language proficiency after surgery.13 Longitudinal studies p ­ ostsurgery also provide insights into additional regions involved in language recovery following “shift” to the nondominant hemisphere.48 More recently, some studies have also used rs-fMRI to assess language dominance49 reporting high concordance between lateralization based on resting-state language networks and the IAT procedure in 23 patients with mixed intractable epilepsy, six of whom were children (10–17 years). Sensitivity, accuracy, and specificity were high (96%). If replicated, these findings are promising for the evaluation of less cooperative patients.

Localization of Language Cortex: fMRI and ECS Studies on Adults ECS is the gold standard method for mapping eloquent cortex before surgery. However, the majority of ECS sites tested intraoperatively are not associated with language-associated deficits.50 Given also the large variability in the location of individual ECS language sites, it is, therefore, desirable to be able to selectively target critical regions preoperatively. This is particularly true when ECS is applied to children where cooperation and motivation during lengthy testing sessions can be an issue, even when performed extraoperatively. The first studies have shown some promising results,51,​52,​53 with sensitivity of fMRI in predicting ECS language sites ranging from 80 to 100%, especially if multiple fMRI language tasks, both auditory and visual, were combined. These studies also showed that the correlation is confounded by fMRI activation in redundant (noncritical) language sites, usually leading to decreased specificity of approximately 50% for fMRI identifying positive ECS sites. The study by Roux et al54 showed less encouraging results (maximal sensitivity 66% for combined verb generation and naming tasks) and highlighted the challenges in combining both modalities. By necessity, different stimulations tasks are used for fMRI and ECS, often using different response modes (covert vs. overt). There are inherent spatial inaccuracies in both fMRI and ECS, amounting up to approximately 1 cm each. Furthermore, the fMRI signal-to-noise ratio may not be sufficient if very short scanning sessions are used.54 Indeed, the reports with consistent correlations51,​53 used at least three different fMRI language tasks of sufficient length to achieve good signal-to-noise ratio and convergence of activation in critical language regions. It also appears that tasks that involve sentence comprehension are better suited to activate temporoparietal language areas than those that use single-word or item processing.52,​53 The issue of statistical threshold for the display of fMRI activation foci is particularly critical because the strength of ­activation may vary considerably across individuals.19 ­FitzGerald et al51 suggested using variable thresholds so the a ­ ctivation results in approximately 1 cm2 or larger extent of cluster size at the cortical surface, which would be in agreement with the estimated extent of ECS language foci. By contrast, Rutten et al53 used an operator-independent, fixed threshold: perhaps favored by a good signal-to-noise ratio in their study. The ability of presurgical language fMRI in identifying intraoperative ECS sites in brain tumor surgery was reviewed by Guissani et al,55 concluding that current evidence does not allow

20  Functional Magnetic Resonance Imaging in Pediatric Epilepsy Surgery fMRI to be considered as an alternative. Sensitivity ranged from 59 to 100% and specificity from 0 to 97% across studies with variable methodology.

Pediatric Studies A comparative study of eight children reported high sensitivity of language fMRI (covert sentence generation) in detecting ECS sites (100%) but again low specificity.56 One major caveat is that different ECS tasks were used, such as counting, reading, or spontaneous speech. The authors also point out that using a fixed statistical threshold for displaying fMRI maps might miss critical sites. fMRI should help increase the yield of ECS in children by optimizing the placement of depth electrodes. Our own experience with extraoperative language ECS in pediatric patients is in agreement with the previously mentioned conclusions. ECS sites are commonly found in proximity to major fMRI activation foci (see example in Fig. 20.2a); however, in agreement with other authors, the presence of fMRI activation does not predict an ECS site with certainty. The past use of a single fMRI task (auditory verb generation), although resulting in robust and reproducible activations, was clearly insufficient to map essential ECS sites if elicited by a different task during ECS (Fig. 20.2b). The use of multiple tasks and reproducible data can overcome this limitation. In addition, fMRI can indicate atypical locations of eloquent cortex, which may escape detection by ECS, especially when located deep in the vicinity of cortical sulci. This has been documented by Rutten et al57 in a 14-year-old patient with a tumor displacing Broca’s region.

In summary, although the first comparative studies were motivated by the desire to replace ECS with fMRI, currently this does not appear to be a realistic expectation. Nevertheless, the evidence accumulated so far suggests that fMRI language mapping can help in planning the extent of craniotomy, guiding the placement of subdural and stereo-electrodes, and in targeting of sites for ECS.

Preoperative Assessment of Memory Lateralization Surgical resection of the anterior temporal lobe is an established treatment for medication-resistant temporal lobe epilepsy (TLE) in both adults and children. A major clinical concern is the risk of verbal memory deficits as a consequence of ATL in the ­language-dominant hemisphere. While the long-term outcome in children is more favorable than in adults, with most children recovering to preoperative memory performance,58 there is nevertheless a risk for significant memory deficits in the short term.59,​60,​61 Indeed, verbal memory deficits are greatest if surgery is performed in the language-dominant temporal lobe, as defined by language fMRI.58,​61 ATL surgery in children is also associated with varying degrees of deficit in semantic memory (memory for facts, which is distinct from memory for episodes).58 Semantic memory is dependent on the anterior temporal neocortex and fMRI can help in lateralizing the degree of involvement using semantic decision or comprehension tasks.62

Fig. 20.2  (a) Illustrations of possible caveats in correlating language functional MRI and electrocortical stimulation (ECS) mapping. (a) Case no. 6: Frontal activation (green arrow) was displaced dorsally by focal cortical dysplasia (circle) within classic Broca’s region and confirmed using extraoperative ECS (red arrow and red dots on cortical reconstruction, bottom left). A second ECS site in the anterior superior temporal gyrus (yellow arrow) was not visible on fMRI. (b) Case no. 7: Intraoperative mapping identified multiple ECS sites to auditory (blue dots) and visual (yellow dots) confrontation naming in a child with a focal cortical dysplasia (circled on MRI and delineated on the photograph). fMRI failed to identify the extent of eloquent cortex (yellow arrow) in the posterior superior temporal (Wernicke’s) region. Only a small fMRI activation focus was found posteriorly to the lesion (green arrow), suggesting that an additional fMRI task may have been useful. fMRI correctly identified the language dominant hemisphere in both cases. Note: the crosshair indicates the local maximum activation using silent verb generation task. Left hemisphere is on the left.

175

176

IIc  Preoperative Neuroimaging So far only a few studies have demonstrated the feasibility of conducting memory fMRI studies at group level in healthy children,63,​64 and there are currently no reports from clinical pediatric populations. We will therefore only briefly review fMRI studies in adult TLE that have focused on the following aspects relevant to epilepsy surgery (see review in Powell et al65): fMRI studies have demonstrated material-specific lateralization and localization of memory processes in the medial temporal lobe (MTL). ­Typically, verbal memory for words is associated with left MTL activation and memory for faces with right MTL; memory for visual objects results in bilateral activation.66 fMRI studies in adult patients have uncovered how unilateral TLE affects the organization of the MTL memory systems. Typically, patients with left-sided TLE (mostly caused by hippocampal sclerosis) show fMRI evidence for reorganization of verbal memory functions to the right MTL (focused on the hippocampus), whereas patients with right-sided TLE show greater lateralization to the left MTL.66 Nonverbal tasks, such as mental navigation or visual scene encoding, show bilateral MTL activation in control subjects and asymmetric activation shifted contralateral to the side of TLE seizure focus.67,​68,​69 fMRI lateralization of memory was concordant with IAT memory scores in the majority of patients, but studies are limited to small numbers of cases.68,​70 Furthermore, hippocampal volume loss was correlated with MTL activation in a material and side-specific way: v ­erbal ­encoding-correlated activation in the left MTL with left hippocampal volume and picture encoding-correlated activation in the right MTL with right hippocampal volume.71 It is therefore not surprising that fMRI activation within the sclerotic ­hippocampus correlated with memory scores: for the left side with verbal memory, and for the right side with nonverbal memory. Conversely, activation within the contralesional hippocampus was negatively correlated with memory performance, suggesting that reorganized MTL function did not ­positively contribute to preoperative memory function. The critical question is whether these fMRI findings are relevant to prediction of postoperative memory changes after ATL. Indeed, correlations of preoperative memory fMRI with postoperative memory changes have been demonstrated in small groups of ATL patients. The consistent finding across these studies in exclusively adult patient groups is that the stronger the ipsilesional MTL activation is or the larger its activation asymmetry is,66,​72 the larger the postoperative memory loss will be.73 This is true for verbal memory change66 as well as nonverbal memory loss.66,​69 These findings are consistent with the functional adequacy model of memory deficits following ATL in adults74 and the fact that higher preoperative memory performance is a predictor of larger postoperative decline. Contralesional MTL fMRI activations, which can be seen as evidence for functional reorganization caused by unilateral TLE, have not been found to correlate with postoperative memory performance, at least, at the short postoperative follow-up periods reported so far. Although memory fMRI is technically challenging and the resulting activation strength within MTL regions is usually low, an alternative approach is the use of fMRI language lateralization in predicting the memory effect of ATL. Indeed, Binder and colleagues39 reported that fMRI improved the prediction of verbal memory outcome in comparison with IAT memory ­lateralization in a large cohort of adult left ATL patients.

­Similarly, a large scale study9 of adults and children with and without epilepsy demonstrated that MTL activation could be derived from a language paradigm (auditory semantic decision task). Two findings emerged from this study. First, language fMRI lateralization predicted a significant proportion of variance in MTL lateralization in adults, but not in children. Second, children showed weaker left MTL lateralization than adults. Future studies should establish whether memory fMRI will add predictive power in addition to neuropsychological ­evaluation, hippocampal volumetry, and language fMRI lateralization.

Localization of Epileptic Discharge Activity Using fMRI Studies on Adults Since the first description of BOLD changes during a focal seizure in a child,75 fMRI has been increasingly used to study brain blood flow changes associated with normal and abnormal electroencephalography (EEG) activity.76 EEG-spike associated BOLD responses have been localized in the absence of structural MRI abnormalities,77 which has led to the expectation that EEG-fMRI may help in identifying potential surgical candidates in whom other source localization techniques have failed to localize a single focus. Only a variable proportion of adult patients show significant fMRI activation correlated with focal EEG discharges; however, a degree of topographical concordance between the two modalities is found in the majority of those cases. Recent studies have demonstrated that EEG-fMRI during ictal and interictal events in patients with malformations of cortical development can determine the involvement of the lesion in epileptogenesis and may help determine the potential surgical target.78 Zijlmans et al79 evaluated EEG-fMRI in adult surgical ­candidates who were considered ineligible because of unclear EEG foci or multifocality. EEG-fMRI either improved source localization leading to reevaluation of surgical candidacy or corroborated the initial negative decision. The following ­guidelines for the application of EEG-fMRI have been proposed by those authors: 1. The best indication is for source localization in extratemporal (MRI-negative) epilepsy and when questions about the depth of the source arise. 2. In the case of presumed multifocality, EEG-fMRI will likely confirm this hypothesis, but, incidentally, it can favor one of the putative sources. 3. A priori allocation of a region of interest is crucial: topographically unrelated co-(de)activation may lack a clinical relevance. 4. EEG-fMRI can guide invasive electrode placement.

Pediatric Studies Encouraging findings have been obtained in a prospective cohort of 53 children with pharmacoresistant focal epilepsy.80 fMRI localization was compared with EEG electrical source imaging (ESI) of interictal discharges and surgical outcome. The epileptic zone was well localized in 29 patients, and seizure outcome was correctly predicted by EEG-fMRI in 8 of 20 surgical patients, and by combined EEG-fMRI/ESI in 9 of 9 patients, including three with no visible lesion on MRI. This

20  Functional Magnetic Resonance Imaging in Pediatric Epilepsy Surgery study suggests that EEG-fMRI combined with ESI may predict surgery outcome better than each individual test, including in MRI-negative patients.

„„ Conclusion fMRI examinations play an important role within a multidisciplinary epilepsy surgery workup. In combination with comprehensive neuropsychological assessments of intellectual, language, and memory abilities, fMRI language lateralization findings help to estimate the risk for postoperative cognitive changes. It is also expected that in the near future, fMRI evaluations of memory functions will be possible in children, much like in adult TLE patients. Despite the existing caveats, in particular in relation to its ability in localizing eloquent cortex, it is important to stress that fMRI methods have made considerable gains in recent years, which have impacted on its usability in children, across a wide age and ability range. This includes improvements in image acquisition and task design as well as post-processing. Higher field strengths have resulted in better signal-to-noise ratio and reduced scanning times. Monitoring of task performance, including overt speech, during scanning can reduce ambiguity of atypical

activation patterns. Finally, methods that reduce the dependence of fMRI lateralization indices on statistical thresholds are now in routine use. fMRI has been proven to be an important diagnostic tool for neurosurgical practice at four key stages: 1. Assessing the feasibility of surgical resection and predicting its risk for cognition. 2. Refining the extent and location of the resection in surgical candidates. 3. Selecting patients for invasive functional mapping procedures. 4. Visualizing functional areas intraoperatively using image guidance systems. Recent practice guidelines for epilepsy surgery evaluation of adults81 recommend fMRI for lateralizing language functions in place of the intracarotid amobarbital procedure in patients with focal epilepsy and for predicting postsurgical language and memory outcome after ATL for TLE. However, the level of evidence to support the replacement of invasive techniques may vary between types of epileptogenic lesions. While such systematic evidence in children is still not yet available, in many pediatric epilepsy centers the use of IAT has dramatically declined since the introduction of fMRI.

References 1. Liégeois F, Cross JH, Gadian DG, Connelly A. Role of fMRI in the decision-making process: epilepsy surgery for children. J Magn Reson Imaging 2006;23(6):933–940 2. Tharin S, Golby A. Functional brain mapping and its applications to neurosurgery. Neurosurgery 2007;60(4, Suppl 2):185–201, discussion 201–202 3. Detre JA. Clinical applicability of functional MRI. J Magn Reson Imaging 2006;23(6):808–815 4. Bartsch AJ, Homola G, Biller A, Solymosi L, Bendszus M. Diagnostic functional MRI: illustrated clinical applications and decisionmaking. J Magn Reson Imaging 2006;23(6):921–932 5. Bargalló N. Functional magnetic resonance: new applications in epilepsy. Eur J Radiol 2008;67(3):401–408 6. Sachs BC, Gaillard WD. Organization of language networks in children: functional magnetic resonance imaging studies. Curr Neurol Neurosci Rep 2003;3(2):157–162 7. O’Shaughnessy ES, Berl MM, Moore EN, Gaillard WD. Pediatric functional magnetic resonance imaging (fMRI): issues and applications. J Child Neurol 2008;23(7):791–801 8. Logothetis NK, Pauls J, Augath M, Trinath T, Oeltermann A. Neurophysiological investigation of the basis of the fMRI signal. Nature 2001;412(6843):150–157 9. Sepeta LN, Berl MM, Wilke M, et al. Age-dependent mesial temporal lobe lateralization in language fMRI. Epilepsia 2016;57(1):122–130 10. Lidzba K, Schwilling E, Grodd W, Krägeloh-Mann I, Wilke M. ­Language comprehension vs. language production: age effects on fMRI activation. Brain Lang 2011;119(1):6–15 11. Vadivelu S, Wolf VL, Bollo RJ, Wilfong A, Curry DJ. Resting-state functional MRI in pediatric epilepsy surgery. Pediatr Neurosurg 2013;49(5):261–273 12. Gaillard WD, Balsamo L, Xu B, et al. fMRI language task panel improves determination of language dominance. Neurology 2004;63(8):1403–1408

13. Liégeois F, Connelly A, Baldeweg T, Vargha-Khadem F. Speaking with a single cerebral hemisphere: fMRI language ­ ­organization after hemispherectomy in childhood. Brain Lang 2008;106(3):195–203 14. Wilke M. An alternative approach towards assessing and accounting for individual motion in fMRI timeseries. Neuroimage 2012;59(3):2062–2072 15. Tierney TM, Weiss-Croft LJ, Centeno M, et al. FIACH: A biophysical model for automatic retrospective noise control in fMRI. Neuroimage 2016;124(Pt A):1009–1020 16. Croft LJ, Rankin PM, Liégeois F, et al. To speak, or not to speak? The feasibility of imaging overt speech in children with epilepsy. Epilepsy Res 2013;107(1–2):195–199 17. Benke T, Köylü B, Visani P, et al. Language lateralization in temporal lobe epilepsy: a comparison between fMRI and the Wada test. Epilepsia 2006;47(8):1308–1319 18. Woermann FG, Jokeit H, Luerding R, et al. Language lateralization by Wada test and fMRI in 100 patients with epilepsy. Neurology 2003;61(5):699–701 19. Wilke M, Lidzba K. LI-tool: a new toolbox to assess lateralization in functional MR-data. J Neurosci Methods 2007;163(1):128–136 20. Liégeois F, Connelly A, Salmond CH, Gadian DG, Vargha-Khadem F, Baldeweg T. A direct test for lateralization of language activation using fMRI: comparison with invasive assess­ ments in children with epilepsy. Neuroimage 2002;17(4): 1861–1867 21. Hertz-Pannier L, Gaillard WD, Mott SH, et al. Noninvasive assessment of language dominance in children and adolescents with functional MRI: a preliminary study. Neurology 1997;48(4):1003–1012 22. Stapleton SR, Kiriakopoulos E, Mikulis D, et al. Combined utility of functional MRI, cortical mapping, and frameless stereotaxy in the resection of lesions in eloquent areas of brain in children. Pediatr Neurosurg 1997;26(2):68–82

177

178

IIc  Preoperative Neuroimaging 23. Liégeois F, Connelly A, Cross JH, et al. Language reorganization in children with early-onset lesions of the left hemisphere: an fMRI study. Brain 2004;127(Pt 6):1229–1236

c­ omparing positron emission tomography visual and computerized region of interest analysis with the Wada test. Ann Neurol 1999;45(5):662–665

24. Anderson DP, Harvey AS, Saling MM, et al. FMRI lateralization of expressive language in children with cerebral lesions. Epilepsia 2006;47(6):998–1008

44. Medina LS, Bernal B, Ruiz J. Role of functional MR in determining language dominance in epilepsy and nonepilepsy populations: a Bayesian analysis. Radiology 2007;242(1):94–100

25. Rasmussen T, Milner B. The role of early left-brain injury in determining lateralization of cerebral speech functions. Ann N Y Acad Sci 1977;299:355–369

45. Rodin D, Bar-Yosef O, Smith ML, Kerr E, Morris D, Donner EJ. Language dominance in children with epilepsy: concordance of fMRI with intracarotid amytal testing and cortical stimulation. Epilepsy Behav 2013;29(1):7–12

26. Duchowny M, Jayakar P, Harvey AS, et al. Language cortex representation: effects of developmental versus acquired pathology. Ann Neurol 1996;40(1):31–38 27. Adcock JE, Wise RG, Oxbury JM, Oxbury SM, Matthews PM. Quantitative fMRI assessment of the differences in lateralization of language-related brain activation in patients with temporal lobe epilepsy. Neuroimage 2003;18(2):423–438 28. Weber B, Wellmer J, Reuber M, et al. Left hippocampal pathology is associated with atypical language lateralization in patients with focal epilepsy. Brain 2006;129(Pt 2):346–351 29. Gaillard WD, Berl MM, Moore EN, et al. Atypical language in lesional and nonlesional complex partial epilepsy. Neurology 2007;69(18):1761–1771 30. Janszky J, Mertens M, Janszky I, Ebner A, Woermann FG. Left-sided interictal epileptic activity induces shift of language lateralization in temporal lobe epilepsy: an fMRI study. Epilepsia 2006;47(5):921–927 31. Pahs G, Rankin P, Helen Cross J, et al. Asymmetry of planum temporale constrains interhemispheric language plasticity in children with focal epilepsy. Brain 2013;136(Pt 10):3163–3175 32. Berl MM, Zimmaro LA, Khan OI, et al. Characterization of atypical language activation patterns in focal epilepsy. Ann Neurol 2014;75(1):33–42 33. Szaflarski JP, Holland SK, Jacola LM, Lindsell C, Privitera MD, Szaflarski M. Comprehensive presurgical functional MRI language evaluation in adult patients with epilepsy. Epilepsy Behav 2008;12(1):74–83 34. Binder JR, Swanson SJ, Hammeke TA, et al. Determination of language dominance using functional MRI: a comparison with the Wada test. Neurology 1996;46(4):978–984 35. Rutten GJ, Ramsey NF, van Rijen PC, Alpherts WC, van Veelen CW. FMRI-determined language lateralization in patients with unilateral or mixed language dominance according to the Wada test. Neuroimage 2002;17(1):447–460 36. Risse GL, Gates JR, Fangman MC. A reconsideration of bilateral language representation based on the intracarotid amobarbital procedure. Brain Cogn 1997;33(1):118–132 37. Lehéricy S, Cohen L, Bazin B, et al. Functional MR evaluation of temporal and frontal language dominance compared with the Wada test. Neurology 2000;54(8):1625–1633 38. Sabsevitz DS, Swanson SJ, Hammeke TA, et al. Use of preoperative functional neuroimaging to predict language deficits from epilepsy surgery. Neurology 2003;60(11):1788–1792 39. Binder JR, Sabsevitz DS, Swanson SJ, Hammeke TA, Raghavan M, Mueller WM. Use of preoperative functional MRI to predict verbal memory decline after temporal lobe epilepsy surgery. Epilepsia 2008;49(8):1377–1394 40. Janecek JK, Swanson SJ, Sabsevitz DS, et al. Language lateralization by fMRI and Wada testing in 229 patients with epilepsy: rates and predictors of discordance. Epilepsia 2013;54(2):314–322 41. Abou-Khalil B. An update on determination of language dominance in screening for epilepsy surgery: the Wada test and newer noninvasive alternatives. Epilepsia 2007;48(3):442–455 42. Kho KH, Leijten FS, Rutten GJ, Vermeulen J, Van Rijen P, Ramsey NF. Discrepant findings for Wada test and functional magnetic resonance imaging with regard to language function: use of electrocortical stimulation mapping to confirm results. Case report. J Neurosurg 2005;102(1):169–173 43. Hunter KE, Blaxton TA, Bookheimer SY, et al. (15)O water ­positron emission tomography in language localization: a study

46. Kadis DS, Iida K, Kerr EN, et al. Intrahemispheric reorganization of language in children with medically intractable epilepsy of the left hemisphere. J Int Neuropsychol Soc 2007;13(3):505–516 47. Hertz-Pannier L, Chiron C, Jambaqué I, et al. Late plasticity for language in a child’s non-dominant hemisphere: a pre- and postsurgery fMRI study. Brain 2002;125(Pt 2):361–372 48. Bulteau C, Jambaqué I, Chiron C, et al. Language plasticity after hemispherotomy of the dominant hemisphere in 3 patients: Implication of non-linguistic networks. Epilepsy Behav 2017;69:86–94 49. DeSalvo MN, Tanaka N, Douw L, et al. Resting-state functional MR imaging for determining language laterality in intractable epilepsy. Radiology 2016;281(1):264–269 50. Sanai N, Mirzadeh Z, Berger MS. Functional outcome after language mapping for glioma resection. N Engl J Med 2008;358(1):18–27 51. FitzGerald DB, Cosgrove GR, Ronner S, et al. Location of language in the cortex: a comparison between functional MR imaging and electrocortical stimulation. AJNR Am J Neuroradiol 1997;18(8):1529–1539 52. Carpentier A, Pugh KR, Westerveld M, et al. Functional MRI of language processing: dependence on input modality and temporal lobe epilepsy. Epilepsia 2001;42(10):1241–1254 53. Rutten GJ, Ramsey NF, van Rijen PC, Noordmans HJ, van Veelen CW. Development of a functional magnetic resonance imaging protocol for intraoperative localization of critical temporoparietal language areas. Ann Neurol 2002;51(3):350–360 54. Roux FE, Boulanouar K, Lotterie JA, Mejdoubi M, LeSage JP, Berry I. Language functional magnetic resonance imaging in preoperative assessment of language areas: correlation with direct cortical stimulation. Neurosurgery 2003;52(6):1335–1345, discussion 1345–1347 55. Giussani C, Roux FE, Ojemann J, Sganzerla EP, Pirillo D, Papagno C. Is preoperative functional magnetic resonance imaging reliable for language areas mapping in brain tumor surgery? Review of language functional magnetic resonance imaging and direct cortical stimulation correlation studies. Neurosurgery 2010;66(1):113–120 56. 56. de Ribaupierre S, Fohlen M, Bulteau C, et al. Presurgical language mapping in children with epilepsy: clinical usefulness of functional magnetic resonance imaging for the planning of cortical stimulation. Epilepsia 2012;53(1):67–78 57. Rutten GJ, van Rijen PC, van Veelen CW, Ramsey NF. Language area localization with three-dimensional functional magnetic resonance imaging matches intrasulcal electrostimulation in Broca’s area. Ann Neurol 1999;46(3):405–408 58. Skirrow C, Cross JH, Harrison S, et al. Temporal lobe surgery in childhood and neuroanatomical predictors of long-term declarative memory outcome. Brain 2015;138(Pt 1):80–93 59. Jambaqué I, Dellatolas G, Fohlen M, et al. Memory functions following surgery for temporal lobe epilepsy in children. Neuropsychologia 2007;45(12):2850–2862 60. Gleissner U, Sassen R, Lendt M, Clusmann H, Elger CE,­ Helmstaedter C. Pre- and postoperative verbal memory in pediatric patients with temporal lobe epilepsy. Epilepsy Res 2002;51(3):287–296 61. Law N, Benifla M, Rutka J, Smith ML. Verbal memory after temporal lobe epilepsy surgery in children: do only mesial structures matter? Epilepsia 2017;58(2):291–299

20  Functional Magnetic Resonance Imaging in Pediatric Epilepsy Surgery 62. Binder JR. Preoperative prediction of verbal episodic memory outcome using FMRI. Neurosurg Clin N Am 2011;22(2): 219–232, ix

­ iffers between left and right temporal lobe epilepsy patients and d correlates with postsurgical verbal learning decrement. Epilepsy Res 2008;78(2–3):161–170

63. Chiu CY, Schmithorst VJ, Brown RD, Holland SK, Dunn S. Making memories: a cross-sectional investigation of episodic memory encoding in childhood using FMRI. Dev Neuropsychol 2006;29(2):321–340

73. Sidhu MK, Stretton J, Winston GP, et al. Memory fMRI predicts verbal memory decline after anterior temporal lobe resection. Neurology 2015;84(15):1512–1519

64. van den Bosch GE, El Marroun H, Schmidt MN, et al. Brain connectivity during verbal working memory in children and adolescents. Hum Brain Mapp 2014;35(2):698–711 65. Powell HW, Koepp MJ, Richardson MP, Symms MR, Thompson PJ, Duncan JS. The application of functional MRI of ­memory in temporal lobe epilepsy: a clinical review. Epilepsia 2004;45(7):855–863 66. Powell HW, Richardson MP, Symms MR, et al. Preoperative fMRI predicts memory decline following anterior temporal lobe ­resection. J Neurol Neurosurg Psychiatry 2008;79(6):686–693 67. Jokeit H, Okujava M, Woermann FG. Memory fMRI lateralizes temporal lobe epilepsy. Neurology 2001;57(10):1786–1793

74. Chelune GJ. Hippocampal adequacy versus functional reserve: predicting memory functions following temporal lobectomy. Arch Clin Neuropsychol 1995;10(5):413–432 75. Jackson GD, Connelly A, Cross JH, Gordon I, Gadian DG. Functional magnetic resonance imaging of focal seizures. Neurology 1994;44(5):850–856 76. Krakow K, Woermann FG, Symms MR, et al. EEG-triggered functional MRI of interictal epileptiform activity in patients with partial seizures. Brain 1999;122(Pt 9):1679–1688 77. Hamandi K, Salek-Haddadi A, Fish DR, Lemieux L. EEG/functional MRI in epilepsy: The Queen Square Experience. J Clin Neurophysiol 2004;21(4):241–248

68. Detre JA, Maccotta L, King D, et al. Functional MRI lateralization of memory in temporal lobe epilepsy. Neurology 1998;50(4):926–932

78. Tyvaert L, Hawco C, Kobayashi E, LeVan P, Dubeau F, Gotman J. Different structures involved during ictal and interictal epileptic activity in malformations of cortical development: an EEG-fMRI study. Brain 2008;131(Pt 8):2042–2060

69. Rabin ML, Narayan VM, Kimberg DY, et al. Functional MRI predicts post-surgical memory following temporal lobectomy. Brain 2004;127(Pt 10):2286–2298

79. Zijlmans M, Huiskamp G, Hersevoort M, Seppenwoolde JH, van Huffelen AC, Leijten FS. EEG-fMRI in the preoperative work-up for epilepsy surgery. Brain 2007;130(Pt 9):2343–2353

70. Golby AJ, Poldrack RA, Illes J, Chen D, Desmond JE, Gabrieli JD. Memory lateralization in medial temporal lobe epilepsy assessed by functional MRI. Epilepsia 2002;43(8):855–863

80. Centeno M, Tierney TM, Perani S, et al. Combined electroencephalography-functional magnetic resonance imaging and electrical source imaging improves localization of pediatric focal epilepsy. Ann Neurol 2017;82(2):278–287

71. Powell HW, Richardson MP, Symms MR, et al. Reorganization of verbal and nonverbal memory in temporal lobe epilepsy due to unilateral hippocampal sclerosis. Epilepsia 2007;48(8):1512–1525 72. Frings L, Wagner K, Halsband U, Schwarzwald R, Zentner J, Schulze-Bonhage A. Lateralization of hippocampal activation ­

81. Szaflarski JP, Gloss D, Binder JR, et al. Practice guideline summary: use of fMRI in the presurgical evaluation of patients with epilepsy: report of the guideline development, dissemination, and implementation subcommittee of the American Academy of Neurology. Neurology 2017;88(4):395–402

179

21

  Application of Positron Emission Tomography and Single-Photon Emission Computed Tomography in Pediatric Epilepsy Surgery Stephen J. Falchek, Ajay Kumar, and Harry T. Chugani

Summary In recent years, functional neuroimaging has proven a useful adjunct to intracranial electroencephalography (EEG) monitoring, in the identification of epilepsy surgery candidates, and their respective seizure onset zones. The most commonly used modalities for this purpose currently are ictal single-photon emission computed tomography (SPECT) and 2-deoxy-2[18F] fluoro-d-glucose positron emission tomography (FDG-PET). The former images localization-specific cerebral perfusion, and the latter cerebral glucose utilization, as the guiding principles in identification of the seizure onset zone. These techniques aid not only in the identification of a surgical target, but also in visualizing secondary epileptic foci, dual diagnoses, cortical function, postsurgical targeting for second resections, and surgical outcome prognostication. These techniques are helpful in the treatment of a large number of otherwise intractable epilepsies, including infantile spasms, Lennox–Gastaut syndrome, multifocal cortical dysplasias, tuberous sclerosis, Sturge–Weber syndrome, and Rasmussen’s encephalitis. Ictal SPECT can correctly localize the epileptic foci in 70 to 90% of cases with unilateral temporal lobe epilepsy and is also highly useful in extratemporal epilepsies. Registration techniques like statistical parametric mapping (SPM) or subtraction ictal SPECT coregistered to MRI (SISCOM) increase the sensitivity and specificity in identifying the epileptogenic region in up to 95% of cases. Other PET radiotracers include 11C-flumazenil (FMZ), which binds to the alpha-subunits of the gammaaminobutyric acid type A receptor (GABAA)-benzodiazepine receptor, and is useful in imaging mesial temporal lobe epilepsies. 11C-alphamethyl-L-tryptophan (AMT), which t­races tryptophan metabolism, is especially useful in tuberous sclerosis (AMT-PET), whereas FDG-PET often produces false ­ positives. PET utilizing 11C-PK11195, a radioligand that binds specifically to the peripheral-type benzodiazepine receptors, is being explored in the imaging of neuroinflammation, such as the early diagnosis of Rasmussen’s syndrome or other cerebral inflammatory conditions, with or without epilepsy. Newly available PET-MR scanners offer many research and clinical advantages beyond simple coregistration of PET and

MR image volumes, including automated motion correction for PET dynamic studies, single-event sedation when necessary, image-derived arterial input function, and lower radioactivity exposure than PET-CT. Keywords:  ictal SPECT, interictal SPECT, PET, PET-MR, SISCOM, radiotracers, diaschisis, temporal lobe epilepsy, extratemporal epilepsy, tuberous sclerosis, infantile spasms, Rasmussen’s encephalitis, Sturge–Weber syndrome

„„ Introduction Epilepsy is the most common neurological disorder, with a prevalence of 1 to 2% and cumulative lifetime incidence exceeding 3%. Almost 25% of epileptic patients do not respond to multiple antiepileptic treatments and will have intractable (i.e., medically refractory) seizures. These patients can be helped by surgically removing the epileptogenic region of the cerebral cortex. However, to accomplish this, the epileptogenic region has to be precisely delineated before surgery. Indeed, the guiding purpose of presurgical evaluation is the identification of an epileptogenic region amenable to resection that will lead to complete seizure control without unacceptable loss of neurological function. This is not always possible, and sometimes difficult decisions must be made, weighing total seizure control versus loss of some neurological function. I­ ntracranial EEG monitoring (electrocorticography) to capture seizures remains the gold standard for epilepsy surgery planning;1,​2 ­however, functional neuroimaging, such as PET and SPECT, combined with EEG, plays a very important role in defining a surgical area of interest. They provide noninvasive, presurgical localization of potential epileptogenic foci in the patient with no brain lesion on CT or MRI, that is, nonlesional cases, with multiple structural lesions of which only one or two are epileptogenic, or in cases with discordant or inconclusive scalp-surface EEG findings. These technologies can help to guide the progression to surgery, including the accurate placement of intracranial grid and strip electrodes for electrocorticography to confirm the epilepsy focus.

21  Application of PET and SPECT in Pediatric Epilepsy Surgery

„„ Rationale for PET and SPECT SPECT and PET are imaging techniques that use radioisotopes or radiolabeled molecules (the term radiotracer will be used for both of them) to study the perfusion or biochemical function of an organ, even at the cellular or molecular level. With both methods, a very small amount of a selected radiotracer is injected into the patient. The radiotracer is selected on the basis of desired purpose, that is, which organs and what functions the physician is interested in exploring. The gamma rays emitted by the radiotracers are detected externally with the help of suitable detectors, and an image of the spatial distribution of these radiotracers is generated. Therefore, theoretically, the function of any organ can be studied, provided appropriate radiotracers are available. The reason for applying PET and SPECT in epilepsy is based on the fact that metabolism (particularly of glucose), receptor density and neurotransmission, and cerebral blood flow in the epileptic region (as well as in the associated seizure propagation network) are altered and can be detected by these imaging techniques.

„„ Principles and Techniques of PET PET is an imaging technique designed to noninvasively image and measure the function of various organs. In PET, positron-emitting radionuclides, such as 18F, 11C, 15O, and 13N, are used to label various natural biological substrates and drugs or pharmaceuticals. These all contain hydrogen, carbon, oxygen, or nitrogen, which can be replaced with a radioactive positron-emitting species. The resulting radioactive substance, also known as a radiotracer or PET tracer, will have similar behavior to the original compound. Following the same physiological pathways while emitting positrons, they signal their destinations by paired antipodal high-­ energy (511 keV) photons, produced when the shed p ­ ositrons collide with the abundant electrons in the relevant tissue bed. These photons can be detected by external detectors, and the whole process can be traced or imaged. Any physiological or metabolic process, such as glucose metabolism, protein synthesis, enzymatic processes, or receptor–ligand interaction, can be studied using an appropriate radiotracer and various kinetic models describing the biochemical behavior or pathway.3 The most commonly used PET tracer in epilepsy is 2-deoxy-2[18F] FDG (half-life: 110 minutes), which measures

glucose metabolism. FDG is transported in tissue and phosphorylated to FDG-6-phosphate in the same manner as glucose. However, FDG-6-phosphate is not a substrate for the next step of glycolysis. Because it cannot immediately leave the cell, phosphorylated FDG remains trapped within the cell, and its location and quantity can be measured by PET imaging. Under steady-state conditions, FDG uptake reflects the glucose metabolic rate. In the brain, this rate is highly related to the synaptic density and functional activity of the brain tissue. Since brain glucose metabolism undergoes age-related changes, particularly in early childhood, practitioners should be aware of age-specific patterns while interpreting pediatric FDG-PET scans. Quantitatively, metabolic rates of glucose in cerebral cortex at the time of birth are, usually, approximately 30% less than those of adult values. They reach adult values by the second year of life, exceed them by the third year, and reach peak values (almost double the adult values) by 3 to 4 years of age, when a plateau is reached. This plateau persists until approximately 10 years of age, then declines to reach adult values by the age of approximately 16 to 18 years. The glucose metabolism pattern and visual appearance on PET scans also change with age. In the newborn, glucose metabolic activity is most prominent in primary sensory and motor cortex, thalamus, brain stem and cerebellar vermis, cingulate cortex, amygdala, and hippocampus. Between ages 2 and 4 months, glucose utilization increases in parietal, temporal, and primary visual cortex (medial occipital or calcarine cortex), as well as in the basal ganglia and cerebellar hemispheres. Between 6 and 8 months of age, glucose utilization increases in the lateral and inferior prefrontal regions, with medial and dorsal frontal cortex becoming active between 8 and 12 months of age. The adult distribution pattern is seen by 1 year of age.4,​5 Interictal FDG-PET typically shows reduced radiotracer uptake (hypometabolism) in the region typically regarded as the epileptogenic region (Fig. 21.1). The hypometabolism can result from a variety of mechanisms, including neuronal loss, diaschisis, reduction in synaptic density, tonic inhibition of an epileptic focus, or various combinations of these factors. Convincing evidence from three-dimensional brain surface analysis colocalized with intracranial EEG monitoring suggests that the margins of the hypometabolic zone are much more commonly the location of the precise epileptic focus. This is congruent with current understanding of the mechanism of epileptogenic brain lesions, in which the epileptic activity originates from the otherwise normal tissue immediately adjacent

Fig. 21.1  A 2-deoxy-2[18F] fluoro-d-glucose positron emission tomography scan shows an area of hypometabolism in the left frontal lobe (arrows) in an 8-year-old child with intractable seizures and normal MRI. Postsurgical histopathology revealed cortical dysplasia.

181

182

IIc  Preoperative Neuroimaging to the lesion.6 It appears that cortical hypometabolism may be associated with duration, frequency, and severity of the seizures. This may be the reason why hypometabolism is usually found in only one-fourth of children with new-onset epilepsy compared with 80 to 85% of adults with intractable seizures.7 Persistent or increased seizure frequency may lead to enlargement of the hypometabolic area, whereas seizure control may be associated with decrease in the size of hypometabolic cortex or even its resolution.8 The time interval between PET acquisition and the most recent seizure also may affect the extent and severity of the cortical abnormality, with shorter duration having a positive effect on the extent and severity of the hypometabolism.9 In other words, the occurrence of a seizure shortly before the FDG injection may elevate glucose metabolism in an otherwise hypometabolic epileptogenic region and obscure the hypometabolism thus yielding a negative study. The time interval between seizure and FDG administration required to produce this effect is not well understood. During the ictal phase, metabolism increases manyfold in the epileptogenic region; however, because FDG-PET shows cumulative FDG uptake over a period of 30 to 45 minutes (i.e., the phosphorylation of FDG to FDG-6-phosphate), the final images of “ictal studies” can be variable and complex, depending on the exact nature of the underlying pathology, seizure duration, seizure evolution, and net summation effects of ictal, postictal, and interictal metabolism. Therefore, ictal FDG-PET is not very reliable and is often difficult to interpret, and EEG monitoring should be performed during the FDG uptake period to rule out any clinical or subclinical seizure. Because ictal FDG-PET is virtually not done, in all the subsequent discussion, we use the term FDG-PET in place of interictal FDG-PET. Typically, FDG-PET usually shows a larger area of hypometabolism extending beyond the epileptogenic region; therefore, it cannot be reliably used for precise determination of the

surgical margin. However, it can be used for lateralization and general localization of the seizure focus. Furthermore, this information can help identifying areas of interest and therefore targets for subsequent subdural electrode placement, which may be especially useful in cases with normal MRI (Fig. 21.2). The use of more specific tracers, as discussed below, can help in providing a more precise delineation of the epileptogenic tissue; this becomes particularly important in the developing pediatric brain, and when the epileptic focus potentially involves eloquent brain regions (primary motor, speech, or visual areas). Other PET tracers with the potential for detecting epileptic brain regions include FMZ, which binds to alpha-subunits of the GABAA-benzodiazepine receptor, and 11C-alphamethyl­-Ltryptophan (AMT), which provides a measure of tryptophan metabolism. FMZ binding or FMZ volume of distribution (FMZVD), a quantitative measurement incorporating receptor density and binding affinity, is high at 2 years of age and then decreases exponentially with age until adult values are reached at approximately age 20 years. The order of brain region (from highest to lowest FMZ binding) at 2 years of age is as follows: primary visual cortex, superior frontal cortex, medial temporal cortex, temporal lobe, prefrontal cortex, cerebellum, basal ganglia, and thalamus. FMZ binding in medial temporal lobe is robust enough for its good visualization (compared with FDG-PET, in which the medial temporal region, particularly hippocampus, is often not well visualized). This is one of the main reasons for an important role of FMZ-PET in epileptic patients in whom medial temporal lobe abnormalities are suspected. AMT is an analog of tryptophan, which is the precursor for serotonin synthesis. Unlike tryptophan, AMT is not incorporated into protein in significant amounts. The intravenously administered AMT tracer is converted to alpha-methylserotonin (AM-5HT) by tryptophan hydroxylase and accumulates in

Fig. 21.2  Three-dimensional brain rendering of cortical 2-deoxy-2[18F] fluoro-d-glucose (FDG) uptake showing hypometabolism in the left inferior parietal and occipital cortex, extending into the left temporal lobe in an 8-year-old child with uncontrolled seizure and normal MRI. Electroencephalography showed diffuse left-sided epileptiform discharges, mostly coming from temporoparietal cortex. Based on the FDG findings, intracranial electrodes (black circles) were placed over inferior parietal and occipital cortex also, which found most of the epileptiform discharges coming from the occipital region. The rest of the hypometabolic area coincided with the electrodes showing seizure spread.

21  Application of PET and SPECT in Pediatric Epilepsy Surgery neurons and nerve terminals along with the releasable pool of serotonin. The AM-5HT accumulates in the brain for the first 20 minutes (< 2% of the injected dose is present in the brain at peak values), after which a plateau is reached and maintained for up to 60 minutes, with no right–left asymmetry in normal subjects. The half-life of the positron-emitting species greatly influences the logistics and practical application of the PET radiotracer. For example, the 20-minute half-life of 11C requires the presence of an on-site cyclotron to produce it, ­thereby ­limiting access to only a few centers. This has prompted research into the use of other species–tracer combinations, for example, 18 F-flumazenil, which has shown comparable utility to 11C FMZ in localization for temporal lobe epilepsy.10 An AMT analog labeled with 18F is not available yet, and so its use is ­confined to cyclotron-linked PET centers.

„„ Principles and Techniques of SPECT In SPECT, rotating gamma-cameras (detectors) are used to image the distribution of the injected radiotracer in the organ of interest. For this purpose, one to three detectors and only gamma-ray–emitting radioisotopes are used. For the purpose of brain and particularly for epilepsy, SPECT is mostly used to study brain perfusion, with hexamethyl propylene amine oxime (HMPAO) and ethylene cysteine dimer (ECD) labeled with 99mTc being the most common radiotracers used for this purpose. HMPAO readily crosses the blood–brain barrier and approximately 80% is extracted by brain during the first pass. Once inside the neurons and glial cells, HMPAO is oxidized by glutathione into a nondiffusible compound, effectively trapping the radiotracer. A total of 4 to 7% of the injected activity is trapped within the brain, reaching its peak in 1 to 2 minutes. ECD is another lipophilic compound that, like HMPAO, gets trapped inside neurons because of its transformation into a hydrophilic compound that cannot diffuse back. Its first-pass extraction is 60 to 70%, with a maximum of 6 to 7% injected activity accumulating in the brain 1 to 2 minutes after the injection. The brain can be imaged subsequently, under sedation if necessary, and the resulting image provides a snapshot of perfusion immediately after the injection. This is the basis for applying brain SPECT in epilepsy, because the ictal and interictal phases are usually associated with increased or decreased blood flow, respectively, in the epileptic foci. However, during a seizure, the cerebral blood flow changes rapidly with time, depending on the seizure type and its mode of propagation. Therefore, early radiotracer injection is imperative to catch the blood flow changes in the epileptic zone, during seizure. Similarly, knowledge of the exact time of injection (because it takes approximately 20–30 seconds for the radiotracer to reach the brain from an arm vein) and seizure duration is very important for the correct interpretation of the SPECT images. Delayed injection of radiotracer may show a variable pattern of blood flow changes associated with seizure evolution in the epileptic zone and mode and pattern of seizure propagation, depending on the time of injection. Also, because seizure ­propagation usually proceeds from the temporal to the frontal

lobe, or from posterior (parieto-occipital lobes) to anterior cortical regions (temporal and frontal lobe),11,​12,​13,​14 interpretation of ictal SPECT can be very challenging if the exact timing of radiotracer injection and seizure onset is not known. For this reason, synchronization of the radiotracer injection with video-EEG monitoring is the preferred method. During the ictal phase, blood flow in the epileptic region can increase up to 300%, which can be seen as an area of hyperperfusion in ictal SPECT.15 A true ictal SPECT (tracer injected immediately after the onset of the seizure and SPECT images showing perfusion at that point of time) shows an area of hyperperfusion in the epileptogenic region, surrounded by an area of hypoperfusion, which becomes more prominent at the end of the ictal phase. This surrounding area of hypoperfusion may be caused by steal syndrome (shift of adjacent blood flow to the seizure focus), or this area may function as an inhibitory zone trying to limit the seizure spread.16 Interictal SPECT (tracer is injected when patient is not having any clinical or subclinical seizure and SPECT images show the baseline perfusion pattern) shows hypoperfusion or normal perfusion in the epileptogenic region. Even when present, hypoperfusion may be very mild and sometimes difficult to distinguish from the surrounding normal brain on visual examination. The main role of interictal SPECT currently is to provide a baseline blood flow and assist in the evaluation of ictal SPECT, visually or quantitatively, using SPM or SISCOM, i.e., the interictal SPECT images are subtracted from the ictal images, and the results are displayed on coregistered MR images. Use of these registration techniques can increase the sensitivity and specificity of ictal SPECT. Studies have shown that SPM can increase the sensitivity of SPECT scan over visual analysis.17,​18,​19 However, lack of agematched control subjects can make its use difficult, particularly in children younger than 6 years of age.20 SISCOM, ­conversely, appears to be very useful in children, particularly in extratemporal nonlesional epilepsy.1,​2 Use of SISCOM can help in revisiting and detecting subtle changes in MRI, which were initially reported normal.21 Studies have shown that the area of the resected SISCOM abnormality is associated with surgical outcome; a larger area of SISCOM abnormality resection corresponds to better outcomes.22,​23 However, conflicting information suggests that, while complete resection of the zone of hyperperfusion on SISCOM yields favorable postoperative seizure outcomes,1,​24 complete resection of the SISCOM focus is not always obtained in cases with a good outcome.1 The most favorable surgical outcomes still are obtained with complete resection of areas displaying abnormality on electrocorticography, even though studies have shown clear s­ urgical failure in cases where the entire epileptogenic zone was thought to have been resected by this measure.1 Such ­surgical failures are probably related to the limitations of electrocorticography, and the fact that potentially epileptogenic cortex (often “secondary” foci) may remain silent during the recordings, only to declare their epileptogenic capability after the resection. The major limitation of SPECT in the pediatric population lies in the difficult acquisition of good interictal or ictal brain SPECT. This is a consequence of the very frequent or brief seizure patterns (such as infantile spasms or myoclonic epilepsy) that commonly occur in children. Another limitation is the poor spatial resolution (10–15 mm) of SPECT, compared with FDGPET (~ 5–6 mm), which becomes even more crucial in small pediatric brains.

183

184

IIc  Preoperative Neuroimaging

„„ The Role of PET and SPECT in Pediatric Epilepsy Surgery The role of PET and SPECT can be summarized as follows: 1. Detection of epileptogenic cortex. 2. Determination of dual pathology (i.e., medial temporal involvement). 3. Assessment for secondary epileptic focus. 4. Evaluation of the functional status outside the ­epileptogenic zone. 5. Evaluation of eloquent cortex. 6. Postsurgical evaluation.

Epileptogenic Region Temporal Lobe Epilepsy In temporal lobe epilepsy, identified regions of interictal glucose hypometabolism are imprecise indicators of both presumed temporal epileptogenic zones and brain tissue bearing pathological changes. These hypometabolic areas usually extend beyond temporal structures to ipsilateral parietal and frontal cortex, as well as to the thalamus, and also occasionally to the contralateral temporal lobe.25,​26,​27,​28 Although this may represent the epileptic network involved in seizure propagation, and may be related to behavioral and neuropsychological changes seen with chronic epilepsy, extratemporal hypometabolic cortical regions should warrant further investigation for intrinsic epileptogenicity associated with a secondary focus. FMZ-PET is highly sensitive in temporal lobe epilepsy and shows decreased FMZ binding in the sclerotic hippocampus,26,​29 and the reduction in FMZ binding is usually more than that expected solely by the loss of hippocampal volume.30,​31 Contrary to FDG-PET, which usually shows extratemporal hypometabolism in parietal and frontal cortex in cases of temporal lobe epilepsy (probably associated with cognitive dysfunction or reflecting diaschisis), decreased FMZ binding usually represents neuronal loss or receptor changes related to epileptogenicity and, therefore, should be more closely scrutinized.26 The sensitivity of FMZ-PET in detection of unilateral hippocampal sclerosis has been reported to be up to 100% with contralateral abnormalities in one-third of patients.10,​26,​29 In MRI-negative patients, FMZ-PET has been found to be abnormal in up to 85% patients with temporal lobe epilepsy.32,​33,​34 Indeed, FMZ-PET appears to be more sensitive than FDG-PET in identifying an epileptogenic region and is associated with better surgical outcome, even when the MRI is normal. Use of SPM can further increase the accuracy of FMZ-PET, with detection of subtle changes in FMZ binding, which is difficult to appreciate visually.26,​35 These SPM studies sometimes found increased FMZ binding also,36,​37,​38,​39 which, in some cases, indicated cortical developmental malformations.37 Use of SPM also revealed increased FMZ binding in the normal-appearing temporal lobe white matter, which was found to harbor areas of microdysgenesis on histopathological examination.40 This is an interesting finding, because these ectopic neuronal clusters may lead to epileptogenesis by providing an aberrant circuitry. Another PET tracer, AMT, does not appear to be very useful in cases of medial temporal lobe epilepsy, p ­ articularly with hippocampal sclerosis.

In temporal lobe epilepsy, the pattern of perfusion on ictal SPECT depends on the origin of the seizures. Patients with medial temporal lesions usually show well-localized areas of hyperperfusion involving ipsilateral medial and lateral temporal lobe, but patients with lateral temporal lesions typically show bilateral hyperperfusion with higher increase in the ipsilateral side.41,​42 This can be explained on the basis of different mechanisms of seizure propagation depending on the neuronal connectivity. The timing of tracer injection is also important in the case of temporal lobe epilepsy, because ictal, postictal, and peri-ictal SPECT scans have different perfusion patterns, which also depend on the area of temporal lobe involved. In the case of medial temporal lobe epilepsy, ictal SPECT scans (tracer injection within 20 seconds of the seizure onset) typically will show hyperperfusion of the entire medial temporal lobe, along with surrounding hypoperfusion of either the orbital cortex or the entire frontal lobe. Peri-ictal scans (slightly delayed injection, 20–60 seconds after the seizure onset) will show hyperperfusion in lateral temporal cortex, orbital cortex, basal ganglia, and also contralateral temporal lobe, probably because of rapid seizure propagation. Postictal scans (injection within 4 minutes after the end of a seizure) will show persistent hyperperfusion in the medial temporal lobe, with hypoperfusion in the lateral temporal lobe, gradually extending to surrounding hyperperfused areas. The medial temporal lobe remains isoperfused for another 10 to 15 minutes, and then gradually becomes hypoperfused, resembling the interictal scan. Extensive published experience regarding SPECT is available for adults; pediatric data are still emerging. Overall, i­ nterictal SPECT has very low sensitivity (< 50%) in the detection of epileptogenic regions in pediatric temporal lobe epilepsy, ­ with false-positive or false-negative findings in 20 to 75% ­cases. However, ictal SPECT can correctly localize the ­epileptic foci in 70 to 90% of cases with unilateral temporal lobe ­epilepsy.43​—​46 As previously mentioned, various registration techniques, such as SPM or SISCOM, can further increase the sensitivity and specificity of ictal SPECT. In children, SISCOM was found to be helpful in identifying the epileptogenic region in up to 95% of cases.21 Compared with intracranial EEG ­f indings, ictal SPECT was found to correctly localize the seizure-­onset zone in 80% of children with intractable epilepsy.47 False localization was caused by rapid seizure propagation or subclinical ­seizure onset. Additionally, in the majority (70%) of ­children with favorable outcome from resective epilepsy surgery, the surgical margin coincided with the SPECT focus. Postictal SPECT is also more sensitive (70–90%) than interictal SPECT and can improve further with use of SISCOM.48,​49 SISCOM can be particularly useful in cases of dysembryoplastic neuroepithelial tumor (DNET), which is more prevalent in the pediatric ­population. SISCOM can demonstrate some additional dysplastic areas around the DNET, and removal of these areas is essential for better surgical outcomes. The use of SISCOM can increase the focus detection rate up to 93%, compared with 74% without it.3,​21,​50

Extratemporal Lobe Epilepsy FDG-PET can play an important role in the presurgical evaluation of extratemporal lobe epilepsy in the pediatric population by providing important lateralizing and localizing information

21  Application of PET and SPECT in Pediatric Epilepsy Surgery that will guide intraoperative electrode placements (Fig. 21.3). Because frontal lobe epilepsy in young children is usually associated with subtle structural changes not apparent in MRI, such as cortical dysplasia or heterotopias, FDG-PET is more informative in children compared with adults. Even in cases of abnormal MRI, FDG-PET can often be very useful. It may show hypometabolism extending beyond the lesion. These areas should be sampled with intracranial EEG during surgery, because perilesional cortex also may be epileptogenic, and lesionectomy alone may not be enough to achieve seizure freedom. In frontal lobe epilepsy, the sensitivity of FDG-PET in localizing the epileptogenic zones is in the range of 45 to 73%.51,​52,​53,​54,​55 However, using a high-resolution PET scanner, we found a sensitivity of 92% and a specificity of 62.5% of FDG-PET in the detection of the epileptogenic region in children with frontal lobe epilepsy.51

In cases of occipital lobe epilepsy, a lower localization value of FDG-PET has been reported.56 In neocortical epilepsy, FMZ-PET has been reported to have 60 to 100% sensitivity compared with intracranial ictal EEG.57​—​60 FMZ abnormalities usually extend beyond the lesions in an eccentric fashion. However, these extensions are usually smaller than the large perilesional hypometabolism seen with FDG-PET and show good correlation with intracranial EEG findings of epileptiform activity.57,​61,​62 In extratemporal lobe epilepsy, FMZ-PET is abnormal in 70% of patients with normal MRI scans.33 FMZ-PET appears to be more sensitive than FDGPET in identifying an epileptogenic region; complete resection of the FMZ abnormality is associated with excellent surgical outcome, even when the MRI is normal (Fig. 21.4).59,​63 Use of SPM can further increase the usefulness of FMZ-PET in

Fig. 21.3  Three-dimensional brain rendering of fluoro-d-glucose (FDG) uptake, showing glucose hypometabolism extending beyond the structural lesion. Intracranial electroencephalography monitoring revealed most of the epileptiform discharges emanating from adjacent to the lesion.

Fig. 21.4  Three-dimensional brain rendering of flumazenil (FMZ) binding (left) and 2-deoxy-2[18F] fluoro-d-glucose (FDG) uptake (right) showing much smaller area of reduced FMZ binding compared with extensive area of reduced FDG uptake in the left temporal and frontal lobe.

185

186

IIc  Preoperative Neuroimaging patients with neocortical epilepsy, including those with or without n ­ ormal MRI.38,​64 Another PET tracer, AMT, appears to have strong clinical applications in selected cases of extratemporal lobe ­epilepsy. This molecule is a substrate for serotonin synthesis and kynurenine pathways. In epileptogenic regions, AMT uptake may be increased, depending on the mechanism underlying the epilepsy. It has proven to be particularly useful in tuberous sclerosis, in non-TS patients with histologically verified cortical dysplasia type 2b, and in polymicrogyria/heterotopias cases.65,​66 While the sensitivity of AMT uptake is about 70%, its specificity for the epileptic focus, as verified by intracranial EEG monitoring, is nearly 100%. AMT uptake is much less altered in cases of nonspecific gliosis.63 It is, similarly, much less effective at identifying the epileptic focus in milder cortical malformations, types 1a and 2a cortical dysplasias, and subependymal heterotopias.65 As in the case of FMZ binding abnormalities, the area of increased AMT uptake is significantly more restricted than the extent of corresponding glucose hypometabolism. In some instances, AMT-PET can identify the epileptogenic cortex even when FDG and FMZ-PET scans are normal. SPECT can provide a basis for the placement of intracranial electrodes in extratemporal lobe epilepsy. However, because of the very short duration of some seizures in children, it may be difficult to acquire an ictal SPECT. Postictal scans are also less helpful because, unlike temporal lobe epilepsy, seizure-­ induced perfusion changes do not always extend into the postictal phase. Although ictal SPECT generally is not as s­ uccessful in ­ extratemporal lobe epilepsy localization as in temporal lobe epilepsy, an accuracy rate of 70% has been reported ­nevertheless.67 Again, the use of SISCOM can further increase its ­localizing value up to 93%.22,​68

study of 140 children with infantile spasms found unifocal and multifocal cortical metabolic abnormalities in 95% of children with an initial diagnosis of cryptogenic infantile spasms.71 In cases of intractable spasms and a single focal PET abnormality, corresponding to the EEG focus, resective surgery can be planned with not only good seizure control but also complete or partial reversal of associated developmental delay. When the pattern of glucose hypometabolism is generalized and symmetric, a lesional etiology is not likely, and neurometabolic or neurogenetic d ­ isorders should be considered in further evaluation and management. PET findings in infants with spasms also suggest complex cortical–subcortical interactions. Prominent glucose metabolism in lenticular nuclei and brain stem, believed to be important in the secondary generalization of focal cortical discharges resulting in spasms, accounts for the bilateral motor involvement and r­elative symmetry of the majority of spasms, even in the presence of a discrete focal lesion.72 Glucose metabolism patterns appear to be significantly correlated with outcomes of infantile spasms. In a series of 23 patients73 with cryptogenic West’s syndrome and normal MRI, imaged with paired FDG-PET at diagnosis and then at age 10 months, and evaluated for seizure outcome and developmental status 13 to 21 years later, the first PET scan did not correlate with seizure or developmental outcomes. However, 5 of 7 patients (71%) with hypometabolism on the second PET subsequently showed intellectual impairment, and 13 of 16 (81%) with normal metabolism had normal intelligence. Similarly, 71% with hypometabolism on the second PET had persistent seizures, and 88% with normal metabolism were seizure free.73

Infantile Spasms

FDG-PET shows hypometabolism in tubers, both epileptic and nonepileptic; therefore, it is not very useful as such in the evaluation of children with tuberous sclerosis. However, as previously discussed, AMT can be used to differentiate epileptogenic from nonepileptogenic tubers in children with tuberous sclerosis, because it shows increased AMT uptake interictally in only ­epileptogenic tubers (Fig. 21.5).

PET can play an important role in the evaluation of infantile spasms. For example, on PET scanning of glucose metabolism, most infants diagnosed with cryptogenic infantile spasms will have focal or multifocal cortical hypometabolism, corresponding to the areas of ictal and interictal EEG abnormalities.69,​70 A

Tuberous Sclerosis

Fig. 21.5  MRI (a) showing multiple tubers in a child with tuberous sclerosis and intractable seizures with the majority of seizures coming from the right frontal lobe. Whereas 2-deoxy-2[18F] fluoro-d-glucose positron emission tomography (FDG-PET) showed glucose hypometabolism in tubers (b), ­­alphamethyl-L-tryptophan PET scan showed intense uptake in a right frontal tuber (c), corresponding to the electroencephalography focus.

21  Application of PET and SPECT in Pediatric Epilepsy Surgery In epileptogenic tubers, there is increased uptake and subsequent intracellular accumulation of the AMT because of activation of the kynurenine pathway,74 which leads to the production of neurotoxic and convulsant metabolites, such as quinolinic acid.75 AMT-PET can identify epileptogenic tuber(s) in almost two-thirds of children with tuberous sclerosis and intractable epilepsy.74,​76,​77 Although the specificity of AMT-PET is very high, its sensitivity is suboptimal and appears to be related to the underlying pathology as well as the method of image analysis. In patients with tuberous sclerosis and intractable epilepsy, MRI-based quantitative assessment increases the sensitivity of AMT-PET to 79% from 44.4% with visual assessment.78 This is because nonepileptogenic tubers typically show decreased AMT uptake, and some epileptogenic tubers that show relatively increased AMT uptake cannot be easily differentiated from adjacent normal cortex without quantitative analysis. Good correlation also exists between resection of epileptogenic tubers suggested by AMT-PET and seizure outcome, with 12 of 14 patients who underwent complete resection of the AMT focus experiencing class I outcomes.79 Tubers with at least 10% increase of AMT uptake compared to normal-­ appearing cortex were all found to be epileptogenic. A cutoff threshold of 1.02 for AMT uptake ratio provided 83% a ­ ccuracy for ­detecting tubers that should be resected to achieve a ­seizure-free outcome.76,​79 Additional information gleaned from the coregistration of AMT with MRI or FDG scans has shown that tubers with AMT uptake equal to or slightly above normal cortex (uptake ratio of 0.98 or greater) are significantly associated with epileptogenic foci, keeping in mind that nonepileptic tubers show lower AMT uptake than normal cortex. This analysis and conclusion would not be possible without the PET-MRI coregistration, as such tubers would not be distinguishable from normal cortex on AMT-PET alone.76 In contrast to AMTPET, our studies using FMZ-PET have not found binding differences between epileptogenic and nonepileptogenic tubers (unpublished data). SPECT also may play some role in identifying epileptic tubers. One small study, in 15 children, found a good correlation between ictal SPECT and ictal scalp EEG.80 SISCOM may increase the yield in defining epileptogenic tubers for surgical resection; thus, individual case reports and small series of both children and adults with tuberous sclerosis report 66 to 100% seizure freedom when applying SISCOM-guided localization.81,​82

Lennox–Gastaut Syndrome Children with Lennox–Gastaut syndrome (a triad of multiple seizure types including tonic seizures, developmental delay, and 1- to 2.5-Hz generalized “slow” spike and wave EEG pattern), may have four metabolic patterns on FDG-PET scan: unilateral focal, unilateral diffuse, and bilateral diffuse hypometabolism, as well as normal patterns.83,​84,​85 Interictal SPECT usually shows multiple areas of hypoperfusion.86 Patients with unilateral focal and unilateral diffuse patterns may be

occasionally considered for cortical resection, provided there is concordance between PET and ictal EEG findings.

Sturge–Weber Syndrome In children with Sturge–Weber syndrome, FDG-PET reveals hypometabolism ipsilateral to the facial nevus and u ­ sually identifies additional areas of abnormal cortex extending ­ beyond the lesion visible on MRI.87,​88 However, infants may show a paradoxical pattern of increased glucose metabolism interictally in the cortex underlying the leptomeningeal angioma; as the disease progresses, the hypermetabolic area becomes hypometabolic.87 In some patients, serial FDG-PET scans show rapidly progressing and severe hypometabolism in the affected area, probably because of rapid demise of the brain tissue associated with the angioma; these patients will have improvement in seizure status and cognitive function and therefore may not require surgical intervention. However, the presence of transient hypermetabolism also seems to increase the likelihood of a surgical intervention; therefore, it may serve as a marker of a more protracted course of disease.89 Early and rapid progression in unilateral cases of Sturge–Weber syndrome leads to e ­ arly and more efficient reorganization in the contralateral ­cortex. Conversely, persistent mild hypometabolism of the lesion may indicate ongoing functional disturbance, inefficient reorganization, and these patients may show persistent seizures and developmental arrest.90 These are the patients who require surgical intervention for seizure control and possible cognitive improvement by forcing effective reorganization in the contralateral hemisphere, while brain plasticity is still at a maximum during development. In Sturge–Weber syndrome, detrimental metabolic compromise often occurs before 3 years of age,91 coinciding with a sharp increase in developmentally regulated cerebral metabolic demand associated with programmed synaptic proliferation.5 Progressive hypometabolism is associated with high seizure frequency in these children. However, metabolic abnormalities may remain limited or even partially recover later in some children with well-controlled seizures. Metabolic recovery accompanied by neurological improvement suggests a window for therapeutic intervention in children with unilateral Sturge– Weber syndrome.88 Perfusion SPECT shows hyperperfusion in some lesions even before seizure onset,91 analogous to the transient hypermetabolism seen on PET.87,​88 After 1 year of age, these areas t­ ypically show hypoperfusion.92 In a single case report on a patient with failed functional hemispherectomy, ictal SPECT showed ­hyperperfusion in the residual lesion with falsely lateralized epileptiform discharges on EEG; further surgery resulted in ­seizure freedom.93

Rasmussen’s Syndrome and Other Epilepsies of Inflammatory Origin Neuroinflammation may be the underlying cause for intractable epilepsy in some cases, such as in chronic focal encephalitis of Rasmussen’s syndrome. Neuroinflammation is mediated

187

188

IIc  Preoperative Neuroimaging by activated microglia, which secrete several proinflammatory molecules such as cytokines (interleukin-1 [IL-1], IL-6, tumor necrosis factor-alpha [TNF-alpha]), chemokines (macrophage inflammatory protein-1 [MIP-1] alpha, beta, monocyte chemoattractant protein-1 [MCP-1]) and neurotoxins, free radicals, nitric oxide, proteinases, eicosanoids, and excitotoxins, which may play an important role in epileptogenesis. Although the exact mechanisms are unclear, it appears that the inflammatory mediators act by increasing glutamatergic neurotransmission, decreasing GABA-mediated currents and inducing neovascularization, and damaging the blood–brain barrier. Detection of microglia was previously not possible with radiological methods or biochemical techniques but required histopathological examination of central nervous system tissues, which is quite invasive or possible only postmortem. Because activated microglia express peripheral-type benzodiazepine receptors (translocator protein), they can be imaged with PET using 11C-PK11195, a radioligand that binds specifically to the peripheral-type benzodiazepine receptors, thus making the in vivo imaging of neuroinflammation possible. PET scanning using 11C-PK-11195 can help in the early diagnosis of Rasmussen’s syndrome or other inflammatory conditions of the brain with or without epilepsy, where CT and MRI are often normal for several months after the clinical manifestation of the disease, and subsequently show only nonspecific abnormalities such as atrophy. Localization of the most affected brain regions may also provide a guide in deciding the site of brain biopsy (if indicated) to avoid sampling errors and guide in the surgical removal of that region.94

Other Epilepsies of Hemispheric Origin In addition to Rasmussen’s encephalitis, patients with conditions such as hemimegalencephaly, pre- and postnatal stroke, and widespread malformations of cortical development often benefit from partial resection and complete disconnection of one cortical hemisphere, termed functional hemispherectomy or hemispherotomy. FDG-PET has been demonstrated to be useful in predicting surgical outcomes (including developmental outcome), based on the presence or absence of bilateral metabolic abnormalities. Of a cohort of 35 patients, 100% of those with unilateral FDG abnormalities were seizure free, versus only 75% of those with bilateral abnormality.95

Dual Pathology Undiagnosed dual pathology (coexistence of neocortical lesion and hippocampal sclerosis) can be the reason for surgical failure in epilepsy surgery, because resection of both the cortical lesion and the affected hippocampus is necessary to achieve seizure freedom. Although FDG-PET can be used to evaluate for an abnormally functioning hippocampus, the higher ­sensitivity of FMZ-PET makes it more useful in such cases, particularly when the MRI does not reveal an abnormal signal or atrophy in the hippocampus. This can be very useful because resection of both the lesion itself and the secondarily affected hippocampus is necessary for good surgical ­outcome.96

Secondary Epileptic Foci “Secondary” epileptic foci have been defined by Morrell97,​98 as “transsynaptic and long-lasting alterations in nerve cell behavior characterized by paroxysmal electrographic manifestations and clinical seizures” induced by seizures from a primary epileptic focus. The secondary epileptic focus is usually located at a different site from the primary focus along the path of seizure propagation. Histopathological examination of these secondary foci usually shows gliosis without the pathology of the primary focus.99 FMZ-PET can play a very important role in the ­detection of secondary epileptic foci, as shown by our experience.57,​99 If the secondary focus has reached independence, complete removal of both primary and secondary foci is required to achieve the best s­ urgical results. In contrast, “dependent” ­secondary foci rely on the primary focus and need not be resected. The c­ hallenge in chronic epilepsy is to be able to differentiate between “­dependent” and “independent” ­secondary e ­ pileptic foci.

Functional Status of the Rest of the Brain FDG-PET can be very valuable in assessing the integrity of brain regions outside of the epileptogenic zone during the surgical planning process in order to provide prognostic information. The presence of nonepileptogenic dysfunctional areas indicates poor substrate for reorganization of functions affected by surgical resections. In children with hemimegalencephaly, FDG-PET often shows additional less-pronounced abnormalities in the opposite hemisphere, which probably accounts for the suboptimal cognitive outcome even with complete seizure control after surgical removal of the profoundly abnormal hemisphere. Thus, FDG-PET can be useful in such cases to assess the functional integrity of the contralateral hemisphere before hemispherectomy and help predict cognitive outcome. In children with epilepsy, the extent of glucose hypometabolism appears to be correlated with overall intellectual ability, and its lateralization with performance in subtests of intellectual capacity. In 78 children, given the Chinese Wechsler Intelligence Scale for Children, the extent of hypometabolism on FDG-PET negatively correlated with fullscale IQ. The verbal/performance discrepancy scores (VIQPIQ) also differed significantly between left hemisphere and right hemisphere hypometabolism groups; left hemisphere hypometabolism showed a negative difference (i.e., performance IQ greater than verbal IQ), and the right a positive discrepancy.100

Eloquent Cortex Although 15O-water PET has been used in the past to assess eloquent cortex (e.g., motor and language cortex), this type of ­evaluation has been, for the most part, replaced by ­functional MRI. The latter can be repeated as indicated and offers the advantage of no radiation exposure. As a result, PET or SPECT has a very limited role in activation studies for ­presurgical evaluation of epileptic patients. Epileptic subjects with ­ implanted metallic devices, which are not MRI-compatible and who require noninvasive functional brain mapping, may be candidates for 15O-water PET activation studies.

21  Application of PET and SPECT in Pediatric Epilepsy Surgery

Postsurgical Evaluation In surgical failures where a second surgery is being considered, very few neuroimaging options are available to pinpoint remaining epileptic tissue. Occasionally, ictal SPECT may be helpful in this regard but, in most cases, the epileptologist is left with the scalp EEG and seizure semiology to guide placement of intracranial electrodes. In one study, SISCOM revealed a localized area of hyperperfusion in almost 80% of patients undergoing reoperation, and in 70% of cases they were concordant with EEG findings.101 Resection of these concordant lesions led to good surgical outcome. AMT-PET can be particularly useful in this setting and, unlike MRI, interictal FDG or FMZ-PET, AMT-PET can differentiate epileptogenic cortex from nonepileptic tissue damage caused by the initial surgery in approximately half of the cases.102 The best results are obtained if the scan is performed between 2 months and 2 years after the first surgery. However, more work is required in this difficult group of patients undergoing reoperation. Finally, FDG-PET can be useful in the postsurgical evaluation and monitoring of previously hypometabolic nonepileptogenic brain regions (which had no EEG correlate) in the presurgical FDG scan. Resolution of this hypometabolism, which occurs in some instances, will suggest the functional nature of the suppression. Conversely, persistence of a remote, but connected, region or appearance of a new area of hypometabolism (not suspected to be diaschisis) may suggest a potential secondary epileptic focus.

„„ Experience with Other PET Tracers in Epilepsy Clinical experience with opiate, histamine, N-methyl-D-aspartate (NMDA), acetylcholine, dopamine, and other neuroreceptor PET tracers in epilepsy is limited. The data available thus far show either increased binding [11C-carfentanil (μ-opioid receptor agonist), 11C-methylnaltrindole (δ-opioid receptor antagonist), 18F-cyclofoxy (μ,k-opioid receptor antagonist),11C-doxepin (H1 receptor antagonist), 11C-L-deprenyl (MAO-B inhibitor)] or decreased binding [11C-diprenorphine (μ,δ,K-opioid receptor antagonist), 11C/18F-FCWAY (5HT1A receptor antagonist), 18 F-MPPF (5HT1A receptor antagonist), 18F-altanserin (5HT2A receptor antagonist), 11C-(S)-[N-methyl]-ketamine (NMDA receptor antagonist), 11C-NMBP (mAch receptor antagonist), 123 I-iododexetimide (mAch receptor antagonist)]. However, because of paucity of data, their current role in presurgical evaluation for intractable epilepsy is not yet established.

„„ Multimodality Imaging with PET-MR Both MRI and PET are powerful imaging tools used to study ­structure and function in pediatric epilepsy surgery, yet they

offer complementary imaging perspectives. The simultaneous use of these tools in a single setting with recently available PET-MR scanners offers many research and clinical advantages (Fig. 21.6). This new technology is a significant advancement to biomedical research, because it allows simultaneously acquired noninvasive and quantitative comparisons of anatomical, functional, and metabolic tissue measures. Indeed, the fundamental advantages of simultaneous acquisition of PET and MR data reach far beyond simple coregistration of PET and MR image volumes. Important issues for children include automated motion correction for PET dynamic studies using MR navigator sequences, both structural and functional image acquisition at the same time requiring only one sedation (if necessary), image-derived arterial input function thus eliminating the need for arterial blood sampling for quantitative modeling, and lower radioactivity exposure, as the need for a transmission CT for attenuation correction used by current PET-CT technology is eliminated. When using the PET-MR, scanning time is dictated more by MRI acquisition (~ 40 minutes depending on desired sequences) than PET (~ 10 minutes). We have taken advantage of this, particularly in children, by reducing the FDG dose by more than 50% and increasing PET acquisition time to equal that required for the MRI. In epilepsy surgery patients, we routinely include diffusion tensor imaging (DTI) in order to study the location of motor or language fibers in relation to the PET and MRI abnormalities (Fig. 21.7) in order to minimize damage to these vital regions during resection. PET-MR is also rapidly becoming an important research tool in the study of ­epilepsy by allowing a powerful approach to study anatomical–functional relationships, or dual functional relationships, for example, MR spectroscopy or f­ unctional MRI in relation to PET molecular imaging.

Fig. 21.6  GE SIGNA PET-MR 3.0T. This scanner has the most sensitive PET detection system available on the market today; in addition, its MR capabilities are state of the art. Simultaneous acquisition of PET and MR signals using novel tracers and novel MR imaging sequences will allow the translation of innovative new diagnostics for pediatric neurological disorders, including epilepsy.

189

190

IIc  Preoperative Neuroimaging

Fig. 21.7  (a) PET brain image of a 10.5-year-old girl with intractable epilepsy. The red ROI was drawn manually to indicate the left cingulate cortex hypometabolism; MRI anatomical images were normal. (b) Image showing in red is the hypometabolism in three-dimensions and its relative location to diffusion tensor imaging-based motor pathways (blue is for leg, green is for finger, and magenta is for mouth/lip).

„„ Conclusion SPECT and PET have proven to be substantial adjuncts to intracranial EEG recording in the effective treatment of epilepsy surgery patients. Current technology allows for precise identification of seizure-onset zones, secondary epileptic foci, dual diagnoses, cortical function, postsurgical targeting for s­econd resections, and surgical outcome prognostication.

These ­techniques have enabled the treatment of a large number of o ­therwise intractable epilepsies, including infantile spasms, Lennox–Gastaut syndrome, multifocal cortical dysplasias, tuberous sclerosis, Sturge–Weber syndrome, and Rasmussen’s encephalitis. Work in progress will allow noninvasive identification of neuroinflammation, among other potential conditions. As this technology becomes more widely available, an even greater proportion of patients will experience major improvements in the treatment of epilepsy.

References 1. Kudr M, Krsek P, Marusic P, et al. SISCOM and FDG-PET in patients with non-lesional extratemporal epilepsy: correlation with intracranial EEG, histology, and seizure outcome. Epileptic Disord 2013;15(1):3–13

8. Benedek K, Juhász C, Chugani DC, Muzik O, Chugani HT. Longitudinal changes in cortical glucose hypometabolism in children with intractable epilepsy. J Child Neurol 2006; 21(1):26–31

2. Perissinotti A, Setoain X, Aparicio J, et al. Clinical role of subtraction ictal SPECT coregistered to MR Imaging and 18F-FDG PET in pediatric epilepsy. J Nucl Med 2014;55(7):1099–1105

9. Bouvard S, Costes N, Bonnefoi F, et al. Seizure-related shortterm plasticity of benzodiazepine receptors in partial epilepsy: a [11C]flumazenil-PET study. Brain 2005;128(Pt 6):1330–1343

3. Carson RE. Precision and accuracy considerations of physiological quantitation in PET. J Cereb Blood Flow Metab 1991;11(2):A45–A50

10. Vivash L, Gregoire MC, Lau EW, et al. 18F-flumazenil: a γ­-­aminobutyric acid A-specific PET radiotracer for the localization of drug-resistant temporal lobe epilepsy. J Nucl Med 2013;54(8):1270–1277

4. Chugani HT, Phelps ME. Maturational changes in cerebral function in infants determined by 18FDG positron emission tomography. Science 1986;231(4740):840–843 5. Chugani HT, Phelps ME, Mazziotta JC. Positron emission tomography study of human brain functional development. Ann Neurol 1987;22(4):487–497 6. Juhász C, Chugani DC, Muzik O, et al. Is epileptogenic cortex truly hypometabolic on interictal positron emission tomography? Ann Neurol 2000;48(1):88–96 7. Gaillard WD, Kopylev L, Weinstein S, et al. Low incidence of abnormal (18)FDG-PET in children with new-onset partial epilepsy: a prospective study. Neurology 2002;58(5):717–722

11. Lee SK, Yun CH, Oh JB, et al. Intracranial ictal onset zone in nonlesional lateral temporal lobe epilepsy on scalp ictal EEG. Neurology 2003;61(6):757–764 12. Noachtar S, Arnold S, Yousry TA, Bartenstein P, Werhahn KJ, Tatsch K. Ictal technetium-99m ethyl cysteinate dimer single-photon emission tomographic findings and propaga­ tion of epileptic seizure activity in patients with extratemporal ­epilepsies. Eur J Nucl Med 1998;25(2):166–172 13. Williamson PD, Boon PA, Thadani VM, et al. Parietal lobe epilepsy: diagnostic considerations and results of surgery. Ann Neurol 1992;31(2):193–201

21  Application of PET and SPECT in Pediatric Epilepsy Surgery 14. Williamson PD, Thadani VM, Darcey TM, Spencer DD, Spencer SS, Mattson RH. Occipital lobe epilepsy: clinical characteristics, seizure spread patterns, and results of surgery. Ann Neurol 1992;31(1):3–13

34. Lamusuo S, Pitkänen A, Jutila L, et al. [11 C]Flumazenil binding in the medial temporal lobe in patients with temporal lobe epilepsy: correlation with hippocampal MR volumetry, T2 relaxometry, and neuropathology. Neurology 2000;54(12):2252–2260

15. Hougaard K, Oikawa T, Sveinsdottir E, Skinoj E, Ingvar DH,­ Lassen NA. Regional cerebral blood flow in focal cortical ­epilepsy. Arch Neurol 1976;33(8):527–535

35. Bouilleret V, Dupont S, Spelle L, Baulac M, Samson Y, Semah F. Insular cortex involvement in mesiotemporal lobe ­epilepsy: a positron emission tomography study. Ann Neurol 2002;51(2):202–208

16. Prince DA, Wilder BJ. Control mechanisms in cortical epileptogenic foci. “Surround” inhibition. Arch Neurol 1967;16(2):194–202 17. Bruggemann JM, Som SS, Lawson JA, Haindl W, Cunningham AM, Bye AM. Application of statistical parametric mapping to SPET in the assessment of intractable childhood epilepsy. Eur J Nucl Med Mol Imaging 2004;31(3):369–377 18. Lee JD, Kim HJ, Lee BI, Kim OJ, Jeon TJ, Kim MJ. Evaluation of ictal brain SPET using statistical parametric mapping in temporal lobe epilepsy. Eur J Nucl Med 2000;27(11):1658–1665 19. Tae WS, Joo EY, Kim JH, et al. Cerebral perfusion changes in mesial temporal lobe epilepsy: SPM analysis of ictal and interictal SPECT. Neuroimage 2005;24(1):101–110 20. Muzik O, Chugani DC, Juhász C, Shen C, Chugani HT. Statistical parametric mapping: assessment of application in children. Neuroimage 2000;12(5):538–549 21. Chiron C, Véra P, Kaminska A, et al. [Ictal SPECT in the epileptic child. Contribution of subtraction interictal images and superposition of with MRI]. [in French] Rev Neurol (Paris) 1999;155(6–7):477–481 22. O’Brien TJ, So EL, Mullan BP, et al. Subtraction ictal SPECT co-­ ­ registered to MRI improves clinical usefulness of SPECT in localizing the surgical seizure focus. Neurology 1998;50(2):445–454 23. O’Brien TJ, So EL, Mullan BP, et al. Subtraction peri-ictal SPECT is predictive of extratemporal epilepsy surgery outcome. Neurology 2000;55(11):1668–1677 24. Kudr M, Krsek P, Maton B, et al. Ictal SPECT is useful in localizing the epileptogenic zone in infants with cortical dysplasia. Epileptic Disord 2016;18(4):384–390 25. Gaillard WD, Bhatia S, Bookheimer SY, Fazilat S, Sato S, Theodore WH. FDG-PET and volumetric MRI in the evaluation of patients with partial epilepsy. Neurology 1995;45(1):123–126 26. Henry TR, Frey KA, Sackellares JC, et al. In vivo cerebral metabolism and central benzodiazepine-receptor binding in temporal lobe epilepsy. Neurology 1993;43(10):1998–2006 27. Swartz BE, Halgren E, Delgado-Escueta AV, et al. Neuroimaging in patients with seizures of probable frontal lobe origin. Epilepsia 1989;30(5):547–558 28. Van Bogaert P, Massager N, Tugendhaft P, et al. Statistical parametric mapping of regional glucose metabolism in mesial temporal lobe epilepsy. Neuroimage 2000;12(2):129–138 29. Koepp MJ, Richardson MP, Brooks DJ, et al. Cerebral benzodiazepine receptors in hippocampal sclerosis. An objective in vivo analysis. Brain 1996;119(Pt 5):1677–1687 30. Koepp MJ, Labbé C, Richardson MP, et al. Regional ­hippocampal [11C]flumazenil PET in temporal lobe epilepsy with unilateral and bilateral hippocampal sclerosis. Brain 1997;120(Pt 10):1865–1876 31. Stanišić M, Coello C, Ivanović J, et al. Seizure outcomes in relation to the extent of resection of the perifocal fluorodeoxyglucose and flumazenil PET abnormalities in anteromedial temporal lobectomy. Acta Neurochir (Wien) 2015;157(11):1905–1916 32. Koepp MJ, Hammers A, Labbé C, Woermann FG, Brooks DJ,­ Duncan JS. 11C-flumazenil PET in patients with r­efractory temporal lobe epilepsy and normal MRI. Neurology 2000;54(2):332–339 33. Koepp MJ, Woermann FG. Imaging structure and function in refractory focal epilepsy. Lancet Neurol 2005;4(1):42–53

36. Hammers A, Koepp MJ, Labbé C, et al. Neocortical abnormalities of [11C]-flumazenil PET in mesial temporal lobe epilepsy. Neurology 2001;56(7):897–906 37. Richardson MP, Friston KJ, Sisodiya SM, et al. Cortical grey matter and benzodiazepine receptors in malformations of ­ ­cortical development. A voxel-based comparison of structural and functional imaging data. Brain 1997;120(Pt 11):1961–1973 38. Richardson MP, Koepp MJ, Brooks DJ, Duncan JS. 11C-flumazenil PET in neocortical epilepsy. Neurology 1998;51(2):485–492 39. Richardson MP, Koepp MJ, Brooks DJ, Fish DR, Duncan JS. Benzodiazepine receptors in focal epilepsy with corti­ cal dysgenesis: an 11C-flumazenil PET study. Ann Neurol 1996;40(2):188–198 40. Hammers A, Koepp MJ, Hurlemann R, et al. Abnormalities of grey and white matter [11C]flumazenil binding in temporal lobe epilepsy with normal MRI. Brain 2002;125 (Pt 10):2257–2271 41. Ho SS, Berkovic SF, McKay WJ, Kalnins RM, Bladin PF. Temporal lobe epilepsy subtypes: differential patterns of cerebral perfusion on ictal SPECT. Epilepsia 1996;37(8):788–795 42. Ho SS, Newton MR, McIntosh AM, et al. Perfusion patterns during temporal lobe seizures: relationship to surgical outcome. Brain 1997;120(Pt 11):1921–1928 43. Benifla M, Otsubo H, Ochi A, et al. Temporal lobe surgery for intractable epilepsy in children: an analysis of outcomes in 126 children. Neurosurgery 2006;59(6):1203–1213, discussion 1213–1214 44. Harvey AS, Bowe JM, Hopkins IJ, Shield LK, Cook DJ, Berkovic SF. Ictal 99mTc-HMPAO single photon emission computed tomography in children with temporal lobe epilepsy. Epilepsia 1993;34(5):869–877 45. Lee JJ, Kang WJ, Lee DS, et al. Diagnostic performance of 18F-FDG PET and ictal 99mTc-HMPAO SPET in pediatric temporal lobe epilepsy: quantitative analysis by statistical ­parametric mapping, statistical probabilistic anatomical map, and ­subtraction ictal SPET. Seizure 2005;14(3):213–220 46. Rowe CC, Berkovic SF, Austin MC, McKay WJ, Bladin PF. ­ Patterns of postictal cerebral blood flow in temporal lobe epilepsy: qualitative and quantitative analysis. Neurology ­ 1991;41(7):1096–1103 47. Kaminska A, Chiron C, Ville D, et al. Ictal SPECT in children with epilepsy: comparison with intracranial EEG and relation to postsurgical outcome. Brain 2003;126(Pt 1):248–260 48. O’Brien TJ, So EL, Mullan BP, et al. Subtraction SPECT co-­ registered to MRI improves postictal SPECT localization of ­seizure foci. Neurology 1999;52(1):137–146 49. Rowe CC, Berkovic SF, Austin M, McKay WJ, Bladin PF. Postictal SPET in epilepsy. Lancet 1989;1(8634):389–390 50. Valenti MP, Froelich S, Armspach JP, et al. Contribution of SISCOM imaging in the presurgical evaluation of temporal lobe epilepsy related to dysembryoplastic neuroepithelial tumors. Epilepsia 2002;43(3):270–276 51. da Silva EA, Chugani DC, Muzik O, Chugani HT. Identification of frontal lobe epileptic foci in children using positron emission tomography. Epilepsia 1997;38(11):1198–1208 52. Henry TR, Sutherling WW, Engel J Jr, et al. Interictal cerebral metabolism in partial epilepsies of neocortical origin. Epilepsy Res 1991;10(2–3):174–182

191

192

IIc  Preoperative Neuroimaging 53. Kim YK, Lee DS, Lee SK, Chung CK, Chung JK, Lee MC. (18)F-FDG PET in localization of frontal lobe epilepsy: comparison of visual and SPM analysis. J Nucl Med 2002;43(9):1167–1174

73. Natsume J, Maeda N, Itomi K, et al. PET in infancy predicts longterm outcome during adolescence in cryptogenic West syndrome. AJNR Am J Neuroradiol 2014;35(8):1580–1585

54. Lee SK, Lee SY, Kim KK, Hong KS, Lee DS, Chung CK. Surgical outcome and prognostic factors of cryptogenic neocortical epilepsy. Ann Neurol 2005;58(4):525–532

74. Chugani DC, Muzik O, Chakraborty P, Mangner T, Chugani HT. Human brain serotonin synthesis capacity measured in vivo with alpha-[C-11]methyl-L-tryptophan. Synapse 1998; 28(1):33–43

55. Swartz BW, Khonsari A, Vrown C, Mandelkern M, Simpkins F, Krisdakumtorn T. Improved sensitivity of 18FDG-positron emission tomography scans in frontal and “frontal plus” epilepsy. Epilepsia 1995;36(4):388–395 56. Patil S, Biassoni L, Borgwardt L. Nuclear medicine in pediatric neurology and neurosurgery: epilepsy and brain tumors. Semin Nucl Med 2007;37(5):357–381 57. Juhász C, Chugani DC, Muzik O, et al. Electroclinical correlates of flumazenil and fluorodeoxyglucose PET abnormalities in lesional epilepsy. Neurology 2000;55(6):825–835 58. Muzik O, da Silva EA, Juhasz C, et al. Intracranial EEG versus flumazenil and glucose PET in children with extratemporal lobe epilepsy. Neurology 2000;54(1):171–179 59. Ryvlin P, Bouvard S, Le Bars D, et al. Clinical utility of flumazenil-PET versus [18F]fluorodeoxyglucose-PET and MRI in refractory partial epilepsy. A prospective study in 100 patients. Brain 1998;121(Pt 11):2067–2081 60. Savic I, Thorell JO, Roland P. [11C]flumazenil positron emission tomography visualizes frontal epileptogenic regions. Epilepsia 1995;36(12):1225–1232 61. Arnold S, Berthele A, Drzezga A, et al. Reduction of benzodiazepine receptor binding is related to the seizure onset zone in extratemporal focal cortical dysplasia. Epilepsia 2000;41(7):818–824 62. Szelies B, Sobesky J, Pawlik G, et al. Impaired benzodiazepine receptor binding in peri-lesional cortex of patients with ­symptomatic epilepsies studied by [(11)C]-flumazenil PET. Eur J Neurol 2002;9(2):137–142 63. Juhász C, Chugani DC, Muzik O, et al. Alpha-methyl-L-­tryptophan PET detects epileptogenic cortex in children with ­intractable epilepsy. Neurology 2003;60(6):960–968 64. Richardson MP, Hammers A, Brooks DJ, Duncan JS. Benzodiazepine-GABA(A) receptor binding is very low in ­dysembryoplastic neuroepithelial tumor: a PET study. Epilepsia 2001;42(10):1327–1334 65. Chugani HT, Kumar A, Kupsky W, Asano E, Sood S, Juhász C. Clinical and histopathologic correlates of 11C-alpha-methylL-­tryptophan (AMT) PET abnormalities in children with intractable epilepsy. Epilepsia 2011;52(9):1692–1698

75. Stone TW. Kynurenines in the CNS: from endogenous obscurity to therapeutic importance. Prog Neurobiol 2001;64(2):185–218 76. Asano E, Chugani DC, Muzik O, et al. Multimodality imaging for improved detection of epileptogenic foci in tuberous sclerosis complex. Neurology 2000;54(10):1976–1984 77. Fedi M, Reutens DC, Andermann F, et al. alpha-[11C]-­ Methyl-L-tryptophan PET identifies the epileptogenic tuber and correlates with interictal spike frequency. Epilepsy Res 2003;52(3):203–213 78. Juhász C, Chugani DC, Asano E, et al. Alpha[11C]methylL-­tryptophan positron emission tomography scanning in 176 patients with intractable epilepsy. Ann Neurol 2002:S118 (abstract) 79. Kagawa K, Chugani DC, Asano E, et al. Epilepsy surgery outcome in children with tuberous sclerosis complex evaluated with alpha-[11C]methyl-L-tryptophan positron emission tomography (PET). J Child Neurol 2005;20(5):429–438 80. Koh S, Jayakar P, Resnick T, Alvarez L, Liit RE, Duchowny M. The localizing value of ictal SPECT in children with tuberous sclerosis complex and refractory partial epilepsy. Epileptic Disord 1999;1(1):41–46 81. Aboian MS, Wong-Kisiel LC, Rank M, Wetjen NM, Wirrell EC, Witte RJ. SISCOM in children with tuberous sclerosis ­complex-related epilepsy. Pediatr Neurol 2011;45(2):83–88 82. Song P, Joo EY, Seo DW, Hong SB. Seizure localization in patients with multiple tubers: presurgical evaluation in tuberous sclerosis. J Epilepsy Res 2012;2(1):16–20 83. Chugani HT, Mazziotta JC, Engel J Jr, Phelps ME. The ­Lennox— Gastaut syndrome: metabolic subtypes determined by 2-­ deoxy-2[18F]fluoro-D-glucose positron emission tomography. Ann Neurol 1987;21(1):4–13 84. Iinuma K, Yanai K, Yanagisawa T, et al. Cerebral glucose metabolism in five patients with Lennox—Gastaut syndrome. Pediatr Neurol 1987;3(1):12–18 85. Theodore WH, Rose D, Patronas N, et al. Cerebral glucose metabolism in the Lennox—Gastaut syndrome. Ann Neurol 1987;21(1):14–21

66. Kumar A, Asano E, Chugani HT. α-[11C]-methyl-L-tryptophan PET for tracer localization of epileptogenic brain regions: clinical studies. Biomarkers Med 2011;5(5):577–584

86. Heiskala H, Launes J, Pihko H, Nikkinen P, Santavuori P. Brain perfusion SPECT in children with frequent fits. Brain Dev 1993;15(3):214–218

67. Hwang SI, Kim JH, Park SW, et al. Comparative analysis of MR imaging, positron emission tomography, and ictal single-­photon emission CT in patients with neocortical epilepsy. AJNR Am J Neuroradiol 2001;22(5):937–946

87. Chugani HT, Mazziotta JC, Phelps ME. Sturge—Weber syndrome: a study of cerebral glucose utilization with positron emission tomography. J Pediatr 1989;114(2):244–253

68. Véra P, Kaminska A, Cieuta C, et al. Use of subtraction ictal SPECT co-registered to MRI for optimizing the localization of seizure foci in children. J Nucl Med 1999;40(5):786–792 69. Chugani HT, Shewmon DA, Shields WD, et al. Surgery for intractable infantile spasms: neuroimaging perspectives. Epilepsia 1993;34(4):764–771 70. Chugani HT, Shields WD, Shewmon DA, Olson DM, Phelps ME, Peacock WJ. Infantile spasms: I. PET identifies focal cortical dysgenesis in cryptogenic cases for surgical treatment. Ann Neurol 1990;27(4):406–413

88. Juhász C, Batista CE, Chugani DC, Muzik O, Chugani HT. ­Evolution of cortical metabolic abnormalities and their clinical correlates in Sturge—Weber syndrome. Eur J Paediatr Neurol 2007;11(5):277–284 89. Alkonyi B, Chugani HT, Juhász C. Transient focal cortical increase of interictal glucose metabolism in Sturge—­ Weber syndrome: implications for epileptogenesis. Epilepsia 2011; 52(7):1265–1272 90. Lee JS, Asano E, Muzik O, et al. Sturge—Weber syndrome: correlation between clinical course and FDG PET findings. ­ ­Neurology 2001;57(2):189–195

71. Chugani HT, Conti JR. Etiologic classification of infantile spasms in 140 cases: role of positron emission tomography. J Child ­Neurol 1996;11(1):44–48

91. Pinton F, Chiron C, Enjolras O, Motte J, Syrota A, Dulac O. Early single photon emission computed tomography in Sturge—Weber syndrome. J Neurol Neurosurg Psychiatry 1997;63(5):616–621

72. Chugani HT, Shewmon DA, Sankar R, Chen BC, Phelps ME. Infantile spasms: II. Lenticular nuclei and brain stem activation on positron emission tomography. Ann Neurol 1992;31(2):212–219

92. Chiron C, Raynaud C, Tzourio N, et al. Regional cerebral blood flow by SPECT imaging in Sturge—Weber disease: an aid for d ­iagnosis. J Neurol Neurosurg Psychiatry 1989;52(12): 1402–1409

21  Application of PET and SPECT in Pediatric Epilepsy Surgery 93. Bilgin O, Vollmar C, Peraud A, la Fougere C, Beleza P, Noachtar S. Ictal SPECT in Sturge—Weber syndrome. Epilepsy Res 2008;78(2–3):240–243

99. Juhász C, Asano E, Shah A, et al. Focal decreases of cortical GABAA receptor binding remote from the primary seizure focus: what do they indicate? Epilepsia 2009;50(2):240–250

94. Kumar A, Chugani HT, Luat A, Asano E, Sood S. Epilepsy surgery in a case of encephalitis: use of 11C-PK11195 positron emission tomography. Pediatr Neurol 2008;38(6):439–442

100. Tang Z, Chen Z, Zhai Q, Hao Y, Zhang Y, Zeng X. Correlation between interictal cerebral glucose hypometabolism and IQ in children with epilepsy. Epilepsy Behav 2014;31(2):15–18

95. Traub-Weidinger T, Weidinger P, Gröppel G, et al. Presurgical evaluation of pediatric epilepsy patients prior to hemispherotomy: the prognostic value of 18F-FDG PET. J Neurosurg Pediatr 2016;25(6):683–688

101. Wetjen NM, Cascino GD, Fessler AJ, et al. Subtraction ictal single-photon emission computed tomography coregis­ tered to magnetic resonance imaging in evaluating the need for repeated epilepsy surgery. J Neurosurg 2006;105(1): 71–76

96. Juhász C, Nagy F, Muzik O, Watson C, Shah J, Chugani HT. [11C] Flumazenil PET in patients with epilepsy with dual pathology. Epilepsia 1999;40(5):566–574 97. Morrell F. Secondary epileptogenesis in man. Arch Neurol 1985;42(4):318–335 98. Morrell F, deToledo-Morrell L. From mirror focus to secondary epileptogenesis in man: an historical review. Adv Neurol 1999;81:11–23

102. Juhász C, Chugani DC, Padhye UN, et al. Evaluation with alpha-[11C]methyl-L-tryptophan positron emission tomography for reoperation after failed epilepsy surgery. Epilepsia 2004;45(2):124–130

193

 

22

  Multimodality Imaging and Coregistration Prashin C. Unadkat, Walid Ibn Essayed, John M. K. Mislow, and Alexandra J. Golby

Summary Multimodal imaging has developed into a key component in the presurgical evaluation of medically intractable epilepsy. This chapter reviews newer imaging techniques including advanced structural and functional imaging, intraoperative imaging, and modern coregistration tools, focusing on their impact on current surgical management of seizures in pediatric patients. Keywords:  coregistration, fMRI, DTI, intractable epilepsy, surgical planning, MEG, PET, SPECT

„„ Introduction In the presurgical evaluation of medically intractable epilepsy, knowledge of the precise location of the seizure focus is of paramount importance. Despite advances in imaging technologies, it is difficult to evaluate data from multiple structural and functional imaging modalities in an integrated fashion. Thus, clinicians may not be able to synthesize all the available data in the most effective fashion. As such, coregistration of multiple modalities has proven to be an invaluable tool in the armamentarium of the surgical team and has led to an increase in the number of patients offered surgery for epilepsy in the United States.1 In addition, as coregistration has improved neurosurgical efficacy, newer imaging techniques have improved presurgical lesion detection as well as demonstrating abnormalities in patients who previously might be considered nonlesional, such as hippocampal atrophy or sclerosis, cortical dysplasia, or small vascular lesions or low-grade tumors.2

„„ Coregistration Direct data fusion technique superimposes multiple images and presents the surgeon with a coarse approximation of the location of functional regions. This may be adequate when the surgical target is single, discrete, and in non-eloquent territory. However, if the target is multifocal, indistinct, or in eloquent cortex, an understanding of the correlation between structure and function is very helpful. Coregistration is the process of volumetrically fusing processed functional images onto a structural image by using imaging data shared by all images3,​4 and as such offers a significant

advantage over direct image overlay. Image-to-image coregistration commonly relies on the identification of mutual points in both images known as tie-points. As such, software packages designed to coregister images, such as statistical parametric mapping (SPM), FreeSurfer, 3D Slicer, AFNI, and Caret, commonly utilize an automated, area-based technique for identifying image tie-points. Coregistration software executes either a linear or nonlinear transformation; linear transformation is technically easier and faster than nonlinear transformation but yields significantly lower-quality image coregistration.5 Initially, functional data derived from positron emission computed tomography (PET) and single-photon emission computed tomography (SPECT) would be volumetrically coregistered with a standardized stereotaxic brain atlas such that of Talairach and Tournoux,6,​7 but variations in gross morphology and microstructure of the human brain8 make an atlas-based methodology unreliable in most presurgical planning for individual patients, particularly for cortical lesions.9 Thus, clinical coregistration generally implies that all data involved in volumetric fusion are derived entirely from the individual patient. One of the first efforts to address the challenge of image data fusion for surgical planning was Mountz and colleagues’ description of the fusion of SPECT and CT images of patients’ brains as a method of providing an accurate and noninvasive method for correlating function (in the form of blood flow) and neuroanatomy (Fig. 22.1).10 The authors demonstrated lesion or pathology in clinical examples of tumor, developmental abnormality (autism), cerebrovascular disease, and notably in epilepsy via injection of tracer during ictal phase of seizure.10 The arrival of MRI heralded clinicians’ ability to define neuroanatomical targets with high precision. As such, MRI replaced CT for surgical planning, and MRI-PET coregistration (Fig. 22.2) eclipsed PET-CT coregistration as standard of care in presurgical planning and clinical evaluation. Evidence of this ­evolution was borne out by Viñas and colleagues’ study of MRIPET coregistration in presurgical planning for epilepsy surgery in eloquent cortex.11 Twelve patients underwent preoperative MRI-PET coregistration with motor, visual, and language mapping and subsequently underwent an awake craniotomy with MRI-assisted image guidance and intraoperative cortical stimulation or visual evoked potentials. The researchers found that PET was reliable in identifying most (but not all) motor, visual, and language cortex, leading to their conclusion that MRI-PET coregistration is a useful tool for identification of eloquent cortex in the setting of neurosurgical preplanning.11 Although the authors concluded that intraoperative cortical stimulation and

22  Multimodality Imaging and Coregistration

Fig. 22.1 Ictal (a) and interictal (b) brain SPECT. Quantification of the coregistered images would help delineate areas of significant hyperperfusion between the two scans. The area of focal hyperperfusion seen in the left anterior frontal lobe lateral to the old resection cavity on ictal SPECT may suggest area of seizure focus.

Fig. 22.2  PET-MRI coregistration demonstrating left temporal hypometabolism, which can prove vital into lateralizing and localizing a seizure focus in nonlesional epilepsy. (Reproduced with permission from Dr. Laura Horky.)

visual evoked potential remained the gold standard for cortical mapping,11 further refinements of SPECT and MRI have given this type of coregistration increased precision and reliability in presurgical planning for epilepsy.12,​13 With further refinements, MRI-PET coregistration has proven to be clinically valuable in patients outside the purview of medically refractory epilepsy: by differentiating actual

hypoperfusion from artefactual hypoperfusion resulting from partial volume effects and to improve the accuracy of asymmetry indices in interictal patients, the use broadened to patients with partial epilepsy.12 Another challenge during the infancy of coregistration was speed—significant offline processing and m ­ anual ­registration work were necessary to achieve acceptably a ­ccurate­

195

196

IIc  Preoperative Neuroimaging ­coregistered MRI-PET images.14 By combining a multiresolution approach with an automatic segmentation of input image volumes into areas of interest and background, Cízek and colleagues demonstrated that with suitable preprocessing the time to coregister PET and MRI images with robust accuracy could be reduced 10-fold.15 The next step in coregistration was the combination of MRI, CT, and PET coregistration to assist in image-guided placement of subdural electrodes, and then using the information gathered by the subdural electrodes (combined with the precise location of the electrodes by CT) to aid in surgical resection of the suspected epileptogenic focus in patients undergoing imageguided surgical treatment of epilepsy.16,​17,​18,​19 The advent of readily available digital cameras has ushered in an even higher level of certainty in epilepsy surgery; as coregistration of digital photographs of the brain cortex with the results of 3D MRI data sets is now possible,20,​21 it allows for identification of anatomical details underlying the subdural grid electrodes and

enhances the intraoperative certainty and precision of the neurosurgeon (Fig. 22.3). In addition, surgical navigation systems found in most modern neurosurgical operating theaters allow for coregistration of preoperative data sets to intraoperative head position of the patient.22 The clinical effectiveness of multimodal coregistration has been borne out in several studies. Localizing the subdural strip and grid electrodes as well as depth electrodes with respect to the surrounding eloquent structures is essential for presurgical planning of the margin of resection. After electrode placement, one of the challenges in using the preoperative MRI as a frame of reference is the soft tissue deformation, which may render the results inaccurate. One method to combat this issue has been to use postoperative images such as MRI or CT scan after electrode implantation, although these may be rendered less than optimal due to the susceptibility artifact on the postoperative MRI and the inferior soft tissue contrast on the postoperative CT.23 Taimouri and colleagues24 were able to over-

Fig. 22.3  Intracranial subdural electrode implantation with prolonged extraoperative monitoring: (a) demonstrates the cortex segmented from the preoperative T2-weighted MRI (c) with the subdural grid electrodes segmented from the postoperative CT (d). The blue dots represent each electrode and the pink dots represent the electrodes with recordings showing sharp or spike waves. A diagrammatic representation (b) of the results of extraoperative electrophysiology illustrating electrodes that showed face and language function. “Brain shift” between the pre- (c) and postimplantation (d) of electrodes, respectively. LLT, left lateral temporal; LPST, left posterior subtemporal; LMST, left middle subtemporal; LAST, left anterior subtemporal. Refer to electrode grid nomenclature. (Reproduced with permission from Dr. Page Pennell and Alexandra Golby.)

22  Multimodality Imaging and Coregistration come these issues by projecting the position of the electrodes extracted from the postoperative CT onto a coregistered three-­ dimensional geometric model of the cortical surface extracted from the preoperative MRI. The accuracy of this method was confirmed by two-dimensional intraoperative photographs ­ of ­electrode placement and demonstrated spatial accuracy of 1.31 ± 0.69 mm.24 On the other hand, one study found that patients with multimodal images coregistered were less likely to undergo invasive electroencephalography (EEG) mapping due to better focus localization and had better postoperative outcomes.25 Another study found that PET-MRI coregistration was better at localizing the epileptogenic zone than PET alone. PET-MRI coregistration found an area of hypometabolism in 80% of the patients with normal MRI findings, and these areas were found to be highly concordant with the seizure-onset zone identified on EEG.26 Computer-aided subtraction ictal SPECT coregistered to the MRI (SISCOM) improved the sensitivity and specificity of SPECT to localize the seizure focus and found that concordance between SISCOM localization and resection site to be predictive of postoperative improvement which was not seen on SPECT localization alone.27 In 2004, Murphy and colleagues evaluated the outcomes of 22 patients selected to undergo multimodal coregistration (PET-MRI, SPECT-MRI, or fluid-attenuated inversion recovery [FLAIR] MRI) for presurgical planning due the following reasons: no lesion visible on conventional MRI sequences, ­multiple lesions, or one very large lesion that could not be completely r­esected without the risk of significant postoperative ­morbidity.1 ­Another group with lesions within eloquent cortex was also included in the study, and underwent further coregistration with subdural electrocorticography grids.1,​16 After an average of 27 months’ postsurgical follow-up, the authors found that 77% of the patients had excellent outcomes for their seizures, and 86% had favorable outcomes. One patient suffered a permanent major deficit, while three patients suffered permanent minor deficits. Since these patients would have been poor surgical candidates without presurgical multimodal coregistration, this study provides a strong argument that this approach may allow the reevaluation of patients previously denied surgery due to the selection criteria outlined by the paper’s authors. The findings of Murphy et al1 were confirmed in 2007 when Doelken et al28 studied 49 temporal lobe epilepsy (TLE) patients with coregistration of MRI, MRI spectroscopy, and SPECT, and compared the imaging results to that of traditional noninvasive EEG video monitoring and evaluated lateralization of affected hemisphere with regard to bilateral affection and postoperative outcome. The authors found that EEG and MRI had a high concordance in establishing unilaterality or bilaterality for TLE and epileptogenic focus, and under the circumstances of ambiguous laterality on MRI or EEG, SPECT and MR spectroscopy aided in identifying an epileptogenic focus, and if so, if the lesion was unilateral or bilateral. As a result, the study demonstrated that multimodal imaging for epilepsy helps identify bilateral involvement, which is important to identify patients who are unlikely to benefit from epilepsy surgery.

„„ Newer Imaging Technologies Magnetic Resonance Imaging Ultra-high-field MRI (7 tesla) with its strong tissue contrast and higher spatial resolution provides greater anatomical and pathological detail, especially in patients with focal epilepsy and hippocampal sclerosis; however, its usefulness in a clinical setting and particularly for surgical planning still needs to be investigated.29,​30,​31

Functional Magnetic Resonance Imaging Functional magnetic resonance imaging (fMRI) has been used in surgical epilepsy patients to lateralize and localize language function32,​33 and to lateralize memory function in medial temporal lobe (MTL) epilepsy34 as well as for localization of motor function in pediatric patients with symptomatic focal epilepsy.35 After image acquisition, some offline processing is still required. This appears to have been overcome by Kesavadas and colleagues36 by using real-time fMRI instead of off-line analysis in a 10-patient study, where a comparison of real-time fMRI and offline processing using analysis software, SPM, was performed for sensorimotor, language, and visual paradigms. Significant concordance between the two techniques was noted, effectively demonstrating that real-time fMRI could be performed easily and effectively for presurgical evaluation of pediatric epilepsy.36 Numerous FDA-approved software packages are available for purchase that can monitor head motion in real time, achieve interim image analysis, process fMRI or diffusion tensor imaging (DTI) data simultaneously, transfer this to picture archiving and communication system (PACS), and output data for integration into neuronavigation systems, yet these software packages can be expensive to purchase.37 The combination of EEG and fMRI was found to be helpful in preoperative evaluation of surgical candidates as well as detection of nonlesional frontal lobe epilepsy.38,​39 EEG or fMRI was also found to yield better outcomes in patients who had areas with interictal epileptic discharges (IED)-related blood oxygen level dependent (BOLD) responses resected during surgery when compared to those that had these areas intact.40 Although in the early stages, the use of various analysis techniques using resting state fMRI has been used to identify potential epileptogenic networks41 as well as impaired functional networks.42,​43

Diffusion Tensor Imaging DTI is an imaging technology that offers great promise in reducing morbidity and mortality in neurosurgery. By evaluating the motion of water at the voxel level via quantitative measures of diffusion and fractional anisotropy, DTI can provide data on the structural integrity of brain tissue.3,​44 This radiological evaluation of the orientation of the preferential diffusion of water can create images of major white matter pathways in the brain, and thus, can deduce the structural basis of cerebral networks.3,​44,​45,​46

197

198

IIc  Preoperative Neuroimaging DTI can also be reliably informative about white matter integrity in pediatric epilepsy patients.47 In a study by Carlson and colleagues47 found that white matter diffusivity variables like mean diffusivity, radial diffusivity, and axial diffusivity, as well as fractional anisotropy have low measurement variability and high inter- and intrarater reliability. DTI has also been useful in predicting risk of postoperative deficit. In a study by James and colleagues48 found that the distance of the anterior limit of the Meyer’s loop to the temporal pole could help predict the risk of visual field defect after anterior temporal lobectomy and thus can be factored into treatment decision- making.

Magnetoencephalography Magnetoencephalography (MEG) is another emerging imaging technology in the field of epilepsy. In MEG, arrays of superconducting quantum interference devices (SQUIDs) detect the slight magnetic fields (10-12 tesla) generated by intraneuronal currents of the human brain in real time.49 MEG offers a direct measurement neural electrical activity with high temporal resolution (< 1 ms) but relatively low spatial resolution.49 An advantage of coregistering MEG data over EEG data with MRI or fMRI is absence of magnetic field distortion and attenuation by conductivities between scalp and EEG electrodes.49 Recent clinical data from RamachandranNair and colleagues50 indicate that presurgical planning with MEG yielded good prediction of what patients would be appropriate surgical candidates, as postoperative seizure freedom was less likely to occur in children with bilateral MEG dipole clusters or only scattered dipoles. In addition, the authors demonstrated that MEG may be accompanied by EEG data to determine what patients may or may not be good surgical candidates, as seizure freedom in the clinical study was most likely to occur when there was concordance between EEG and MEG localization and least likely to occur when these results were divergent.50 One study found that, when compared to intracranial EEG (ICEEG), clustered spike sources from MEG recordings were highly accurate albeit with lower sensitivity in identifying the irritative zone of the ICEEG. However, MEG was also found to be poorly correlative with ICEEG in identifying the seizure-onset zone, which maybe a potential limitation.51

Intraoperative Imaging Intraoperative MRI within the surgical suite provides realtime acquisition of MRI without moving the patient, online image-guided stereotaxy without preoperative imaging, and real-time tracking of instruments in the operative field registered to the MR images.52 High-resolution intraoperative MRI has been shown to improve surgical outcomes in glioma surgery and has found some usefulness in lesional epilepsy.53,​54 However, relatively high costs and a demanding surgical environment have somewhat restricted its widespread use. Intraoperative MRI helps the surgeon compensate for the architectural distortion from “brain shift” after craniotomy and lesion resection, and with ­newer

imaging algorithms, coregistered preoperative images such as fMRI, PET, and DTI can be altered and reregistered with the new intraoperative MRI to negotiate eloquent territories.55,​56,​57 A study involving 415 patients found intraoperative MRI combined with neuronavigation to be beneficial and led to favorable seizure outcomes especially in those patients with epilepsyassociated tumors.58 On the other hand, intraoperative ultrasound offers a low-cost solution, which is minimally disruptive to the surgical workflow. With the ability to navigate the ultrasound probe, the ultrasound images can be coregistered to the preoperative multimodal images and can assist in correcting for “brain shift.”59,​60,​61 Besides “brain shift” correction, ultrasound has been used to delineate focal epileptogenic lesions and guide hemispherectomy; convincing evidence for its applications in epilepsy surgery is still lacking.62,​63 Intraoperative CT has been shown to improve the accuracy of intracranial electroencephalography by guiding the placement of subdural and depth electrodes, decreasing the need for revision surgery and improving the precision of intracranial electroencephalography.64 Implantation of responsive neurostimulation devices may also benefit from intraoperative CT by improving accuracy and allowing for intraoperative target correction and conformation.65 Lastly, using intraoperative MRI thermometry, laser interstitial thermal therapy (LITT) procedures have shown to be an effective and safe treatment option in focal epilepsy as well as mesial temporal lobe epilepsy.66,​67

Future Perspectives With increasing information available to the clinician, integration and visualization of these multimodal images for optimal decision-making, during the pre- and intraoperative period is a topic of great interest. With increased computing power, various interactive visualization platforms as well as augmented and virtual reality systems have been proposed and may help to improve the quality of care, although its overall impact on clinical care remains largely unexplored.68,​69,​70

„„ Conclusion Accuracy and precision remain the hallmarks of neurosurgery, and recent advances in image guidance technology represent a considerable asset to the armamentarium of tools at the disposal of the neurosurgeon. With the advent of image coregistration and introduction of newer imaging modalities, multiple images including intraoperative MRI, fMRI, CT, PET, SPECT, DTI, EEG, and MEG can all be volumetrically fused to paint an extraordinarily intricate and precise portrait of a patient’s functional neuroanatomy (Fig. 22.4). By collecting both functional and anatomical connectivity data via high-resolution neuroimaging methods such as MRI, fMRI, MEG, EEG, and DTI, the neurosurgeon may now visualize the three-dimensional structure of the patient’s brain and evaluate what pathways may be disrupted or displaced by a lesion (sclerosis from TLE, tumor, etc.).3,​22,​71 All these imaging

22  Multimodality Imaging and Coregistration

Fig. 22.4  DTI-fMRI coregistration: DTI (a, b) illustrates the corticospinal tract and fMRI represents hand clench task (yellow); (c, d) illustrate the arcuate fasciculus, and fMRI represents a sentence completion task (blue). Coregistration of these two modalities demonstrates that it is the corticospinal tract of the hand region and arcuate fasciculus between language-related cortex, illustrating that the combination of image modalities is more useful than either modality alone.

­ odalities, particularly functional imaging, have demonstrated m great potential in improving the diagnostic and prognostic outcome in epilepsy patients. While advances in current imaging techniques (e.g., 3+ tesla MRI, faster image acquisition for fMRI and DTI, and higher spatial accuracy for MEG) have improved the accuracy in detecting and localizing epileptogenic foci, these advances are only clinically effective if accurate coregistration can be achieved.

„„ Acknowledgment The four authors gratefully acknowledge the following support: NIH P41-EB015898, 5P41-EB015902–20, R01-NS049251, R21-CA198740, R25-CA 089017, U01CA199459.

References 1. Murphy MA, O’Brien TJ, Morris K, Cook MJ. Multimodality image-guided surgery for the treatment of medically refractory epilepsy. J Neurosurg 2004;100(3):452–462

8. Fischl B, Sereno MI, Tootell RB, Dale AM. High-resolution intersubject averaging and a coordinate system for the cortical surface. Hum Brain Mapp 1999;8(4):272–284

2. Brázdil M, Mikl M, Chlebus P, et al. Combining advanced neuroimaging techniques in presurgical workup of non-lesional intractable epilepsy. Epileptic Disord 2006;8(3):190–194

9. Brett M, Johnsrude IS, Owen AM. The problem of functional localization in the human brain. Nat Rev Neurosci 2002;3(3):243–249

3. Rykhlevskaia E, Gratton G, Fabiani M. Combining structural and functional neuroimaging data for studying brain connectivity: a review. Psychophysiology 2008;45(2):173–187 4. Wells WM III, Viola P, Atsumi H, Nakajima S, Kikinis R. Multimodal volume registration by maximization of mutual information. Med Image Anal 1996;1(1):35–51 5. Sugiura M, Kawashima R, Sadato N, et al. Anatomic validation of spatial normalization methods for PET. J Nucl Med 1999;40(2):317–322

10. Mountz JM, Zhang B, Liu HG, Inampudi C. A reference method for correlation of anatomic and functional brain images: validation and clinical application. Semin Nucl Med 1994;24(4):256–271 11. Viñas FC, Zamorano L, Mueller RA, et al. [15O]-water PET and intraoperative brain mapping: a comparison in the localization of eloquent cortex. Neurol Res 1997;19(6):601–608 12. Shiga T, Morita K, Takano A, et al. Clinical advantages of interictal SPECT coregistered to magnetic resonance imaging in patients with epilepsy. Clin Nucl Med 2001;26(4):334–339

6. Talairach J, Tournoux P. Coplanar Stereotaxic Atlas of the Human Brain: 3-Dimensional Proportional System—an Approach to Cerebral Imaging. New York, NY: Thieme Medical; 1988

13. Shin WC, Hong SB, Tae WS, Seo DW, Kim SE. Ictal hyperperfusion of cerebellum and basal ganglia in temporal lobe epilepsy: SPECT subtraction with MRI coregistration. J Nucl Med 2001;42(6):853–858

7. Collins DL, Neelin P, Peters TM, Evans AC. Automatic 3D intersubject registration of MR volumetric data in standardized Talairach space. J Comput Assist Tomogr 1994;18(2):192–205

14. Pietrzyk U. Registration of MRI and PET images for clinical applications. In: Hajnal J, Hawkes D, Hill D, eds. Medical Image Registration. Boca Raton, FL: CRC Press; 2001:199–216

199

200

IIc  Preoperative Neuroimaging 15. Cízek J, Herholz K, Vollmar S, Schrader R, Klein J, Heiss WD. Fast and robust registration of PET and MR images of human brain. Neuroimage 2004;22(1):434–442

34. Golby AJ, Poldrack RA, Illes J, Chen D, Desmond JE, Gabrieli JD. Memory lateralization in medial temporal lobe epilepsy assessed by functional MRI. Epilepsia 2002;43(8):855–863

16. Hogan RE, Lowe VJ, Bucholz RD. Triple-technique (MR imaging, single-photon emission CT, and CT) coregistration for imageguided surgical evaluation of patients with intractable epilepsy. AJNR Am J Neuroradiol 1999;20(6):1054–1058

35. De Tiège X, Connelly A, Liégeois F, et al. Influence of motor functional magnetic resonance imaging on the surgical management of children and adolescents with symptomatic focal epilepsy. Neurosurgery 2009;64(5):856–864, discussion 864

17. Kovalev D, Spreer J, Honegger J, Zentner J, Schulze-Bonhage A, Huppertz HJ. Rapid and fully automated visualization of subdural electrodes in the presurgical evaluation of epilepsy patients. AJNR Am J Neuroradiol 2005;26(5):1078–1083

36. Kesavadas C, Thomas B, Sujesh S, et al. Real-time functional MR imaging (fMRI) for presurgical evaluation of paediatric epilepsy. Pediatr Radiol 2007;37(10):964–974

18. Ken S, Di Gennaro G, Giulietti G, et al. Quantitative evaluation for brain CT/MRI coregistration based on maximization of mutual information in patients with focal epilepsy investigated with subdural electrodes. Magn Reson Imaging 2007;25(6):883–888 19. Zhang Y, van Drongelen W, Kohrman M, He B. Three-dimensional brain current source reconstruction from intra-cranial ECoG recordings. Neuroimage 2008;42(2):683–695 20. Dalal SS, Edwards E, Kirsch HE, Barbaro NM, Knight RT,­ Nagarajan SS. Localization of neurosurgically implanted electrodes via photograph-MRI-radiograph coregistration. ­ J ­Neurosci Methods 2008;174(1):106–115 21. Mahvash M, König R, Wellmer J, Urbach H, Meyer B, Schaller K. Coregistration of digital photography of the human cortex and cranial magnetic resonance imaging for visualization of subdural electrodes in epilepsy surgery. Neurosurgery 2007;61(5, Suppl 2):340–344, discussion 344–345 22. O’Shea JP, Whalen S, Branco DM, Petrovich NM, Knierim KE, Golby AJ. Integrated image- and function-guided surgery in eloquent cortex: a technique report. Int J Med Robot 2006;2(1):75–83 23. Yang AI, Wang X, Doyle WK, et al. Localization of dense intracranial electrode arrays using magnetic resonance imaging. Neuroimage 2012;63(1):157–165 24. Taimouri V, Akhondi-Asl A, Tomas-Fernandez X, et al. Electrode localization for planning surgical resection of the epileptogenic zone in pediatric epilepsy. Int J CARS 2014;9(1):91–105 25. Perry MS, Bailey L, Freedman D, et al. Coregistration of multimodal imaging is associated with favourable two-year seizure outcome after paediatric epilepsy surgery. Epileptic Disord 2017;19(1):40–48 26. Fernández S, Donaire A, Serès E, et al. PET/MRI and PET/MRI/ SISCOM coregistration in the presurgical evaluation of refractory focal epilepsy. Epilepsy Res 2015;111:1–9 27. O’Brien TJ, So EL, Mullan BP, et al. Subtraction ictal SPECT co-­ registered to MRI improves clinical usefulness of SPECT in localizing the surgical seizure focus. Neurology 1998;50(2):445–454 28. Doelken MT, Richter G, Stefan H, et al. Multimodal coregistration in patients with temporal lobe epilepsy—results of different imaging modalities in lateralization of the affected hemisphere in MR imaging positive and negative subgroups. AJNR Am J Neuroradiol 2007;28(3):449–454 29. Breyer T, Wanke I, Maderwald S, et al. Imaging of patients with hippocampal sclerosis at 7 Tesla: initial results. Acad Radiol 2010;17(4):421–426 30. Veersema TJ, van Eijsden P, Gosselaar PH, et al. 7 tesla T2*weighted MRI as a tool to improve detection of focal cortical dysplasia. Epileptic Disord 2016;18(3):315–323 31. Santyr BG, Goubran M, Lau JC, et al. Investigation of hippocampal substructures in focal temporal lobe epilepsy with and without hippocampal sclerosis at 7T. J Magn Reson Imaging 2017;45(5):1359–1370 32. Gabrieli JD, Poldrack RA, Desmond JE. The role of left prefrontal cortex in language and memory. Proc Natl Acad Sci USA 1998;95(3):906–913 33. Wagner AD, Desmond JE, Glover GH, Gabrieli JD. Prefrontal cortex and recognition memory. Functional-MRI evidence for context-dependent retrieval processes. Brain 1998;121 (Pt 10):1985–2002

37. Yousem DM. The economics of functional magnetic resonance imaging: clinical and research. Neuroimaging Clin N Am 2014;24(4):717–724 38. Moeller F, Tyvaert L, Nguyen DK, et al. EEG-fMRI: adding to standard evaluations of patients with nonlesional frontal lobe epilepsy. Neurology 2009;73(23):2023–2030 39. Zijlmans M, Huiskamp G, Hersevoort M, Seppenwoolde JH, van Huffelen AC, Leijten FS. EEG-fMRI in the preoperative work-up for epilepsy surgery. Brain 2007;130(Pt 9):2343–2353 40. Thornton R, Laufs H, Rodionov R, et al. EEG correlated functional MRI and postoperative outcome in focal epilepsy. J Neurol Neurosurg Psychiatry 2010;81(8):922–927 41. Pizarro R, Nair V, Meier T, et al. Delineating potential epileptogenic areas utilizing resting functional magnetic resonance imaging (fMRI) in epilepsy patients. Neurocase 2016;22(4):362–368 42. Maneshi M, Vahdat S, Fahoum F, Grova C, Gotman J. Specific ­resting-state brain networks in mesial temporal lobe epilepsy. Front Neurol 2014;5:127 43. Zhang Z, Lu G, Zhong Y, et al. Impaired perceptual networks in temporal lobe epilepsy revealed by resting fMRI. J Neurol 2009;256(10):1705–1713 44. Duncan JS. Imaging the brain’s highways-diffusion tensor imaging in epilepsy. Epilepsy Curr 2008;8(4):85–89 45. Karis JP; Expert Panel on Neurologic Imaging. Epilepsy. AJNR Am J Neuroradiol 2008;29(6):1222–1224 46. Wu W, Rigolo L, O’Donnell LJ, Norton I, Shriver S, Golby AJ. Visual pathway study using in vivo diffusion tensor imaging tractography to complement classic anatomy. Neurosurgery 2012;70(1, Suppl Operative):145–156, discussion 156 47. Carlson HL, Laliberté C, Brooks BL, et al. Reliability and variability of diffusion tensor imaging (DTI) tractography in pediatric epilepsy. Epilepsy Behav 2014;37:116–122 48. James JS, Radhakrishnan A, Thomas B, et al. Diffusion tensor imaging tractography of Meyer’s loop in planning resective surgery for drug-resistant temporal lobe epilepsy. Epilepsy Res 2015;110:95–104 49. Knowlton RC. Can magnetoencephalography aid epilepsy surgery? Epilepsy Curr 2008;8(1):1–5 50. RamachandranNair R, Otsubo H, Shroff MM, et al. MEG predicts outcome following surgery for intractable epilepsy in children with normal or nonfocal MRI findings. Epilepsia 2007;48(1):149–157 51. Kim D, Joo EY, Seo DW, et al. Accuracy of MEG in localizing irritative zone and seizure onset zone: quantitative comparison between MEG and intracranial EEG. Epilepsy Res 2016;127:291–301 52. Moriarty TM, Kikinis R, Jolesz FA, Black PM, Alexander E III. Magnetic resonance imaging therapy. Intraoperative MR imaging. Neurosurg Clin N Am 1996;7(2):323–331 53. Kurwale NS, Chandra PS, Chouksey P, et al. Impact of intraoperative MRI on outcomes in epilepsy surgery: preliminary experience of two years. Br J Neurosurg 2015;29(3):380–385 54. Olubiyi OI, Ozdemir A, Incekara F, et al. Intraoperative magnetic resonance imaging in intracranial glioma resection: a singlecenter, retrospective blinded volumetric study. World Neurosurg 2015;84(2):528–536 55. Nimsky C, Ganslandt O, Cerny S, Hastreiter P, Greiner G, Fahlbusch R. Quantification of, visualization of, and ­compensation

22  Multimodality Imaging and Coregistration for brain shift using intraoperative magnetic resonance imaging. Neurosurgery 2000;47(5):1070–1079, discussion 1079–1080 56. Upadhyay UM, Golby AJ. Role of pre- and intraoperative imaging and neuronavigation in neurosurgery. Expert Rev Med Devices 2008;5(1):65–73 57. Archip N, Clatz O, Whalen S, et al. Compensation of geometric distortion effects on intraoperative magnetic resonance imaging for enhanced visualization in image-guided neurosurgery. Neurosurgery 2008;62(3, Suppl 1):209–215, discussion 215–216 58. Roessler K, Hofmann A, Sommer B, et al. Resective surgery for medically refractory epilepsy using intraoperative MRI and functional neuronavigation: the Erlangen experience of 415 patients. Neurosurg Focus 2016;40(3):E15 59. Comeau RM, Sadikot AF, Fenster A, Peters TM. Intraoperative ultrasound for guidance and tissue shift correction in imageguided neurosurgery. Med Phys 2000;27(4):787–800 60. Prada F, Del Bene M, Mattei L, et al. Fusion imaging for intraoperative ultrasound-based navigation in neurosurgery. J Ultrasound 2014;17(3):243–251 61. Miga MI, Sun K, Chen I, et al. Clinical evaluation of a modelupdated image-guidance approach to brain shift compensation: experience in 16 cases. Int J CARS 2016;11(8):1467–1474 62. Miller D, Knake S, Bauer S, et al. Intraoperative ultrasound to define focal cortical dysplasia in epilepsy surgery. Epilepsia 2008;49(1):156–158 63. Kanev PM, Foley CM, Miles D. Ultrasound-tailored functional hemispherectomy for surgical control of seizures in children. J Neurosurg 1997;86(5):762–767

64. Lee DJ, Zwienenberg-Lee M, Seyal M, Shahlaie K. Intraoperative computed tomography for intracranial electrode implantation surgery in medically refractory epilepsy. J Neurosurg 2015;122(3):526–531 65. Kerolus MG, Kochanski RB, Rossi M, Stein M, Byrne RW, Sani S. Implantation of responsive neurostimulation for ­epilepsy using intraoperative computed tomography: technical nuances and accuracy assessment. World Neurosurg 2017;103: 145–152 66. Kang JY, Wu C, Tracy J, et al. Laser interstitial thermal therapy for medically intractable mesial temporal lobe epilepsy. Epilepsia 2016;57(2):325–334 67. Le S, Ho A, Fisher RS, et al. Laser Interstitial Thermal Therapy (LITT): Seizure outcomes refractory mesial temporal lobe epilepsy. Epilepsy Behav 2018;89:37–41 68. Golby AJ, Kindlmann G, Norton I, Yarmarkovich A, Pieper S, ­Kikinis R. Interactive diffusion tensor tractography visualization for neurosurgical planning. Neurosurgery 2011;68(2):496–505 69. Pelargos PE, Nagasawa DT, Lagman C, et al. Utilizing virtual and augmented reality for educational and clinical enhancements in neurosurgery. J Clin Neurosci 2017;35:1–4 70. Wendt MA. Bani-Hashemi, Sauer F. Intra-operative image-guided neurosurgery with augmented reality visualization. 2001, Google Patents 71. Archip N, Clatz O, Whalen S, et al. Non-rigid alignment of preoperative MRI, fMRI, and DT-MRI with intra-operative MRI for enhanced visualization and navigation in image-guided ­neurosurgery. Neuroimage 2007;35(2):609–624

201

III Part III

Surgical Anatomy and ­Mapping Techniques

23 Cerebral Cortex: Embryological Development and Topographical Anatomy 

205

24 Tractographic Anatomy of White Matter

228

25 Localization of Motor Cortex and ­Subcortical Pathways Using ­Functional Magnetic Resonance Imaging and ­Diffusion Tensor Imaging 240 26 The Wada Test: Lateralization of Language and Memory

251

27 Language Lateralization and Localization: Functional Magnetic Resonance Imaging

255

28 Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia 259 29 Cortical Stimulation and Mapping 277 30 Subcortical Mapping During Intracranial Surgery in Children

284

203

23

  C  erebral Cortex: Embryological Development and Topographical Anatomy Oğuz Çataltepe

Summary The gross morphology of the developing brain undergoes striking changes during fetal period. Brain development continues after birth as well. The most dynamic and critical period of postnatal brain development is the first 2 years of life, and the developing human brain undergoes dynamic structural changes in cortical thickness and cortical folding patterns that are tightly linked to specific functionality and axonal connectivity during the early neonatal and later fetal periods. The development of the cortical patterns occurs in two stages: migration and gyrification. The human cerebral cortex is composed of numerous foldings with vast number of variations. The sulci define the gyral topography and provide intrinsic landmarks to functionally distinct regions of the cerebral hemispheric surfaces. Intimate knowledge of the cortical anatomy is essential, even mandatory, for every neurosurgeon because precise definition of the sulcal anatomy and cortical surface landmarks is indispensable to planning and performing safe and accurate surgery in the brain. Keywords:  cerebral cortex, sulcus, gyrus, frontal lobe, parietal lobe, occipital lobe, temporal lobe, insula

„„ Introduction The human cerebral cortex is composed of numerous foldings with vast number of variations. This complex topography creates a significant challenge to neurosurgeons for locating the lesions and navigating to the target safely. Thanks to the advancement of imaging and neuronavigation technologies, translating neuroanatomical information into the operating room has become much more feasible today than ever before. However, intimate knowledge of the cortical anatomy remains essential, even mandatory, for every neurosurgeon. It is not only significant as a basic anatomical information but also for very practical reasons. These include assessing the topographical location of a lesion relative to well-defined cortical landmarks and planning the most appropriate surgical approach for a safer navigation in microsurgical corridors close to functionally critical areas. Therefore, precise definition of the sulcal anatomy and cortical surface landmarks is indispensable to planning and performing safe and accurate surgery in the brain. Sulci are extensions of cortical subarachnoid spaces and provide natural

microneurosurgical corridors even to the deep-seated surgical targets. Hence, intimate familiarity with the cortical anatomy of the brain allows neurosurgeons to safely navigate along predetermined surgical corridors by using transsulcal approaches or subpial dissection techniques more efficiently. Intimate knowledge of cortical anatomical landmarks would also enable the neurosurgeon to securely reach certain targets in the brain, such as, finding a small temporal horn by simply following the collateral sulcus, or reaching the frontal horn by following the direction of the anterior limiting sulcus at the bottom of the pars triangularis, etc. Cortical anatomy knowledge is especially helpful in epilepsy surgery cases, because these patients frequently have a nondistorted anatomy with reliable cortical landmarks.1,​2,​3,​4,​5,​6,​7,​8

„„ The Human Cerebral Cortex The cerebral cortex has a large surface area (1,200–1,500 cm2 in adults) with numerous foldings. Only one-third of the cortical surface area is visible and the rest stays hidden within, or under, the cortical foldings.4,​8,​9,​10 Cerebral cortex can be defined based on its different properties such as topographical features, functional characteristics, cytoarchitecture, connectivity, etc. Although all these mapping and parcellation efforts provide a valuable perspective in understanding the cerebral cortex, defining the topographical and functional anatomy based on cortical patterns and landmarks verifiable both on neuroimaging studies and during surgery remains the most valuable approach from a neurosurgical standpoint.11 Yaşargil5 describes the cortical anatomy by dividing each cerebral hemisphere into seven lobes—the frontal, central, parietal, occipital, temporal, insular, and limbic lobes—defined by specific sulci and gyri. In addition to traditional lobar divisions of the cerebral hemisphere, he defines three additional lobes based on their common embryological, anatomical, physiological, and functional characteristics: the central lobe, the insular lobe, and the limbic lobe. This is a very practical and useful definition in reviewing the cortical anatomy from a neurosurgical standpoint. Cerebral cortex does indeed have a complex and difficult-to-define topography with extensive infoldings, variable configurations, and series of interconnected gyri with deep gyral bridges hidden within the sulcal depths.

206

III  Surgical Anatomy and Mapping Techniques Yaşargil5 puts a special emphasis on the “gyral continuum” concept (Fig. 23.1). The gyral continuum can be defined as an uninterrupted continuum of gyri in which the main gyri are connected by “short and long, small or voluminous transverse gyri” (Fig. 23.2).5 According to Tamraz and Alkadhi, similar observations on cortical anatomy were made earlier by Gratiolet and Paul Broca.12,​13 They used the “pli de passage” term to describe an anatomical bridge connecting one gyrus to another or a transition of one gyrus into another without any anatomical barrier between them. Yaşargil5 nicely defined the uninterrupted connection of the entire cortex: “Starting at a polar area (whether frontal, temporal, or occipital) and tracing the contour of any prominent gyrus, it is possible to follow the same gyrus without interruption along the whole length of the hemisphere…”.5

The perisylvian gyral connection is a good example of the gyral continuum. We can easily follow all gyri surrounding the three sides of this fissure as a single gyrus starting from the rostral end of the superior temporal gyrus and advancing first to the posterior and then anterior limbs of the supramarginal gyrus, and then to the lower postcentral gyrus, the subcentral gyrus, lower precentral gyrus, the inferior frontal gyrus including the pars opercularis, the pars triangularis, and finally ending at the pars orbitalis. The gyral continuum concept is not only significant as an anatomical concept, but it is also critical from a neurosurgical standpoint. The neurosurgeon can easily and unintentionally enter into a functionally critical adjacent gyrus through a connecting short gyral bridge while subpially emptying the gyral content of the same sulcal bank of another gyrus.5,​6,​7

Fig. 23.1  The gyral continuum concept according to Dr. Yaşargil: uninterrupted connection, either on the surface or intrasulcally, between every gyrus within the hemisphere. (Reproduced with permission from Yaşargil MG. Microneurosurgery Vol IVA; Thieme, 1994.)

Fig. 23.2  The transverse gyri (arrows) within the sulci are visible in a rubberized brain.

23  Cerebral Cortex: Embryological Development and Topographical Anatomy

„„ Embryology and Cortical Development of the Cerebral Cortex The prenatal development of the cerebral cortex occurs in two interrelated periods: embryonic (fifth to eighth week) and fetal (eighth week to birth). The fetal period is characterized by an extraordinary surface expansion and cortical folding process.1,​14 The gross morphology of the developing brain undergoes striking changes during this time. The human brain begins as a smooth, “lissencephalic” structure and gradually develops its characteristic pattern of gyral and sulcal foldings.14,​15,​16 Initially, the only visible depression on the lateral surface of the brain is the sylvian fissure with underlying insula as a shallow indentation as early as the 14th week. The formation of gyri and sulci follows an orderly sequence. Sulci are described as primary, secondary, and tertiary sulci. Sulci found in all gyrencephalic primates are called primary sulci and develops before the 30th week of gestation.12 Primary sulci are first seen as grooves positioned in eloquent brain regions. Secondary branches then begin to form off of the primary sulci, followed by the tertiary branches. The first

fissure to form is the longitudinal fissure that separates two cerebral hemispheres. It starts as a superior midline depression of the telencephalic vesicle. Its development begins in the rostral part as early as gestational week of 8 (GW8) and proceeds caudally until it is complete at GW22. Then, the transverse fissure appears between the telencephalon and diencephalon vesicles. Gradually, other primary sulci develop: the sylvian fissure, cingulate, parieto-occipital, and calcarine (GW14–16); the central and superior temporal (GW20–24); and the superior frontal, precentral, inferior frontal, postcentral, and intraparietal (GW25–26) sulci. Secondary sulci emerge between GW30 and GW35, the formation of tertiary sulci begins during GW36 and extends well into the postnatal period (Fig. 23.3). The central, precentral, and postcentral sulci initially appear as two separate upper and lower parts, and then they become single sulci. Opercularization begins at 20 to 22 weeks and progresses from the posterior to the anterior. The anterior part of the opercularization stays incomplete until the term of gestation and the insula remains somewhat exposed (Fig. 23.4, Fig. 23.5, and Fig. 23.6).10,​12,​14,​15,​17,​18,​19,​20,​21,​22 The average surface area of the human cerebral cortex is 20 cm2 at the 12th week of gestation and becomes 716 cm2 at the end of the fetal period.19 Throughout this process, cortical thickness increases from 350 μm at the 11th week of

Fig. 23.3  Developmental timeline for gyrification. (Reproduced with permission from White et al 2010.15)

207

208

III  Surgical Anatomy and Mapping Techniques

Fig. 23.4  Development of sulci of the superolateral cerebral surface of the fetus at 19 (a), 20 (b), 21 (c), 24 (d), 29 (e), and 36 weeks (f). The arrows illustrate the central sulci development. w, weeks; CS, central sulcus; IFS, inferior frontal sulcus; IHF, interhemispheric fissure (longitudinal cerebral fissure); IPS, intraparietal sulcus; ITS, inferior temporal sulcus; OrbS, orbital sulcus; OTS, occipitotemporal sulcus; PostCS, postcentral sulcus; PreCS, precentral sulcus; SFS, superior frontal sulcus; STS, superior temporal sulcus; SyF, lateral sulcus (sylvian fissure); TrOS, transverse occipital sulcus. (Reproduced with permission from Nishikuni and Ribas 2013.18)

Fig. 23.5  Development of sulci of the inferior cerebral surface of the fetus at 17 (a), 20 (b), 23 (c), 30 (d), 33 (e), and 37 weeks (f). ColS, collateral sulcus; HS, hippocampal sulcus; OlfBu, olfactory bulb; OlfS, olfactory sulcus; OlfTr, olfactory tract; RhiS, rhinal sulcus. (Reproduced with permission from Nishikuni and Ribas 2013.18 )

23  Cerebral Cortex: Embryological Development and Topographical Anatomy

Fig. 23.6  Development of sulci of the medial cerebral surface of the fetus at 12 (a), 17 (b), 18 (c), 20 (d), 33 (e), and 35 weeks (f). The white arrows in (a) indicate the appearance of transitory furrows during the initial phase of cerebral development. The black arrows in (f) indicate two parallel cingulate sulci. CaF, calcarine sulcus; CaS, callosal sulcus; CC, corpus callosum; CiS, cingulate sulcus; MaCiS, marginal branch of the cingulate sulcus; PaCS, paracentral sulcus; PaOlfS, paraolfactory sulcus; POS, parietooccipital sulcus (calcarine fissure); SubPS, subparietal sulcus. (Reproduced with permission from Nishikuni and Ribas 2013.18)

gestation to 1,500 to 1,700 μm at the 40th week of gestation.19 This impressive increase in the surface area without affecting the total brain volume occurs because of the infolding process that creates numerous sulci and fissures. The end result of this unique developmental process is the presence of a large amount of hidden cortical area (two-thirds of the total cortical surface) in the depths of the sulci and fissures.4 The development of the sulcal patterns on the lateral, basal, and medial surfaces of the cerebrum shows some differences. While the directions of the sulci on the lateral and ­basal surfaces are toward the lateral ventricle, their orientation on the medial hemispheric surfaces is influenced by the development of the corpus callosum.4,​12 Brain development continues after birth as well. The most dynamic and critical period of postnatal brain development is the first 2 years of life.17,​21 The brain volume increases significantly and more than doubles in size during the time period between birth and 2 years of age, and the brain reaches 80% of adult size at the end of this period.23 The cortical gray matter volume doubles in size in the first year of life and further increases in the second year (14–18%).17 The primary motor and sensory cortical areas grow more slowly, while the association cortices grow relatively rapidly. The central lobe, the Heschl’s gyrus, the cuneus, the peri-calcarine cortex, and the superior temporal gyrus exhibit the slowest growth rate in the first year of life. The fastest growing regions during this period are the insula, the inferior frontal gyrus (opercular part), the superior frontal gyrus (orbital part), the inferior temporal gyrus, the temporal pole, the median cingulate, the paracingulate gyri, the angular gyrus, and the fusiform gyrus. The growth of the hippocampus is also less than the amygdala in the first year of life. In the second year of life, the angular and supramarginal gyri, dorsolateral and medial superior frontal gyri, the middle

f­rontal gyrus, and the temporal pole of the middle temporal gyrus exhibit the fastest growth rate.17,​21,​23 Although not as much as the first 2 years, the brain continues to grow and mature until the teenage years. Total brain volume reaches 95% of its maximum size by 6 years of age.24 Cerebral cortex on the parietal and frontal lobes reaches its maximal volume at 10 and 11 years of age, respectively.24 This peak occurs for the temporal lobe at 16 years. The last maximization on cortical volume in the temporal lobe occurs in the superior temporal gyrus, which integrates audiovisual input and object recognition functions (along with prefrontal and inferior parietal cortices). Again, remarkable maturation is seen in certain temporal lobe areas such as the amygdala and hippocampus between the ages of 4 and 18 years.24 Although the thickness and surface area of the cerebral cortex peak before the age of 10 years, it declines during adolescence years. The total volume of gray matter decreases, while the total volume of white matter increases because of synaptic pruning and myelination through the 20th year of life.25 In summary, the developing human brain undergoes dynamic structural changes in cortical thickness and cortical folding patterns that are tightly linked to specific functionality and axonal connectivity during the early neonatal and later fetal periods.21,​26

„„ The Cerebral Sulci and Gyri On the basis of comparative anatomy, the human brain has carried some mammalian evolutionary trends to extreme limits with a disproportionate growth of the isocortex and its closely connected central core structures such as thalamus, internal capsule, and the corpus callosum. The cortical folding and the transformation of the lissencephalic brain into a gyrencephalic

209

210

III  Surgical Anatomy and Mapping Techniques brain were driven by the disproportionate expansion of isocortex without changing the fundamental topological organization of the vertebrate brain.10,​27 The development of the cortical patterns is not a random process. It occurs in two stages: migration and gyrification. Throughout this process, regional axonal distribution plays a significant role in gyrification and associated sulcal arrangements by anchoring certain regions together and letting other regions drift apart with cortical expansion. This and some other dynamics of the infolding process play a ­significant role in gyral and sulcal patterns, such as sulcal orientation toward the nearest ventricular cavity on the lateral and basal surfaces, and sulcal orientation around the corpus callosum on interhemispheric cortical surfaces.4,​8,​10,​28 The sulci define the gyral topography and provide intrinsic landmarks to functionally distinct regions of the cerebral hemispheric surfaces. The sulci exhibit numerous variations in their length, depth, and continuity. Their depth varies from a few millimeters to 3 cm. Another variable characteristic of the sulci is their continuity. Continuous, uninterrupted sulci are frequently related to specialized areas. The central sulcus, the sylvian fissure, the callosal, the collateral, the parieto-occipital, and the calcarine sulci are almost always uninterrupted. The remaining sulci are frequently interrupted to varying degrees. If a sulcus is wider, deeper, and anatomically more constant than usual, it is known as a “fissure.” The sulci are also defined as “primary,” “secondary,” and “tertiary” based on the timetable of their embryological development as described above.5,​6,​7,​8,​10,​12 The gyral anatomy is defined by the surrounding sulci. The gyri are continuous, irregular, undulated convolutions with uninterrupted interconnections through smaller transvers gyri within the sulcal spaces. Yaşargil5 defines this characteristic with the “gyral continuum” term. Although each gyrus is named and defined as if it was an individual topographical unit, it should be seen as a part of larger functional unit. The gyri also exhibit numerous variations just like the sulci, with varying width, cortical thickness, and continuity. The cortical thickness varies widely within the range of 1.5 to 4.5 mm.9,​29 The thickest cortex is found in the primary motor area (4.5 mm) and the thinnest in the primary visual cortex (1.5 mm).29 The cortical thickness and surface area of the gyrus are closely related to the hemispheric dominance, function, and specialization of that region. Therefore, the same gyri on both hemispheres may show a certain degree of asymmetry based on the cerebral dominance. The surface area of the Heschl’s gyrus is usually larger on the left side, and similarly, the surface area of the planum temporale is found to be much larger on the dominant hemisphere. The cortical thickness and gray matter volume also co-vary in functionally related cortical areas such as Broca’s and Wernicke’s areas. Typically, if the cortical thickness of Broca’s area is greater in an individual brain, then the cortical thickness of Wernicke’s area is found to be greater as well.4,​23

Frontal Lobe The frontal lobe covers most anterior part of the hemisphere and occupies the largest area in each hemispheric surface. It contains almost 40% of the human cortex (Fig. 23.7 and Fig. 23.8).27 The frontal lobe has medial, lateral, and basal ­surfaces. The frontal pole constitutes a transitional area between the lateral, medial, and basal surfaces of the frontal lobe and is shaped by

merging superior, middle frontal, orbital gyri and gyrus rectus. It is the only lobe that is anatomically separated from the other lobes with well-defined borders on all sides.

Lateral Surface Upper border of the lateral surface is defined by the dorsal edge of the hemisphere, inferior border and posterior borders are defined by the sylvian fissure and the precentral sulcus, respectively (Fig. 23.9). Horizontally oriented two sulci, the superior and inferior frontal sulci, along with several rami arising from the sylvian fissure and the precentral sulcus define the gyral anatomy of the lateral surface of the frontal lobe. The superior frontal sulcus arises from orbital margin of the hemisphere and extends parallel to the interhemispheric fissure by gradually separating from the fissure as it goes posteriorly. It is a shallow sulcus and frequently interrupted with short transvers gyri.12 The superior frontal sulcus ends at the precentral sulcus with T-shaped branching. The inferior frontal sulcus arises just behind the lateral orbital gyrus, frequently with a Y-shape bifurcation, and extends parallel to the sylvian fissure. It is a very deep sulcus and comes very close to the insular plane in some areas. It is an interrupted sulcus in almost 50% of the cases and ends at the inferior precentral sulcus.10,​12 The precentral sulcus shows a continuum with posterior ascending and descending bifurcations of the superior and inferior frontal sulci and essentially formed by these two branches. The precentral sulcus is divided into the superior and inferior precentral sulci in 75% of cases by a bridging cortical connection between the precentral and middle frontal gyri.10,​12 The precentral sulcus extends parallel to the central sulcus. Its upper end does not reach to the dorsal edge of the hemisphere, but lower end is frequently connected with the sylvian fissure.5,​6,​7,​8,​10 Three wide gyri, along the sulci described above, constitute the lateral surface of the frontal lobe: the superior, middle, and inferior frontal gyri. The superior frontal gyrus (F1) is the longest frontal gyrus and its width is 1 to 2 cm on the lateral surface. It extends between the superior part of the precental gyrus posteriorly and the frontal pole anteriorly. It is connected to the precentral gyrus with a short cortical bridge. It extends to the dorsal margin of the hemisphere and then continues on the medial surface of the frontal lobe. The superior frontal gyrus is the only gyrus extending on all three surfaces of the frontal lobe. Interhemispheric portion of F1 is called the medial (= mesial) frontal gyrus and lies between the dorsal edge of F1 and the cingulate sulcus. Posterior interhemispheric part of the superior frontal gyrus is known as the supplementary motor area. Borders of this important region are poorly defined and shows individual variations. The superior frontal gyrus is posteriorly connected to the precentral gyrus and anteriorly connected to the middle frontal gyrus as well as the orbital and rectus gyri at the frontal pole.5,​6,​7,​8,​12,​30 The middle frontal gyrus (F2) is the widest frontal gyrus and lies between the superior frontal sulcus and inferior frontal sulcus. It is separated from the precentral gyrus posteriorly by the precentral sulcus, but this is not a complete separation. It is still connected to the precentral gyrus with a short cortical bridge at its lower part. The middle frontal gyrus anteriorly merges with other frontal gyri at the frontal pole as described above. The premotor cortex is located on the posterior part of the superior and middle frontal gyri and constitutes a transition area

23  Cerebral Cortex: Embryological Development and Topographical Anatomy

Fig. 23.7  Superior view of the gyral convolutions of both hemispheres. a, superior frontal sulcus; b, paracentral sulcus; c, precentral sulcus; d, central sulcus; e, postcentral sulcus; f, marginal ramus of cingulate sulcus; g, intraparietal sulcus. 1, superior frontal gyrus; 2, middle frontal gyrus; 3, precentral gyrus; 4, postcentral gyrus; 5, superior parietal gyrus; 6, superior occipital gyrus; 7, middle occipital gyrus. (Anatomical specimens in Figures 23.7−23.14 were prepared by Prof. Dr. Uğur Türe.)

between the frontal polar cortex and the primary motor cortex. It covers approximately six times larger area than the primary motor cortex, although boundaries are not clearly defined.29 The premotor cortex does not have a somatotopic organization and involves in voluntary motor functions dependent on visual, auditory, and somatosensory inputs. It coordinates synergistic movements in collaboration with the cerebellum. Its electrical stimulation produces stereotyped gross movements that require coordinated activities of multiple muscle groups, such as turning the head, eye, and trunk; elevating or flexing the arm, elbow, etc. Electrical stimulation threshold of the premotor cortex is higher than the primary motor cortex. Excision of this area may produce a motor deficit related to sequential complex movements (= apraxia). The frontal eye field area is located on the middle frontal gyrus anterior to the premotor cortex. It plays a role in directing the gaze. Electrical stimulation or lesions in this area produces conjugate eye deviation to contralateral side.29 The inferior frontal gyrus (F3) covers the smallest area in the lateral surface of the frontal lobe and extends between the sylvi-

an fissure and the inferior frontal sulcus. The inferior segment of the precentral sulcus defines the posterior limit of F3. It has an irregular pattern with frequent interruptions by various small branches of the inferior frontal sulcus and several prominent rami of the sylvian fissure. Anteriorly, the inferior frontal gyrus terminates by merging with the middle frontal gyrus. Although, the precentral sulcus delineates posterior border of F3, there is a cortical bridge, “pli de passage,” connecting F3 to the precentral gyrus.5,​6,​7,​12,​30 The horizontal and ascending rami of the sylvian fissure typically emerge from a widened subarachnoid space in the anterior portion of the sylvian fissure. Retracted nature of the pars triangularis causes this visible widening in the anterior portion of the sylvian fissure, and this area is called “anterior sylvian point.” The horizontal and ascending rami of the sylvian fissure divides the inferior frontal gyrus into three parts: the pars orbitalis, the pars triangularis, and the pars opercularis. The pars orbitalis constitutes more prominent part of F3. These three parts of the inferior frontal gyrus continue without interruption as a folded gyral ribbon. The pars triangularis is anatomically more retracted than

211

212

III  Surgical Anatomy and Mapping Techniques

Fig. 23.8  Frontal and temporal poles; anterior view. a, superior frontal sulcus; b, H-shaped orbital sulcus; c, olfactory sulcus; d, sylvian fissure; e, superior temporal sulcus; 1, superior frontal gyrus; 2, middle frontal gyrus; 3, anterior orbital gyrus; 4, medial orbital gyrus; 5, lateral orbital gyrus; 6, posterior orbital gyrus; 7, superior temporal gyrus; 8, temporal pole; 9, middle temporal gyrus; 10, inferior temporal gyrus; 11, gyrus rectus.

Fig. 23.9  Lateral aspect of left hemisphere. a, superior frontal sulcus; b, inferior frontal sulcus; c, horizontal ramus; d, ascending ramus; e, precentral sulcus; f, central sulcus; g, postcentral sulcus; h, intraparietal sulcus; i, temporal occipital incisura; j, sylvian fissure; k, superior temporal sulcus; l, inferior temporal sulcus; 1, frontal pole; 2, superior frontal gyrus; 3, middle frontal gyrus; 4, inferior frontal gyrus; 5, pars orbitalis of inferior frontal gyrus; 6, pars triangularis of inferior frontal gyrus; 7, pars opercularis of inferior frontal gyrus; 8, subcentral gyrus; 9, precentral gyrus; 10, postcentral gyrus; 11, parietal operculum; 12, supramarginal gyrus; 13, angular gyrus; 14, superior parietal lobule; 15, occipital pole; 16, inferior occipital gyrus; 17, superior temporal gyrus; 18, middle temporal gyrus; 19, inferior temporal gyrus; 20, temporal pole; 21, limen insula.

the others and located between the horizontal and ascending rami of the sylvian fissure. A small descending sulcus branching from the inferior frontal sulcus quite often cut into the pars triangularis. The pars triangularis is typically V shaped in dominant hemisphere and Y shape in the nondominant hemisphere. The pars opercularis is delineated by the anterior sylvian point anteroinferiorly, and it is a consistently U-shaped folding. It lies between the ascending ramus and the precentral sulcus as well as the anterior subcentral branch of the sylvian fissure. Posterior half of the pars opercularis continues with the subcentral gyrus. The pars triangularis and the pars opercularis are generally more developed in the dominant hemisphere and correspond to the Broca’s speech area. They are folded into the depth of the sylvian fissure and reach into the anterior and superior circular sulcus of insula.4,​5,​6,​7,​8,​10,​12,​30,​31

Medial Surface The interhemispheric part of the superior frontal gyrus (= the medial frontal gyrus) and the anterior cingulate gyrus constitute the medial aspect of the frontal lobe (Fig. 23.10). The anterior cingulate gyrus extends between the anterior callosal and cingulate sulci. The cingulate sulcus starts just below the rostrum of the corpus callosum and follows the curve of the corpus callosum by sweeping around the genu and separating the cingulate and the medial frontal gyri. It ends with the marginal ramus that separates the paracentral lobule and the precuneus. Although we review the anterior cingulate as part of the frontal lobe here, it is also described as part of the “limbic lobe.” The cingulate gyrus abuts the subcallosal gyrus anteriorly and continues with the parahippocampal gyrus through the isthmus posteriorly.4,​5,​6,​7,​8,​12,​30

23  Cerebral Cortex: Embryological Development and Topographical Anatomy

Fig. 23.10  Medial view of the left hemisphere. a, subcallosal sulcus; b, susorbital sulcus; c, cingulate sulcus; d, paracentral sulcus; e, central sulcus’ notch; f, marginal ramus of cingulate sulcus; g, parieto-occipital sulcus; h, calcarine sulcus; i, uncus; j, collateral sulcus; k, rhinal sulcus; 1, subcallosal gyrus; 2, medial frontal gyrus; 3, cingulate gyrus; 4, paracentral lobule; 5, precuneus; 6, cuneus; 7, fourth occipital gyrus; 8, isthmus; 9, lingual gyrus; 10, fusiform gyrus; 11, parahippocampus.

The superior frontal gyrus constitutes superior medial aspect of the frontal lobe and covers a larger area than the cingulate gyrus. It extends superiorly to the dorsal edge of the hemisphere and inferiorly to the cingulate sulcus. Although anterior and posterior borders are not very well defined, the gyrus rectus delineates its anterior border and the paracentral sulcus defines its posterior border. Posteriorly, the superior frontal gyrus is connected to the precentral gyrus. The supplementary motor area (SMA) is located in the posterior portion of the medial frontal gyrus and mainly covers an area anterior to the paracentral lobule. The SMA borders are not well defined anteriorly. There is no clear sulcal demarcation between the SMA and the primary motor cortex. The cingulate sulcus makes its inferomedial border, the precentral sulcus delineates the SMA posteriorly. Some authors, however, extend its borders into the cingulate gyrus inferiorly and over the lateral convexity toward superior frontal sulcus superolaterally. The SMA is connected reciprocally with the ipsilateral primary motor, premotor, and somatosensory areas as well as the contralateral supplementary motor area through the corpus callosum. The SMA plays a crucial role in initiation of movement, speech, and postural adjustments with sequential performance of multiple movements.5,​6,​7,​12,​29,​30,​32 Although the supplementary motor area is described as two parts, the pre-SMA and SMA proper, this division is essentially based on differing functions and cortical–subcortical connections.32 There is no well-defined topographical border between the pre-SMA and the SMA proper. SMA stimulation gives rise to a complex movement described as “fencing ­posture”: abducted contralateral arm with externally rotated shoulder and flexed elbow. Threshold for electrical stimulation of the SMA is higher than the primary motor cortex, and both ipsi- and bilateral responses can be seen with the SMA stimulation.29 The SMA proper is organized somatotopically and directly connected to the primary motor cortex and spinal cord by the corticospinal tract. The face and upper limb areas are represented rostral to the lower limbs and trunk areas on the SMA proper. The pre-SMA, on the other hand, has a somatosensory organization and higher-order functions. Removal of the SMA may result with postoperative deficits including hypertonia, motor and speech initiation problems ranging from delayed initiation to complete suppression with

akinetic mutism and motor paralysis.6,​7,​29 This is called “the SMA syndrome,” and it is almost always reversible within a few weeks because of compensation mechanisms via the contralateral SMA complex. However, disturbance of alternating movements of two hands might be permanent.

Basal Surface The mediobasal and fronto-orbital surfaces show a continuum and make basal surface of the frontal lobe (Fig. 23.11). The rostral sulcus shapes the mediobasal surface along the anterior cingulate sulcus. It may be seen as the inferior, superior, and an accessory rostral sulcus in some cases. It starts from the paraolfactory sulcus and then courses around the rostrum of the corpus callosum parallel to the cingulate sulcus and ends just behind the frontal pole on the mesial surface. The gyrus rectus extends between the olfactory sulcus and the interhemispheric fissure. It is anatomically the most constant and straight gyrus. It covers a narrow segment on both the mesial and basal surface of the frontal lobe and merges with the superior frontal gyrus on both surfaces. The olfactory sulcus is a very deep and straight sulcus extends parallel to the interhemispheric fissure. It starts just behind the frontal pole on the basal surface and ends in the anterior perforated substance with the medial and lateral striae. It has a paramedian location close to midline and separates the gyrus rectus and the orbital gyrus. An H-shaped orbital sulcus (= cruciform sulcus of ­Rolando) divides the orbital gyrus to four separate, smaller gyri: the anterior, posterior, medial, and lateral orbital gyri (Fig. 23.11). These four gyri together cover the largest portion of the frontobasal surface.4,​5,​6,​7,​8,​10,​12,​30

Central Lobe The central lobe is composed of the precentral, postcentral, subcentral gyri and the paracentral lobule (Fig. 23.7, Fig. 23.9, and Fig. 23.10). All these gyri together make a distinct morphological and functional unit that deserves to be named as a separate lobe. It is divided vertically by the central sulcus and surrounded by multiple sulci and rami including the precentral, the postcentral, and the paracentral sulci as well as the ­anterior subcentral, the posterior subcentral rami, and the m ­ arginal

213

214

III  Surgical Anatomy and Mapping Techniques

Fig. 23.11  Inferior view of both hemispheres. a, H-shaped orbital sulcus; b, olfactory sulcus; c, lateral olfactory stria; d, medial olfactory stria; e, anterior perforated substance; f, collateral sulcus (medial occipitotemporal sulcus); g, lateral occipitotemporal sulcus; 1, anterior orbital gyrus; 2, lateral orbital gyrus; 3, medial orbital gyrus; 4, posterior orbital gyrus; 5, gyrus rectus; 6, temporal pole; 7, uncus; 8, parahippocampal gyrus; 9, fusiform gyrus; 10, inferior temporal gyrus; 11, lingual gyrus; 12, fourth occipital gyrus; 13, inferior occipital gyrus.

ramus of cingulate sulcus. It has medial, dorsolateral, and opercular–insular surfaces and corresponds to one of the most eloquent areas of the brain: the primary sensorimotor cortex. The precentral and postcentral gyri extend on an obliquely oriented course on the lateral surface of the brain. These two gyri are connected at upper and lower ends and have close functional interactions. Connected upper part on the mesial surface of the hemisphere is called the paracentral lobule, and short gyral connection at the lower end of the central sulcus is called the subcentral gyrus.5,​6,​7,​8,​12,​30,​33 The precentral and postcentral sulci define anterior and posterior borders of the central lobe, and the central sulcus divides the central lobe into two parallel gyri that extends from the interhemispheric fissure to the sylvian fissure by following an anteriorly oblique and sinuous trajectory. The central sulcus (= fissure of Rolando) is the most constant sulcus on the human brain. It is almost always uninterrupted with a variable depth (12–17 mm), and its superior and inferior ends are capped with two gyral connections: the paracentral lobule and the subcentral gyri.10,​12,​33 It frequently (80%) starts on the dorsal edge of the hemisphere with a small backward notch in the paracentral lobule, “crochet Rolandique,” on the mesial aspect of the hemisphere.3,​12 It follows a sinusoidal course by making three curves (superior, middle, and inferior genu) and ends without direct connection to the sylvian fissure.12,​33 While the superior genu and the inferior genu are anteriorly concave curves, the middle genu makes a posteriorly concave curve. Middle genu is the deepest portion of the central sulcus

that also corresponds to hand–arm area. The lower end of the central sulcus frequently (80–85%) does not reach to the sylvian fissure and very well aligned with the central sulcus of the insula (Fig. 23.12 and Fig. 23.13).3,​10,​12 The precentral sulcus follows a similar trajectory to the central sulcus. Its upper end reaches to the interhemispheric fissure less frequently than the central sulcus. Its lower end defines the posterior border of the pars opercularis of the inferior frontal gyrus. The precentral sulcus is an interrupted sulcus. A few gyral bridges (1–3) connect the precentral gyrus to the superior, middle, and inferior frontal gyri. The postcentral sulcus is deeper than the central sulcus and its depth reaches up to 2 cm. It opens directly into the interhemispheric and sylvian fissures in about half of the cases. It has constant connection to the intraparietal sulcus in most cases.3,​5,​6,​7,​8,​10,​12,​30 The precentral gyrus corresponds to the motor cortex and constitutes one of the widest gyri of the hemisphere (9–15 mm in width and 10–12 cm in length).5,​12 It extends between the interhemispheric and sylvian fissures by following the same trajectory and orientation with the central sulcus. It follows a perpendicular and slightly oblique orientation from posterior–superior to anterior–inferior direction. It has multiple connections to the adjacent gyri. It is connected to the postcentral gyrus both at the upper and lower ends. It merges with the postcentral gyrus at the upper end to make the paracentral lobule on the medial aspect of the hemisphere and at the lower end to make the subcentral gyrus (= Rolandic

23  Cerebral Cortex: Embryological Development and Topographical Anatomy

Fig. 23.12  Insula and superior aspect of superior temporal gyrus after removal of the frontoparietal opercular areas. a, inferior frontal sulcus; b, anterior limiting sulcus; c, superior circular sulcus; d, central sulcus (Rolandic sulcus); e, precentral insular sulcus; f, central insular sulcus; g, postcentral insular sulcus; h, sulcus acusticus; i, transverse temporal sulcus; j, inferior circular sulcus; k, limen insula; 1, precentral gyrus; 2, postcentral gyrus; 3, anterior shorth insular gyrus; 4, posterior shorth insular gyrus; 5, anterior long insular gyrus; 6, posterior long insular gyrus; 7, Heschl’s gyrus (transverse temporal gyrus); 8, planum temporale; 9, superior temporal gyrus.

Fig. 23.13  Mesial temporal structures and insula after removal of superior and lateral temporal cortex. a, postcentral sulcus; b, central sulcus; c, precentral sulcus; d, ascending ramus; e, horizontal ramus; f, H-shaped orbital gyrus; g, precentral insular sulcus; h, central insular sulcus; i, postcentral insular sulcus; 1, parietal opercula; 2, subcentral gyrus; 3, pars opercularis, inferior frontal gyrus; 4, pars triangularis, inferior frontal gyrus; 5, pars orbitalis, inferior frontal gyrus; 6, lateral orbital gyrus; 7, anterior orbital gyrus; 8, medial orbital gyrus; 9, posterior orbital gyrus; 10, apex of insula; 11, short insular gyri; 12, long insular gyri; 13, amygdala; 14, hippocampus; 15, postcentral gyrus; 16, precentral gyrus.

­ perculum). The precentral gyrus has several cortical bridges o with three frontal gyri anteriorly. The precentral gyrus also has a connection with the postcentral gyri through a short transverse gyrus (“pli de passage frontoparietal moyen of ­Broca”) that is not visible on the surface and hidden at the depth of the central sulcus. This area is located on the middle genu of the central sulcus and corresponds to hand area that can be localized on MRI easily because of its resemblance to a “hook” or “omega” sign.4,​5,​6,​7,​8,​10,​12,​30,​34 The postcentral gyrus corresponds to the somatosensory area and has a similar topological configuration to the precentral gyrus. It follows the central sulcus just like the precentral gyrus. It is generally narrower than the precentral gyrus except its opercular part.12 While the postcentral gyrus’ extension in the paracentral lobule is just a small isthmus and much less prominent than that of the precentral gyrus, its opercular part is significantly wider than the precentral gyrus. Lower end of the postcentral gyrus reaches the sylvian fissure just above the Heschl’s gyrus (Fig. 23.12). The postcentral gyrus is posteriorly connected to the parietal lobules with several gyral bridges.33

The paracentral lobule is located on the mesial aspect of the central lobe and covers an area between the distal part of the cingulate sulcus and the dorsal edge of the hemisphere (Fig. 23.10). Its posterior border is well delineated by the marginal ramus of the cingulate sulcus. Its anterior border is defined by the paracentral sulcus that is located just anterior to the upper end of the precentral sulcus. However, anterior border of the paracentral lobule is frequently unclear because both the paracentral sulcus and the precentral sulcus are quite shallow at this area. The subcentral gyrus (= Rolandic operculum) is composed of merged lower ends of the precentral and the postcentral gyri at the bottom of the central sulcus, and it covers posterior half of the insula (Fig. 23.13).4 The anterior and posterior subcentral rami of the sylvian fissure delineate its anterior and posterior borders, respectively. The precentral gyrus, as a functional unit, has well-defined topological representations of different contralateral muscle groups. The muscle groups are represented in a precise but disproportionate manner on the precentral gyrus. The face, lip, hand, thumb, and index finger representations on the precentral gyrus are disproportionately larger than other parts. The face

215

216

III  Surgical Anatomy and Mapping Techniques and tongue areas have also bilateral representations. The postcentral gyrus representations of body areas are mirror images of the areas on the precentral gyrus. The lower part of the central lobe, including the opercular area, has face representation and covers an area up to 3 cm above the sylvian fissure. The pharynx, tongue, and jaw areas are located at the most ventral portion. The thumb area at the upper part of this segment is a transition region between the face and hand–arm areas that starts with thumb representation and ends with shoulder representation. The superior segment of the central lobe has trunk representation on the lateral surface and leg–foot representation on the mesial aspect of hemisphere. The anterior border of the medial superior segment of the central lobe is not well demarcated and merges with the SMA.4,​12,​30 Stimulation of the precentral gyrus gives rise to discrete and isolated contralateral movements. However, the areas representing the extraocular muscles, face, tongue, jaw, larynx, and pharynx may produce bilateral responses. Ablation or removal of an area on the motor cortex results in a flaccid paralysis with loss of all reflexes. On the other hand, removal of the tongue and face area frequently has little consequences because of bilateral cortical representation of these areas. Ablation of the postcentral gyrus results in loss of touch, pressure, pain, temperature senses. Pain, temperature, and light touch senses frequently return, but complete loss of discriminative touch and proprioception senses persist. As stated above, removal of the tongue and face area on the postcentral gyrus has little consequences but finger–hand–foot area resections cause significant impairment of proprioception sense.12,​29,​30,​34

Insula and the Peri-insular Region The insula (= island of Reil) and the peri-insular region have a complex anatomy that deserves special attention, especially in epilepsy surgery cases. The insula is hidden deep in the sylvian fissure as an invaginated cortical island. It is only visible when the sylvian fissure is widely open and the opercular cortices are retracted (Fig. 23.12 and Fig. 23.13). The sylvian fissure is the only true fissure on the lateral surface of the hemisphere and constitutes a major landmark. It can be divided as basal and lateral parts. Anteriorly located basal (= stem) segment extends from the lateral border of the anterior perforated substance and covers the limen insula by separating the lateral orbital gyrus and the temporal pole. The lateral part of the sylvian fissure has three segments with deep depressions on both ends: the anterior, horizontal, and posterior segments. The anterior and posterior segments, along the basal part of the fissure, extend deep into the cortical foldings and envelope the insula anteriorly and posteriorly. The anterior segment has two branches cutting into the inferior frontal gyrus: the horizontal and the vertical rami. The anterior ascending ramus continues with the circular sulcus around the insula. The horizontal segment of the sylvian fissure courses on the lateral surface of the hemisphere. The posterior segment turns up vertically by bifurcating into the ascending and descending rami on the surface and also extends deep into the upper posterior superior surface of the superior temporal gyrus by separating the Heschl’s gyri from the temporal planum.5,​6,​7,​8,​10,​12,​30,​35,​36,​37,​38,​39,​40 The insula is a paralimbic structure and part of the mesocortex. It lies mostly under the horizontal segment of the sylvian fissure and covered by the frontal, parietal, and temporal opercula. The insula has an inverted triangular pyramid shape

with the base on the top and the apex at the bottom. The limen insula makes the apex of the insular pyramid and connects the anterior and lateral surfaces of the insula to the anterior perforated substance and frontal lobe.5,​6,​7,​8,​10,​12,​30,​36,​37,​38,​39,​40 The orbitofrontal operculum covers the anterior surface of the insula, while the frontal, parietal and temporal opercula cover the lateral surface of the insula. The insula is surrounded by the anterior limiting, superior and inferior circular (= peri-insular) sulci circumferentially. The anterior limiting sulcus is not a true sulcus but an infolding separating the anterior surface of the insula from the posterior orbital gyrus. The superior circular sulcus is known as being the only sulcus on the lateral surface of the brain without an artery along its axis. The insular arteries cross the superior circular sulcus perpendicularly without entering the sulcal space itself.5,​8,​10,​12,​30,​35,​36,​38,​39,​40 The insula has lateral and anterior surfaces and forms the floor of the sylvian fissure to make the cortical coverage of the central core. The anterior part of the insula covers head of the caudate nucleus while the posterior part lies over the body of caudate and the thalamus. The anterior surface of the insula has one vertically oriented lateral gyrus (the accessory gyrus) and a short transversely oriented medial gyrus (the transverse gyrus of Eberstaller) that connects the limen insula to the posterior orbital gyrus and the lateral olfactory stria. The lateral surface of the insula is defined by a prominent central sulcus extending from the superior circular sulcus to the limen insula. The central sulcus of the insula obliquely traverses the insula from superior–posterior to inferior–anterior direction by dividing the lateral surface of the insula into anterior and posterior parts. The anterior part of the lateral surface has three short gyri and covers a larger area. The posterior part of the insula has two long, obliquely oriented gyri and covers a smaller part of the insular surface. These five insular gyri have a fan-like arrangement to fill triangular-shaped lateral insular surface. The short insular gyri composing the anterior part of the insula are named as the anterior, middle, and posterior short insular gyri and separated by the anterior and the precentral insular sulci. These three gyri are merged with the gyri of the anterior surface of the insula to form the insular apex. The anterior insula is connected exclusively to the frontal lobe. The posterior part of the insula is composed of the anterior and posterior long insular gyri that are separated by the postcentral insular sulcus. The posterior insula is connected to both the parietal and temporal lobes. The central sulcus of the insula has a relatively constant and continuous relation with overlying central sulcus of Rolando. These two sulci exhibit continuity without interruption in 60% of the cases.5,​6,​7,​8,​10,​12,​30,​35,​36,​37,​38,​39,​40 The insula has important physiological functions, including pain integration, motor planning of speech, auditory processing, vestibular function, gustatory sensation, visceral sensorimotor phenomena, cardiac function control, and even motor control.2,​35,​39,​40 The anterior insula participates motor function, including speech, and the posterior insula is associated with somatosensory processing. Stimulation of the anterior insula causes word-finding difficulties and speech apraxia especially on the left side. Stimulation of the insula also produces sensation of rotation and motion. Right-sided stimulations cause increased sympathetic tone, and left-sided stimulations cause increased parasympathetic tone.2,​29

23  Cerebral Cortex: Embryological Development and Topographical Anatomy

Temporal Lobe The temporal lobe has many unique features with a complex, distorted anatomy at the medial surface. It is the second largest lobe and covers more than 20% of the cerebral cortex including most of the basal surface of the cerebral hemisphere. The temporal lobe is also known as morphologically most heterogenous lobe. It contains the areas with histologically different cortical organizations including the major parts of the allocortex and mesocortex of the human cerebrum.5,​6,​7,​8,​12,​27,​30,​41,​42 The temporal lobe has a shape of pyramid with lateral, superior, medial, and basal surfaces as well as a pole (Fig. 23.8, Fig. 23.9, Fig. 23.10, and Fig. 23.11). The lateral surface of the temporal lobe is separated clearly from the frontal and parietal lobes by the sylvian fissure superiorly. However, its separation from the occipital and parietal lobes on the lateral ­surface and from the occipital lobe on the basal surface is quite arbitrary and defined by two imaginary lines: the parietotemporal and temporo-occipital lines. The parietotemporal line runs from superior end of the parieto-occipital fissure to the preoccipital notch and defines the posterior border between temporal and occipital lobes on the lateral surface. The temporo-occipital line runs from posterior end of the ­sylvian fissure to the midpoint of the parietotemporal line and defines the posterior border between the temporal and parietal lobes. The basal border between the temporal and occipital lobes is again defined by another imaginary line between the preoccipital notch and inferior end of the parieto-occipital fissure.5,​6,​7,​8,​12,​27,​30,​41,​42,​43

Lateral Surface The lateral surface of the temporal lobe starts from the sylvian fissure and extends toward the base of the temporal fossa. It has three gyri (the superior, middle, and inferior temporal gyri) separated by two parallel sulci (the superior and inferior temporal sulci). The superior temporal sulcus runs parallel to the sylvian fissure but is longer than that. It is a long, very well-defined, almost always uninterrupted and quite a deep (2.5–3 cm) sulcus. It reaches almost to the inferior end of the insular surface.12 Anterior tip of the superior temporal sulcus reaches to the temporal pole and its posterior end dives into the inferior parietal lobule. The angular gyrus caps the posterior end of this sulcus. The inferior temporal sulcus separates the middle and inferior temporal gyri. It is frequently a discontinuous sulcus and defining it might be difficult at times.5,​6,​7,​8,​10,​12,​30,​41,​42,​43 The superior temporal gyrus lies between the sylvian fissure and the superior temporal sulcus and has two surfaces: the lateral and superior surfaces. The lateral surface runs just below the sylvian fissure and extends all the way back to the inferior parietal lobule. It anteriorly merges with the inferior temporal gyrus to make the temporal pole and continues posteriorly with the supramarginal gyrus. Its superior surface continues with the anterior transverse temporal gyrus (= Heschl’s gyrus) in the sylvian fissure. The middle temporal gyrus is the widest and the most prominent gyrus on the lateral surface of the temporal lobe. It runs between the superior and inferior temporal sulci and always exhibits a wavy, interrupted topography. It barely reaches to the temporal pole anteriorly and merges with the angular gyrus and occipital lobe posteriorly without any clear boundary.

The inferior temporal gyrus is located between the inferior temporal and occipitotemporal sulci. Its upper part covers the lower end of the lateral surface of the temporal lobe, and its lower part extends into the basal surface of the temporal lobe. The inferior temporal gyrus is frequently (70%) discontinuous.12 Its anterior end merges into the temporal pole and posterior end continues with the inferior occipital gyrus.

Superior Surface Upper aspect of the superior temporal gyrus is hidden deep into the sylvian fissure and constitutes the superior surface of the temporal lobe. It is divided into three parts: the planum polare, the anterior transverse temporal gyrus (Heschl’s gyrus), and the planum temporale (Fig. 23.12). The planum polare is a flat area that covers the anterior part of the superior surface of the temporal lobe by extending between the uncus and the Heschl’s gyrus. It has two gyri: short and oblique. The Heschl’s gyrus is located between the planum polare and planum temporale as a single or two gyri separated by the intermediate transverse temporal sulcus. Its borders are defined by the anterior limiting sulcus of Holl (= the sulcus acusticus) anteriorly and the transverse subtemporal sulcus (= Heschl’s sulcus) posteriorly. It originates from midposterior edge of the superior temporal gyrus and dives into the sylvian fissure by crossing obliquely with an anterior-to-posterior direction. Its outer end is aligned with the base of the postcentral gyrus and inner end reaches to the lateral wall of the occipital atrium. The Heschl’s gyrus is larger and more obliquely oriented on the left side. The planum temporale is composed of two gyri: the middle and posterior transverse temporal gyri. It is a triangular-shaped area with an apex corresponding to the junction of the superior and inferior circular sulcus at the bottom of the sylvian fissure. Its posterior border is delineated by the posterior ascending ramus of the sylvian fissure.5,​6,​7,​8,​12,​30,​31,​41,​43

Basal Surface The temporal lobe has a large basal surface and its posterior border continues with the basal surface of the occipital lobe. Three gyri (lower half of the inferior temporal gyrus, the fusiform and parahippocampal gyri) cover the basal temporal surface (Fig. 23.10 and Fig. 23.11). The inferior and fusiform (= occipitotemporal gyrus) gyri are separated by the lateral occipitotemporal sulcus and the fusiform and the parahippocampal gyri are separated by the collateral (= the medial occipitotemporal) sulcus. The lateral occipitotemporal sulcus is frequently interrupted and its posterior end comes close to the preoccipital notch. The collateral sulcus is a well-defined, deep, and frequently uninterrupted sulcus. It is connected to the rhinal sulcus anteriorly and the calcarine sulcus posteriorly. Bottom of the collateral sulcus reaches to the temporal horn and makes an impression (the collateral eminence) in the temporal horn. The rhinal sulcus is a short sulcus and delineates the border between the uncus and the fusiform gyrus.5,​6,​7,​8,​10,​12,​27,​30,​41,​43 Although defining its precise borders are not easy, the primary auditory cortex is located on the anterior transverse ­temporal gyrus (the Heschl’s gyrus), and auditory association area is located on the planum temporale. There is possibly a different tonotopic organization within the Heschl’s gyrus; high frequencies are represented medially and low frequencies laterally.44 The Heschl’s gyrus and the planum temporale are general-

217

218

III  Surgical Anatomy and Mapping Techniques ly wider and longer on the left side. The posterior part of the superior temporal gyrus, the Heschl’s gyrus, and the angular gyrus area (Wernicke’s area) participate the functions related to the comprehension of spoken and written language. The temporal association cortex includes the middle and inferior temporal gyri and plays a role on remembering names and words. The inferotemporal region includes visual association areas plays a role in recognizing visual stimuli and shape of the objects. Stimulation of the posterior part of the Heschl’s gyrus produces buzzing, humming, and tinnitus type of sensations. Lesion in this area causes decreased hearing bilaterally and also causes impairment of sound localization in space. Lesion in Wernicke’s area in dominant hemisphere causes receptive aphasia.29

Medial Temporal Lobe The medial temporal lobe is separated from the rest of the temporal lobe by the collateral and rhinal sulci (Fig. 23.10 and Fig. 23.11). This region includes the parahippocampal gyrus, uncus, hippocampus, fimbria, dentate gyrus, and amygdala. The cortical structures belong to the medial temporal region were defined by several short and shallow sulci and one very prominent sulcus, collateral sulcus. These are conspicuous surface landmarks to define the structures in this region. The rhinal sulcus is a shallow sulcus in humans and defines the anterolateral border of the mesial temporal structures. On the other hand, collateral sulcus is a quite deep and prominent sulcus and defines the entire lateral border of mesial temporal region.5,​6,​7,​8,​10,​12,​30,​41,​42,​43,​45,​46 The mesial temporal lobe has a notoriously complex cortical anatomy. Although it is part of the temporal lobe, its connections are very different than the rest of the temporal lobe. It

is connected intimately to the cingulate gyrus through the i­sthmus, to the insula through the temporal stem, to the globus pallidus through the amygdala, and to the basal frontal lobe through the limen insula. The mesial temporal structures and some other embryologically related and anatomically connected brain structures create a functional network and collectively named as the “limbic lobe.” The limbic lobe is composed of a series of well-connected neural structures. The “limbic gyrus” is the main connecting structure of the limbic lobe and extends from the primary olfactory cortex to the uncus by encircling the corpus callosum and the upper brain stem by following a distorted C-shaped “limbic fissure.” The limbic fissure is formed by several separate, discontinuous sulci including the subcallosal, cingulate, subparietal, anterior calcarine, collateral, and rhinal sulci. The limbic gyrus is composed of the subcallosal, cingulate, isthmus, and parahippocampal gyri. It is also connected to the hippocampus, amygdala, and temporal pole. The limbic lobe bridges the phylogenetic gap between the neocortex and the primitive central areas.5,​6,​7,​8,​10,​12,​30,​41,​42,​43,​45,​46

Uncus The anterior mesial temporal region includes the uncus and the entorhinal cortex. The uncus is formed by a medial-backward bend of the rostral hippocampus or parahippocampus (Fig. 23.14). The uncus extends medially and then curves posteriorly to rest on the parahippocampus. This distortion creates a hook-like convolution (= uncus, in Latin). Its posterior third is formed by the hippocampus, and its anterior two-thirds is formed by the entorhinal gyrus ventrally, the periamygdaloid cortex dorsally, and the amygdala laterally. The uncal

Fig. 23.14  Medial basal view of the left temporal lobe. a, fimbria; b, gyrus dentatus; c, collateral sulcus; d, uncal sulcus; e, rhinal sulcus; 1, apex of uncus; 2, band of Giacomini; 3, gyrus uncinatus; 4, gyrus ambiens; 5, semilunar gyrus; 6, entorhinal area; 7, parahippocampus; 8, isthmus; 9, lingual gyrus; 10, fusiform gyrus; 11, splenium.

23  Cerebral Cortex: Embryological Development and Topographical Anatomy sulcus separates the uncus from the parahippocampal gyrus anteriorly. The uncus is a conical structure with an apex made by its merging anterior and posterior segments. Its anteromedial surface face the carotid cistern.12,​30,​41,​43,​46 The anterior segment of uncus contains two visible protrusions on the surface: the semilunar and ambient gyri. The semiannular sulcus separates these two gyri. The semilunar gyrus covers the cortical nucleus of the amygdala, and it is separated by the entorhinal sulcus superolaterally and the semiannular sulcus anteromedially. The ambient gyrus is located inferiorly, and it is mainly formed by medially protruding entorhinal cortex. It covers the amygdala. The ambient gyrus has a visible impression on it secondary to the mechanical effect of the tentorial edge: the uncal notch. The posterior segment of the uncus is divided into the inferior and superior parts by the uncal sulcus. The inferior part is formed by the entorhinal area, and its lateral border is defined by the rhinal sulcus anteriorly and the collateral sulcus posteriorly. The superior part of the posterior uncal segment is formed by the hippocampal head. The velum terminale connects the rostral ends of the fimbria and the stria terminalis at the roof of the temporal horn and defines the posterior border of the uncus. This part of the uncus has three small gyri: the uncinate gyrus, the band of Giacomini, and the intralimbic gyrus. The uncus continues along the globus pallidus at its superior surface without any defined borders.8,​12,​30,​41,​43,​46,​47

The parahippocampus is also described by various subdivisions; the subiculum, presubiculum, parasubiculum, and entorhinal areas. These areas do not have anatomically well-defined borders. The entorhinal area is anteriorly located and merges with the periamygdaloid cortex dorsomedially and with the presubiculum caudomedially. Its posterior extension is not clearly defined. The subiculum is located in posterior part. It covers the most medial aspect of the parahippocampus and known as “the bed of the hippocampus.” It is a transitional area from six-layer to three-layer cortex. Superiorly, the parahippocampal gyrus is separated from the dentate gyrus by the hippocampal sulcus.5,​6,​7,​8,​10,​12,​30,​41,​42,​45,​46,​47,​48

Hippocampus The hippocampus occupies the floor and the medial surface of the temporal horn of lateral ventricle (Fig. 23.13, Fig. 23.14, and Fig. 23.15). The extraventricular aspect of the hippocampus makes the medial surface of the temporal lobe and includes the dentate gyrus, fimbria fornix, and hippocampal sulcus. The hippocampus rests on the subiculum of the parahippocampus

Amygdala The amygdala, along the hippocampus, constitutes one of the two major telencephalic components of the limbic system. It is mainly contained within the uncus and located dorsal to the hippocampus and rostral to the tip of the temporal horn. Its borders have no clear demarcations. Its medial and anterosuperior aspects are contained by the uncus, and its anteroinferior aspect is related to the entorhinal area. The amygdala’s posteroinferior portion fuses with the hippocampal head, and its posterosuperior portion blends into the claustrum and the globus pallidus by coming close to contact with optic tract. Grossly, the a ­ mygdala is recognized with its relatively brownish, hazelnut-like color. The amygdala is composed of the central, corticomedial, and basolateral groups of nuclei and has extensive connections with various regions. It receives cortical p ­ rojections from various limbic and associated cortical areas, and its main output channels are the stria terminalis and the ventral amygdalofugal pathway.5,​8,​12,​30,​41,​42,​45,​46,​48

Parahippocampus The parahippocampal gyrus constitutes a transitional area between the basal and the mesial surfaces of the temporal lobe (Fig. 23.10, Fig. 23.11, and Fig. 23.14). It starts from the uncus, runs along the tentorial edge, and reaches all the way back to the calcarine fissure. It continues at this junction with the lingual gyrus inferiorly and the isthmus of the cingulate gyrus superiorly. Its lateral border is defined by the rhinal and collateral sulci. The parahippocampus can be divided into the anterior and posterior segments. The anterior segment covers a larger area, and it is also called the piriform lobe that includes the entorhinal area and the uncus. The posterior segment is narrower and has a flat superior surface (the subiculum) that is separated from the hippocampus by the hippocampal sulcus.

Fig. 23.15  Intraventricular aspect of the hippocampus after temporal horn has been opened and the choroid plexus removed. 1, hippocampal body; 2, head and digitationes hippocampi (internal digitations); 3, hippocampal tail; 4, fimbria; 5, crus of fornix; 6, subiculum; 7, splenium of the corpus callosum; 8, calcar avis; 9, collateral trigone; 10, collateral eminence; 11, uncal recess of the temporal horn. (Reproduced with permission from Duvernoy 1998.46)

219

220

III  Surgical Anatomy and Mapping Techniques and formed by two interlocking U-shaped cortical laminae: Ammon’s horn (= hippocampus proper) and dentate gyrus (= fascia dentate). The hippocampal sulcus starts from the uncus anteriorly and separates the dentate gyrus of the hippocampus from the subiculum. The dentate gyrus accompanies to the fimbria fornix until it reaches to the splenium. The dorsal and lateral aspects of the hippocampus are located in the temporal horn and covered by a lamina of white matter fibers: the alveus. Its medial aspect is related to the choroidal fissure. The alveus converges to the fimbria fornix when it reaches to the choroidal fissure and constitutes the principal output channel of the hippocampus. The fimbria is a flattened band of fibers and lies superior to the dentate gyrus. The fimbria fornix starts from the anterior– inferior end of the choroidal fissure (= inferior choroidal point) and becomes thicker as it proceeds backward. The inferior choroidal point is a significant landmark that marks the posterior border of the uncus. The anterior choroidal artery enters into the temporal horn, and the inferior temporal vein exits the temporal horn at the inferior choroidal point.5,​6,​7,​8,​12,​30,​41,​42,​45,​46,​47,​48 The hippocampus is divided into three parts: head, body, and tail. Total length of the hippocampus, from head to tail, is approximately 3.5 to 4 cm in adults.46,​49 The head of the hippocampus is located anterior to the hippocampal fissure and it merges with the amygdala. It is transversely oriented and contains a large area (1.5–2 cm in width).46,​49 It is the only region of the hippocampus that is not covered by the choroid plexus. It has several digitations on its ventricular surface. The junction of the head and body of the hippocampus corresponds the inferior choroidal point (= velum terminale of Aeby), where the uncus appears and the taeniae of the fimbria fornix and stria terminalis unite. The body of the hippocampus is sagittally oriented and is about 1 cm in width. It starts from the inferior choroidal point and covered by the choroid plexus unlike the head of the hippocampus. The collateral eminence defines the intraventricular lateral border of the hippocampus. The hippocampal tail does not have a clear-cut starting point. It starts approximately at the pulvinar level of the posterior intraventricular region. The tail of the hippocampus is separated from the body when the hippocampus proper becomes thinner and the fimbria becomes thicker. The dentate gyrus and the fimbria fornix take different destinations at the tail. While the dentate gyrus runs toward the supracallosal area and becomes the indusium griseum, the fimbria fornix takes an infracallosal course toward the mamillary bodies.5,​6,​7,​8,​12,​30,​ 41,​42,​45,​46,​47,​48

The medial temporal lobe, mainly the parahippocampus or hippocampus, plays a critical role in learning, spatial memory and consolidating memories into lasting memories. The amygdala plays a role in the emotional, motivational, autonomic, and behavioral responses. Stimulation of the amygdala induces negative emotions such as fear, sadness, anxiety, and defensive, even violent behaviors.

Parietal Lobe The parietal lobe has a quite variable topography with several poorly defined serpiginous gyri (Fig. 23.7 and Fig. 23.9). Although its anterior and medial borders are well defined, its borders with occipital and temporal lobes are arbitrary. The postcentral sulcus defines the anterior border of the parietal lobe, and an arbitrary line between the preoccipital notch and the dorsal end of the parieto-occipital sulcus defines its posterior border on the lateral surface. The parietal lobe is divided by a very deep sulcus: the intraparietal sulcus. The intraparietal sulcus divides the lateral surface of the parietal lobe into the superior and inferior lobules. It is a quite deep sulcus and almost reaches to the lateral ventricle in some cases. It runs parallel to the interhemispheric fissure. Anteriorly, it originates from approximately at the midpoint of the postcentral sulcus and frequently continues with both the inferior and superior postcentral sulci (40%) and posteriorly merges with the occipital sulcus.4,​5,​6,​7,​8,​10,​12,​30 The parieto-occipital sulcus is a very well-defined, constant, and deep sulcus on the medial surface of the hemisphere. It starts from the dorsal margin of the hemisphere, extends downward toward the splenium, and reaches to the calcarine fissure. It separates the precuneus from the cuneus. Its dorsal end is seen as a notch at the edge of the hemisphere. The precuneus constitutes the medial aspect of the superior parietal lobule with well-defined borders including the marginal ramus of the cingulate sulcus, the subparietal sulcus, and the parieto-­ occipital sulcus. The inferior parietal lobule contains the supramarginal, angular, and postparietal gyri. The supramarginal gyrus encircles the terminal ascending ramus of the sylvian fissure, and the angular gyrus encircles the ascending ramus of the superior temporal sulcus. The supramarginal and angular gyri are separated by the intermediate sulcus of Jensen and together they make the parietal tuberosity. The supramarginal gyrus caps the posterior end of the sylvian fissure and anteriorly continues with the postcentral gyrus, inferiorly with the superior temporal and the angular gyri. On the other hand, the angular gyrus circles the posterior end of the superior temporal sulcus and continues with the middle temporal gyrus anteriorly and the middle occipital gyrus posteriorly. The supramarginal and the angular gyri along with the posterior parts of the superior and middle temporal gyri correspond the Wernicke’s receptive speech area. The posterior speech area also extends into the Heschl’s gyrus and the planum temporale.4,​5,​12,​37 The angular gyrus and surrounding areas on the dominant hemisphere contain the major association cortex that functions as a higher-order, complex multisensory perception area. It plays a significant role on communication skills, and its lesions cause receptive and expressive aphasia, inability to write (agraphia), inability to recognize multisensory perceptions (agnosia), left–right confusion, difficulty in recognizing different fingers (finger agnosia), and inability to c­ alculate (acalculia). This complex set of impairments is called

23  Cerebral Cortex: Embryological Development and Topographical Anatomy the Gerstmann syndrome. Unilateral lesions in nondominant parietal lobe produce neglect of contralateral half of body and visual space. Bilateral lesions of this area cause inability to move hands toward an object that is clearly seen.29

Occipital Lobe The occipital lobe is the smallest lobe situated posterior to an imaginary line between the parieto-occipital notch and the temporo-occipital incisura on the lateral surface of the hemisphere (Fig. 23.9, Fig. 23.10, and Fig. 23.11). Its topography shows a wide range of gyral variations with many shallow sulci. The occipital gyri of the lateral surface are difficult to characterize morphologically, and they are loosely defined as the superior (O1), middle (O2), and inferior (O3) occipital gyri and fourth occipital gyri in some cases. The intraoccipital sulcus continues between O1 and O2 as the posterior extension of the intraparietal sulcus. All three gyri merge and make the occipital pole. The borders on the medial surface of the occipital lobe are quite well defined with anatomically constant sulci and gyri (Fig. 23.10). The main sulcus in the medial aspect of the occipital lobe is the calcarine fissure. It starts just below the isthmus of the cingulate gyrus and the lingual gyrus junction and extends just above the inferomedial border of the hemisphere toward the occipital pole. The parieto-occipital sulcus is ­connected to the calcarine fissure at a point close to the

i­ sthmus. It divides the calcarine fissure into proximal and distal parts. It is a quite deep fissure (2.5–3 cm) and made an impression onto occipital horn to form the calcar avis.4,​12 The calcarine fissure and the parieto-occipital sulcus are separated with the cuneolingual gyri. The calcarine fissure divides the medial occipital lobe into wedge-shaped cuneus and the lingual gyri that continue with the parahippocampal gyrus. The basal surface of the occipital lobe continues with the basal surface of the temporal lobe without any delineated border (Fig. 23.11). The posterior portion of the fusiform gyrus continues as the inferior occipital gyrus.4,​5,​6,​7,​8,​10,​12,​30 The primary visual cortex is located around the calcarine fissure, more specifically on the cuneal and the lingual surfaces of its distal part. Stimulation of the visual cortex produces contralateral sensation of bright flashes of light. The lesions involving the inferior (lingual) calcarine cortex produce upper contralateral quadrantanopia, and the lesions involving the superior (cuneal) calcarine cortex produce inferior contralateral quadrantanopia. The visual association cortex is located around the primary visual cortex and covers a larger area. Its functions include spatial perception, visual localization and identification of objects, and controlling the eye movements. There are secondary visual cortical areas in the basal occipital lobe, and bilateral lesions of the occipitotemporal junction cause failure to recognize familiar faces (prosopagnosia). Unilateral lesion of the inferior occipitotemporal cortex causes loss of color vision on the contralateral half of visual field (hemichromatopsia).12,​29

221

222

III  Surgical Anatomy and Mapping Techniques

„„ Addendum: Topographical Anatomy of the Cerebral Cortex on MRI

yy

Oğuz Çataltepe and Sathish Kumar Dundamadappa High-resolution MRI provides excellent topographical images of the cerebral cortex. In addition to this, many anatomical landmarks and signs to identify functionally significant sulci and gyri on MRI have been described to facilitate interpretation of MRI images.50,​51,​52,​53,​54,​55,​56,​57,​58,​59,​60,​61 Here, we describe well-established signs and landmarks on magnetic resonance images that can be helpful to locate the main cerebral sulci and gyri. However, variations in cortical anatomy are common. Only some of the imaging signs or landmarks are definitive enough for identification of anatomy by their sole use (e.g., differential thickness of cortex on either side of the central sulcus, hand knob, etc.). Multiple signs should be used in combination for identifying the anatomy on imaging.

yy

yy

Frontal Lobe

yy

yy Central sulcus (Fig. 23.16): A deep sulcus reaching up to interhemispheric fissure (arrow). yy Precentral knob: Corresponds to the hand motor area, seen on axial images as either “‘inverted omega” (Fig. 23.16; green

yy

yy

Fig. 23.16  The central sulcus (black arrow), hand motor area (inverted omega sign; green line-1 and horizontal epsilon sign; green line-2), pars marginalis (bracket sign; green line-3), upper T sign (green line-4), L sign (blue line-5), bifid postcentral gyrus sign (blue line-6), parietal T sign (green line-7).

line-1), or less commonly “horizontal epsilon” (Fig. 23.16; green line-2). In sagittal plane, this is seen as a posteriorly directed hook at the level of the insula (Fig. 23.17; red star). Differential thickness of pre- and postcentral gyri: The sagittal dimension of precentral gyrus (Fig. 23.17; red star) is thicker than that of postcentral gyrus (Fig. 23.17; red circle), approximately 1.5 times. The posterior cortex of the precentral gyrus (primary motor area) (Fig. 23.18; white arrow) is thicker than the corresponding anterior cortex of the postcentral gyrus (Fig. 23.18; black arrow). Pars marginalis: The posterior portion of the cingulate sulcus (Fig. 23.19; green line-1), known as pars marginalis extends upward to reach cerebral margin. In the axial plane, past marginalis from both hemispheres appear as horizontal bracket, :“bracket sign” (Fig. 23.16; green line-3). The medial end of the central sulcus is immediately anterior to this bracket (Fig. 23.16; arrow). In sagittal plane, the medial end of the central sulcus, which is located just anterior to the pars marginalis, appears as posteriorly curved notch (Fig. 23.19; blue line-2). The “upper T sign”: The intersection of easily identifiable superior frontal sulcus with the precentral sulcus has the shape of letter “T” (Fig. 23.16; green line-4). The “lower T sign” (Fig. 23.20): The inferior frontal sulcus posteriorly terminates in the precentral sulcus, also in the shape of letter “T” (green line), that is dorsally bordered by precentral gyrus. The “L sign”: The superior frontal gyrus terminates in the precentral gyrus, which turns laterally from the posterior end of the superior frontal gyrus, together forming “L” (Fig. 23.16; blue line-5: will appear as flipped “L” on the right side). M shape of the inferior frontal gyrus: The anterior horizontal (ahr) and ascending rami (ar) of sylvian fissure extend into inferior frontal gyrus giving the shape of letter “M” (Fig. 23.21; green line-1). The first vertical line of the M represents pars orbitalis, the middle portion represents pars triangularis, and the posterior vertical line represents pars opercularis. Note, a small descending ramus branch of inferior frontals sulcus projecting into pars triangularis. This “M”

Fig. 23.17  The Heschl’s gyrus (HG), planum temporale (PT), central sulcus of insula (cs-i), short gyri (s), long gyri (L), insular apex (green star), accessory gyrus (ag), precentral knob (green line-red star), postcentral gyrus (red circle).

23  Cerebral Cortex: Embryological Development and Topographical Anatomy terminates posteriorly and medially into precentral gyrus. yy The “U sign”: Identified in sagittal plane. At the level of inferior frontal gyrus, the first continuous sulcus that is enclosed by a perisylvian U-shaped gyrus (subcentral gyrus) (Fig. 23.21; green line-2) can be identified as central sulcus. yy Inferior surface of frontal lobe (Fig. 23.22): The medial most gyrus is the gyrus rectus (GR), which is bordered by olfactory sulcus (os). Roughly “H-shaped” orbital sulci (green line) divide the orbital gyri into medial (MOG), lateral (LOG), anterior (AOG), and posterior orbital gyri (POG).

Fig. 23.18  The posterior cortex of the precentral gyrus (white arrow) is thicker than corresponding anterior cortex of the postcentral gyrus (black arrow).

Fig. 23.20  The “lower T sign” of the inferior frontal sulcus (green line).

Fig. 23.19  The cingulate sulcus (green line-1), central sulcus (blue line-2), parieto-occipital sulcus (pos), precuneus (PC), cuneus (C), lingual gyrus (LG), isthmus (Isth), and anterior calcarine sulcus (acs).

Fig. 23.21  The M sign (green line-1), U sign (green line-2), central sulcus (Cs), supramarginal gyrus (green line-3), Heschl’s gyrus (blue line-HG), planum temporale (red line), the sylvian fissure’s anterior horizontal ramus (ahr), anterior ascending ramus (ar), posterior horizontal ramus (phr), posterior ascending ramus (par), anterior and posterior subcentral sulci (scs) and transverse temporal sulci (black arrow).

223

224

III  Surgical Anatomy and Mapping Techniques

Fig. 23.23  The superior frontal gyrus (SFG), cingulate gyrus (CG), cingulate sulcus (cing), paracentral lobule (PL), pars marginalis (curved arrow) of the cingulate sulcus, paracentral sulcus (straight arrow), parieto-occipital sulcus (pos), precuneus (PC), cuneus (C), calcarine sulcus (cal), anterior calcarine sulcus (acs), lingual gyrus (LG). Fig. 23.22  The gyrus rectus (GR), olfactory sulcus (os), “H-shaped” orbital sulci (green line), medial orbital gyri (MOG), lateral orbital gyri (LOG), anterior orbital gyri (AOG), and posterior orbital gyri (POG).

Parietal Lobe yy The bifid postcentral gyrus sign: Toward interhemispheric fissure, the postcentral gyrus is medially fissured, and enclosing the pars marginalis with both legs (Fig. 23.16; blue line-6). yy Intraparietal sulcus: The anterior end of the intraparietal sulcus tends to perpendicularly intersect the postcentral sulcus (Fig. 23.16; green line-7). This sulcus divides lateral aspect of the parietal lobe into the superior and inferior parietal lobules. yy Supramarginal gyrus: Horseshoe-shaped gyrus around the posterior ascending ramus of the sylvian fissure (Fig. 23.21; green line-3). yy Angular gyrus (Fig. 23.24): Horseshoe-shaped gyrus (green line) around the posterior end of superior temporal sulcus (sts). yy Medial surface of parietal lobe (Fig. 23.23): Prominent parieto-occipital sulcus (pos) marks division of parietal and occipital lobes in the medial surface and traverses nearly parallel to pars marginalis. Roughly H-shaped subparietal sulcus (green lines) marks the inferior limit of precuneus (PC) and separates it from the cingulate gyrus (CG).

Temporal Lobe Fig. 23.24  Angular gyrus (horseshoe-shaped, green line), superior temporal sulcus (sts).

yy Medial surface of frontal lobe (Fig. 23.23): Interhemispheric part of superior frontal gyrus (SFG) and cingulate gyrus (CG), separated by cingulate sulcus (cing). Paracentral lobule (PL) lies just anterior to pars marginalis (curved arrow) of cingulate sulcus and is bordered anteriorly by paracentral sulcus (straight arrow).

yy Lateral and inferior surfaces (Fig. 23.25): Superior (sts) and inferior (its) temporal sulci divide the lateral surface into superior temporal gyrus (S), middle temporal gyrus (M), and inferior temporal gyrus (I). Inferior temporal gyrus also forms portion of lateral aspect of the basal surface. Collateral sulcus (cls) and lateral occipitotemporal sulcus (lots) divide the basal surface into parahippocampal gyrus (PH), fusiform gyrus, also known as occipitotemporal gyrus (F) and inferior temporal gyrus (I). Hippocampal (H) lies over parahippocampus.

23  Cerebral Cortex: Embryological Development and Topographical Anatomy

Fig. 23.25  The superior temporal gyrus (S), superior temporal sulcus (sts), middle temporal gyrus (M), inferior temporal sulcus (its), inferior temporal gyrus (I), lateral occipitotemporal sulcus (lots), fusiform gyrus (= occipitotemporal gyrus) (F), collateral sulcus (cls), parahippocampal gyrus (PH), hippocampus (H), Heschl’s gyrus (green line-HG), planum temporale (red line-PT).

yy Heschl’s gyrus: In sagittal plane (Fig. 23.17 and Fig. 23.21) just lateral to the insula, it is easily identified given the characteristic shape. Depending on the number of Heschl’s gyri, it has the appearance of a single omega (Fig. 23.17; green line-HG; Fig. 23.21; blue line-HG), a mushroom, a heart or double omega shape. In the axial plane, it is seen at the level of massa intermedia of the thalamus as gyrus or gyri coursing in an anterolateral direction (Fig. 23.26; arrow, HG). In the coronal plane, it is identified at the level of forniceal convergence and seventh cranial nerves (Fig. 23.25; green line-HG). Planum temporale (Fig. 23.25; red line-PT; Fig. 23.26; PT) refers to flat superior surface of temporal lobe that extends from Heschl’s sulcus at the posterior border of Heschl’s gyrus to posterior end of the sylvian fissure.

Insular Region yy Sylvian fissure (Fig. 23.21): Five major rami are anterior horizontal ramus (ahr), anterior ascending ramus (ar), posterior horizontal ramus (phr), posterior ascending ramus (par), and posterior descending ramus (not shown). Minor arms include anterior and posterior subcentral sulci (scs) and transverse temporal sulci (black arrow). yy Insular gyri (Fig. 23.17): The central sulcus of insula (cs-i) obliquely traverses the insula from superoposterior to

Fig. 23.26  Heschl’s gyrus (HG and arrow), planum temporale (PT).

inferoanterior direction. The anterior part has three short gyri (s) and the posterior part has two long obliquely oriented gyri (L). The three short gyri converge in the anteroinferior aspect to form the insular apex (green star). Accessory gyrus (ag) and transversely oriented medial gyrus make the anterior face of the insula.

Occipital Lobe yy Medial surface (Fig. 23.19 and Fig. 23.23): Parieto-occipital sulcus (pos) separates precuneus (PC) of parietal lobe anteriorly from the cuneus (C) of occipital lobe posteriorly. It courses parallel to pars marginalis, joins with anterior end of calcarine sulcus (cal), and continues anteriorly as anterior calcarine sulcus (acs). Calcarine sulcus separates cuneus superiorly from medial temporo-occipital gyrus (lingual gyrus, LG) inferiorly. Anterior calcarine sulcus separates the cingulate gyrus anteriorly from the lingual gyrus posteriorly.

„„ Acknowledgment All anatomical specimens (Fig. 23.7–Fig. 23.14) used in this chapter were prepared by Prof. Dr. Uğur Türe, and I am grateful for his courtesy and generosity.

References 1. Sun A, Hou LC, Cheshier SH, et al. The accuracy of topographical methods in determining central sulcus: a statistical correlation between modern imaging data and these historical predications. Cureus 2014;6(6):e186

2. Ulmer OJ. Neuroanatomy and cortical landmarks. In: Ulmer S, Jansen O, eds. fMRI—Basics and Clinical Applications. Berlin-­ ­ Heidelberg: Springer-Verlag; 2013:7–16

225

226

III  Surgical Anatomy and Mapping Techniques 3. Frigeri T, Paglioli E, de , Oliveira E, Rhoton AL Jr. Microsurgical anatomy of the central lobe. J Neurosurg 2015;122(3):483–498 4. Ribas GC. The cerebral sulci and gyri. Neurosurg Focus2010;28(2):E2 5. Yaşargil MG. Topographic anatomy for microsurgical approaches to intrinsic brain tumors. In: Yaşargil MG, ed. Microneurosurgery IVA. New York, NY: Thieme; 1994:2–115 6. Olivier A, Boling WW. Surgical Anatomy. In: Olivier A, Boling WW, Tanriverdi T, eds. Techniques in Epilepsy Surgery. Cambridge: Cambridge University Press; 2012:30–40 7. Boling W, Olivier A. Anatomy of important cortex. In: Byrne RW, ed. Functional Mapping of the Cerebral Cortex. Heidelberg: Springer; 2016:23–40 8. Ribas GC. The microneurosurgical anatomy of cerebral cortex. In: Duffau H, ed. Brain Mapping: from Neural Basis of Cognition to Surgical Application. Wien, New York: Springer; 2012:7–26 9. Creutzfeldt OD. Phylogenetic, ontogenetic, and functional development of the cerebral cortex and the general structural organization of the neocortex. In: Creutzfeldt OD, ed. Cortex cerebri: Performance, Structural and Functional Organization of the ­Cortex. New York, Oxford: Oxford University Press; 1995:9–89 10. Ono M, Kubick S, Abernathy CD. Atlas of the cerebral sulci. ­Stuttgart, New York: Georg Thieme Verlag; 1990 11. Glasser MF, Coalson TS, Robinson EC, et al. A multi-modal parcellation of human cerebral cortex. Nature 2016;536(7615):171–178 12. Tamraz JC, Comair YG. Atlas of regional anatomy of the brain using MRI with functional correlations. Berlin, Heidelberg: ­Springer-Verlag; 2000 13. Alkadhi H, Kollias SS. Pli de passage fronto-pariétal moyen of broca separates the motor homunculus. AJNR Am J Neuroradiol 2004;25(5):809–812 14. Dubois J, Benders M, Cachia A, et al. Mapping the early cortical folding process in the preterm newborn brain. Cereb Cortex 2008;18(6):1444–1454 15. White T, Su S, Schmidt M, Kao CY, Sapiro G. The development of gyrification in childhood and adolescence. Brain Cogn 2010;72(1):36–45 16. Ribas GC, Yasuda A, Ribas EC, Nishikuni K, Rodrigues AJ Jr. Surgical anatomy of microneurosurgical sulcal key points. ­ ­Neurosurgery 2006;59(4, Suppl 2):ONS177–ONS210, discussion ONS210–ONS211 17. Gilmore JH, Shi F, Woolson SL, et al. Longitudinal development of cortical and subcortical gray matter from birth to 2 years. Cereb Cortex 2012;22(11):2478–2485 18. Nishikuni K, Ribas GC. Study of fetal and postnatal morphological development of the brain sulci. J Neurosurg Pediatr 2013;11(1):1–11 19. Marin-Padilla M. Human motor cortex: development and cytoarchitecture. In: Marin-Padilla M, ed. The Human Brain: Prenatal Development and Structure. Berlin, Heidelberg: Springer-Verlag; 2011:11–34 20. Donkelaar HJT. Fetal development of the brain. In: Donkelaar HJT, Lammens M, Hori A, eds. Clinical Neuroembryology. Berlin, ­Heidelberg, New York: Springer; 2006 21. Stiles J, Jernigan TL. The basics of brain development. Neuropsychol Rev 2010;20(4):327–348 22. Chi JG, Dooling EC, Gilles FH. Gyral development of the human brain. Ann Neurol 1977;1(1):86–93 23. Alexander-Bloch A, Giedd JN, Bullmore E. Imaging structural co-variance between human brain regions. Nat Rev Neurosci 2013;14(5):322–336 24. Lenroot RK, Giedd JN. Brain development in children and adolescents: insights from anatomical magnetic resonance imaging. Neurosci Biobehav Rev 2006;30(6):718–729 25. Gogtay N, Giedd JN, Lusk L, et al. Dynamic mapping of human ­cortical development during childhood through early adulthood. Proc Natl Acad Sci USA 2004;101(21):8174–8179 26. Vasung L, Fischi-Gomez E, Hüppi PS. Multimodality evaluation of the pediatric brain: DTI and its competitors. Pediatr Radiol 2013;43(1):60–68

27. Gloor P. Comparative anatomy of the temporal lobe and of the limbic system; macroscopic anatomy of temporal isocortex and some adjacent mesocortical areas; macroscopic anatomy and morphology of the human mesial temporal region. In: Gloor P, ed. The Temporal Lobe and Limbic System. New York, Oxford: Oxford University Press; 1997:21–112, 113–157, 326–348 28. Maudgil DD, Free SL, Sisodiya SM, et al. Identifying homologous anatomical landmarks on reconstructed magnetic resonance images of the human cerebral cortical surface. J Anat 1998;193:559–571 29. Afifi AK, Bergman RA. Functional Neuroanatomy. New York, NY: McGraw Hill; 1998:37–58 30. Duvernoy H. Surface anatomy of the brain. In: Duvernoy H, ed. The Human Brain: Surface, Three-Dimensional Sectional Anatomy and MRI. New York, NY: Springer; 1991:3–45 31. Wen HT, Rhoton AL Jr, de , Oliveira E, Castro LH, Figueiredo EG, Teixeira MJ. Microsurgical anatomy of the temporal lobe: part 2— sylvian fissure region and its clinical application. Neurosurgery 2009;65(6, Suppl):1–35, discussion 36 32. Bozkurt B, Yagmurlu K, Middlebrooks EH, et al. Microsurgical and tractographic anatomy of the supplementary motor area complex in humans. World Neurosurg 2016;95:99–107 33. Sindou M, Guenot M. Surgical anatomy of the temporal lobe for epilepsy surgery. Adv Tech Stand Neurosurg 2003;28:315–343 34. Boling W, Reutens DC, Olivier A. Functional topography of the low postcentral area. J Neurosurg 2002;97(2):388–395 35. Afif A, Mertens P. Description of sulcal organization of the insular cortex. Surg Radiol Anat 2010;32(5):491–498 36. Türe U, Yaşargil MG, Al-Mefty O, Yaşargil DCH. Arteries of the insula. J Neurosurg 2000;92(4):676–687 37. Chen CY, Zimmerman RA, Faro S, et al. MR of the cerebral operculum: topographic identification and measurement of interopercular distances in healthy infants and children. AJNR Am J Neuroradiol 1995;16(8):1677–1687 38. Tanriover N, Rhoton AL Jr, Kawashima M, Ulm AJ, Yasuda A. Microsurgical anatomy of the insula and the sylvian fissure. J Neurosurg 2004;100(5):891–922 39. Naidich TP, Kang E, Fatterpekar GM, et al. The insula: anatomic study and MR imaging display at 1.5 T. AJNR Am J Neuroradiol 2004;25(2):222–232 40. Türe U, Yaşargil DCH, Al-Mefty O, Yaşargil MG. Topographic anatomy of the insular region. J Neurosurg 1999;90(4):720–733 41. Al-Otaibi F, Baeesa SS, Parrent AG, Girvin JP, Steven D. Surgical techniques for the treatment of temporal lobe epilepsy. Epilepsy Res Treat 2012;2012:374848 42. Wen HT, Rhoton AL Jr, de Oliveira E, et al. Microsurgical anatomy of the temporal lobe: part 1: mesial temporal lobe anatomy and its vascular relationships as applied to amygdalohippocampectomy. Neurosurgery 1999;45(3):549–591, discussion 591–592 43. Kucukyuruk B, Richardson RM, Wen HT, Fernandez-Miranda JC, Rhoton AL Jr. Microsurgical anatomy of the temporal lobe and its implications on temporal lobe epilepsy surgery. Epilepsy Res Treat 2012;2012:769825 44. Leaver AM, Rauschecker JP. Functional topography of human auditory cortex. J Neurosci 2016;36(4):1416–1428 45. Van Hoesen GW. Anatomy of the medial temporal lobe. Magn Reson Imaging 1995;13(8):1047–1055 46. Duvernoy HM. The Human Hippocampus. 2nd ed. Berlin, ­Heidelberg: Springer-Verlag; 1998:39–72 47. DeFelipe J, Fernández-Gil MA, Kastanausk, aite A, Bote RP, Presmanes YG, Ruiz MT. Macroanatomy and microanatomy of the temporal lobe. Semin Ultrasound CT MR 2007;28(6):404–415 48. Campero A, Tróccoli G, Martins C, Fernandez-Miranda JC, Yasuda A, Rhoton AL Jr. Microsurgical approaches to the medial temporal region: an anatomical study. Neurosurgery 2006;59(4, Suppl 2):ONS279–ONS307, discussion ONS307 –ONS308 49. Tubbs RS, Loukas M, Barbaro NM, Cohen-Gadol AA. Superficial cortical landmarks for localization of the hippocampus: appli-

23  Cerebral Cortex: Embryological Development and Topographical Anatomy cation for temporal lobectomy and amygdalohippocampectomy. Surg Neurol Int 2015;6:16 50. Wagner M, Jurcoane A, Hattingen E. The U sign: tenth landmark to the central region on brain surface reformatted MR imaging. AJNR Am J Neuroradiol 2013;3;4(2):323–326 51. Naidich TP, Valavanis AG, Kubik S. Anatomic relationships along the low-middle convexity: part I—Normal specimens and magnetic resonance imaging. Neurosurgery 1995;36(3):517–532 52. Yousry TA, Schmid UD, Alkadhi H, et al. Localization of the motor hand area to a knob on the precentral gyrus. A new landmark. Brain 1997;120(Pt 1):141–157 53. Naidich TP, Blum JT, Firestone MI. The parasagittal line: an anatomic landmark for axial imaging. AJNR Am J Neuroradiol 2001;22(5):885–895 54. Lehman VT, Black DF, Bernstein MA, Welker KM. Temporal lobe anatomy: eight imaging signs to facilitate interpretation of MRI. Surg Radiol Anat 2016;38(4):433–443 55. Naidich TP, Brightbill TC. The intraparietal sulcus: a landmark for localization of pathology on axial CT scans. Int J Neuroadiol 1995;1:3–16

56. Naidich TP, Brightbill TC. The pars marginalis, II: a “bracket” sign for the central sulcus in axial plane CT and MRI. Int J Neuroradiol 1996;2:3–19 57. Naidich TP, Brightbi, ll TC. Systems for localizing fronto-­ parietal gyri and sulci on axial CT and MRI. Int J Neuroradiol 1996;2:313–338 58. Naidich TP, Valavanis AG, Kubik S, Taber KH, Yaşargil M. Anatomic relationships along the low-middle convexity. Part II: lesion localization. Int J Neuroradiol 1997;3:393–409 59. Yousry TA, Schmid UD, Jassoy AG, et al. Topography of the cortical motor hand area: prospective study with functional MR imaging and direct motor mapping at surgery. Radiology 1995;195(1):23–29 60. Yousry TA, Fesl G, Buttner A, Noachtar S, Schmid UD. Heschl’s gyrus: anatomic description and methods of identification on magnetic resonance imaging. Int J Neuroradiol 1997;3:2–12 61. Meyer JR, Roychowdhury S, Russell EJ, Callahan C, Gitelman D, Mesulam MM. Location of the central sulcus via cortical thickness of the precentral and postcentral gyri on MR. AJNR Am J Neuroradiol 1996;17(9):1699–1706

227

24

  Tractographic Anatomy of White Matter Sandip S. Panesar, David T. Fernandes-Cabral, Antonio Meola, Fang-Cheng Yeh, Maximiliano Nunez, and Juan C. Fernandez-Miranda

Summary Classical descriptions of white matter neuroanatomy were derived from postmortem dissection. The introduction of diffusion magnetic resonance tractography at the end of the 20th century allowed researchers and clinicians the ability to conduct large-scale studies into white matter connectivity, laterality, and subdivision over large subject sets and to a high level of accuracy previously unavailable by postmortem techniques. In this chapter, we explore the application of tractography to the study of white matter neuroanatomy. We further discuss tractographic contributions to the understanding of the anatomy of major white matter systems (projection, association, limbic, commissural, and cranial nerves). Keywords:  tractography, white matter anatomy, neurosurgical anatomy, white matter pathways

„„ Introduction The use of diffusion-weighted MRI to conduct in vivo tractography of the human central nervous system (CNS) has provided clinicians with novel means to visualize white matter structure and arrangement in both healthy and diseased subjects. When compared to the traditionally used method of post-mortem white matter dissection, tractography confers several notable advantages: firstly, it does not involve the removal of overlying cortex or white matter, thus completely preserving hemispheric architectural integrity. Secondly, it allows for observation of white matter cortical terminations, giving direct insight into connectivity. Thirdly, white matter volumetry may be calculated, allowing determination of the hemispheric lateralization of white matter systems. Tractography involves three stages: the acquisition of raw diffusion-weighted MRI data, reconstruction of the data to calculate axonal orientations, and application of either probabilistic or deterministic fiber-tracking algorithms for tract delineation. Several reconstruction modalities are available to investigators; diffusion tensor imaging (DTI) is perhaps the best known of these. It involves the calculation of a single directional tensor per voxel, derived from the diffusion properties of protons confined to axons. The tensor’s principal direction subsequently corresponds with axonal direction. Tractographic visualization is subsequently achieved by stepwise connection

of similarly oriented directional tensors (or equivalent for nontensor techniques). Shortcomings of tensor-based tractography are its inability to accurately trace crossing fibers and the inability to demonstrate cortical terminations with accuracy.1,​2,​3 These limitations have been partially addressed by introduction of advanced tractographic modalities such as the Q-ball model, diffusion spectrum imaging (DSI), and high angular resolution diffusion imaging.2,​4 CNS white matter arrangement can be subdivided into a series of anatomically and functionally interconnecting systems: the limbic fibers, the projection fibers, commissural fibers, and association fibers.5,​6,​7,​8 Historically, anatomical knowledge of these systems was derived from the pioneering work of neuroanatomists and clinicians who utilized post mortem dissection with clinical correlations to make anatomical inferences. Naturally, controversies arose as research progressed, many of which persisted. Tractography has addressed some of these controversies by allowing in vivo studies in large groups of healthy and diseased subjects. Nevertheless, because of the variance in tractographic modalities and methodologies, new controversies regarding the structure, subdivision, connectivity, and even existence of white matter systems have arisen.2 This chapter aims to provide an overview of the current anatomical knowledge of CNS white matter systems, as derived from both classical descriptions and tractographic insights.

„„ Limbic System The limbic system, primary neural substrate of memory and emotion, consists of gray matter nuclei interconnected by white matter fibers: The eponymously named Papez circuit consists of a loop beginning at the hippocampus and connecting to the mammillary body via the fornix (Fig. 24.1). It continues to the anterior thalamic nucleus as the mammillothalamic tract. Short projections, known as the anterior thalamic radiations, then project from the thalamus to the cingulum. The cingulum travels around and dorsal-to the parasagittal corpus callosum, before terminating within the entorhinal cortex and completing the circuit.9 Other gray matter structures including the ­amygdala, olfactory bulb, septal area, and hypothalamus are included in contemporary descriptions of the limbic system.10,​11

24  Tractographic Anatomy of White Matter

Fig. 24.1  All tractography images in this chapter were created using DSI Studio software (http://dsi-studio.labsolver.org). Diffusion spectrum imaging tractography was conducted using a template consisting of averaged diffusion data from 842 healthy subjects. (a) Tractographic rendering of the fornix, viewed from a lateral sagittal perspective. Cerebral isosurface has been overlaid for orientation. (b) Tractographic rendering of the fornix, viewed from a superior axial perspective. Cerebral isosurface has been overlaid for orientation.

Fornix

Cingulum

The fornix is the primary efferent hippocampal pathway and is readily visualized using tractographic algorithms. It originates from the subiculum and entorhinal cortex of the hippocampus as the fimbria. Each hemispheric fimbria passes superomedially along the medial border of the posterior pulvinar in a forward-looping course. It is joined by its antihemispheric counterpart at the junction of the atrium and body of the lateral ventricle, to form the forniceal body, which lies beneath the splenium of the corpus callosum. The two hemispheric components of the body communicate via the hippocampal commissure, or psalterium. The body continues anteriorly, along the superomedial surface of the posterior aspect of the pulvinar before separating again into antihemispheric continuations above the anterior surface of the foramen of Monro. These anterior divisions each further subdivide into anterior and posterior columns. The anterior column terminates within the hypothalamus, accumbens nucleus, and septal region. The posterior column travels ventrally, posterior to the anterior commissure to terminate in the medial mammillary nucleus, adjacent hypothalamus, and anterior thalamic nucleus (Fig. 24.1).5,​9,​11

The cingulum is a two-part, long white matter system, comprised of several shorter white matter populations. Anteriorly, it originates below the genu of the corpus callosum from the subcallosal gyrus and paraolfactory area. It courses around the genu and splenium of the corpus callosum, traveling posteriorly within the white matter of the cingulate gyrus, assuming its sagittal contour. This portion is known as the anterior (dorsal) component. The posterior (ventral) component continues as cingulate fibers travel via the medial temporal lobe to terminate within the uncus, parahippocampal gyrus, and amygdala. Along its course, short fibers interconnect cortical areas over both long and short distances.5,​9,​11 As the cingulum consists of smaller populations of short fibers of varying length and trajectory, accurate tractographic depiction may prove difficult, a problem further compounded by the cingulum’s proximity to the fluid-filled ventricles. Nevertheless, tensor-based studies have tracked and subsequently proposed either a bipartite12 or tripartite13 segmentation of the cingulum based on its cortical terminations and anatomical course (Fig. 24.2).

Mammillothalamic Tract

The thalamic radiations are short fibers that project radially from the gray matter of the thalamus to end points within the frontal, parietal, temporal, and occipital lobes. As such, thalamic radiations lie within the white matter of the internal capsule’s anterior limb, genu, posterior limb, retrolenticular, and sublenticular portions, respectively.14 Specific to the limbic system, anterior thalamic nuclei, receiving input from the fornix and mammillothalamic tract, convey information to the anterior

From the medial mammillary nucleus, very short fibers ascend dorsally to the anterior nucleus of the thalamus, a connection known as the mammillothalamic tract of Vicq d’Azyr. Postthalamic fibers travel further dorsally, becoming interspersed with fibers from the anterior limb of the internal capsule, to terminate within the cingulum and cingulate gyrus.9

Thalamic Radiations

229

230

III  Surgical Anatomy and Mapping Techniques

Fig. 24.2  Tractographic rendering of the right cingulum, viewed from a lateral sagittal perspective. Cerebral isosurface has been overlaid for orientation.

Fig. 24.3  Left-sided, oblique posterior cutaway view demonstrating left precentral gyrus (PrCG), postcentral gyrus (PoCG), thalamus (L-Thal), descending motor fibers (DMF), thalamocortical sensory projections (TCSP), and superior cerebellar peduncle projections (SCP). (Reproduced with permission from Meola et al 2016.21)

cingulate and orbitofrontal cortex via the thalamic radiations.11 In keeping with the thalamic role as a sensory conduit, ascending sensory projection fibers are also included in descriptions of the internal capsular thalamocortical radiations (see below).

dissection, with immunohistochemical methods required for delineation.17 Tractography cannot differentiate between afferent and efferent projection fibers; however, these smaller projections can still be visualized as part of the larger fiber bundles they traverse within.

„„ Projection System

Extrathalamic Serotonergic Projections

The projection system subserves both cortical sensory input and motor output. It consists of both discrete and somatotopically correlated white matter bundles. The optic radiations, acoustic radiations, corticothalamic and extrathalamic (i.e., lemniscal), and superior cerebellar tracts constitute the sensory projection pathways. The motor projection pathways include geniculate and cerebrospinal fibers traversing within the genu and posterior aspect of the internal capsule and the corticopontine fibers.

Thalamocortical Sensory Projections The thalamocortical sensory fibers ascend in a somatotopic arrangement from the thalamus to the postcentral gyrus, medial to the corticospinal and corticopontine fibers within the internal capsule. They originate at the ventroposteromedial and ventroposterolateral nuclei of the thalamus together with components of the corona radiata. Toward their terminations within the postcentral cortex the thalamocortical sensory projections are said to “rotate” posteriorly (i.e., clockwise when viewed in an axial plane) to mirror the classically described somatotopy of the sensory cortex, an arrangement which has been replicated tractographically (Fig. 24.3).15,​16

Extrathalamic Projections There are four types of extrathalamic afferent projections traversing between subthalamic nuclei and cortical areas. They convey serotonergic, noradrenergic, dopaminergic, and cholinergic afferents from the mesencephalic raphe nuclei, pontine nucleus locus coeruleus, substantia nigra/ventral tegmental area complex, and nucleus basalis of the substantia innominata, respectively. These pathways remain obscure to cadaveric

The dorsal longitudinal fasciculus is the primary extrathalamic serotonergic afferent. It originates from the dorsal raphe nuclei as two smaller bundles (dorsal and ventral), which ascend in the paramedian brain stem and merge at the caudal diencephalon before separating again. One group of fibers ascends within the internal capsule and terminates within lateral cerebral cortex. The other group curves ventrally, traveling to the hypothalamus via the medial forebrain bundle.18 The medial forebrain bundle has been tractographically demonstrated using a tensor-based approach,19,​20 and brainstem portions of the dorsal longitudinal fasciculus can be visualized using nontensor methods (Fig. 24.4).21

Extrathalamic Noradrenergic Projections Extrathalamic neocortical projections from the locus c­oeruleus project diffusely throughout the white matter to reach ­geographically separated cortical areas. Research concerning their anatomy has been limited largely to nonhuman immunohistochemical tracer studies; however their pathway from the locus coeruleus nuclei is suggested to travel to frontal cortical areas via the brainstem, before sweeping rostrocaudally through the hemisphere via the cingulum and giving off widespread c­ onnections throughout its course.17,​22 The cingulum has been the subject of considerable tractographic study (see section on Cingulum).

Extrathalamic Dopaminergic Projections Extrathalamic dopaminergic afferents arise from the substantia nigra, within the ventral tegmental area of the midbrain. They travel via the medial forebrain bundle (see Chapter 20, Section “Studies on Adults”) in a course passing inferior to the thalamus, before joining the ventral anterior limb of the internal capsule and continuing to terminations within the ­prefrontal

24  Tractographic Anatomy of White Matter

Fig. 24.4  (a) Sagittal view of medial hemispheric surface of a cadaveric specimen. Corpus callosum has been transected across the midline. Labelled are the rostrum, genu, body, and splenium. Anterior commissure (AC) has been transected. Other visible structures are the thalamus (Thal), mammillary body (MB), red nucleus (RN), cerebral aqueduct (CA), dorsal longitudinal fasciculus (DLF), medial longitudinal fasciculus (MLF), fourth ventricle (4V), and pons. (b) Left-sided, oblique posterior cutaway view demonstrating left-sided red nucleus (L-RN), dorsal longitudinal fasciculus (DLF) terminating in hypothalamus (Hypothal.), medial longitudinal fasciculus (MLF), cerebral aqueduct of Sylvius (CA), and fourth ventricle (4V). (c) Right-sided, oblique anterior cutaway view demonstrating right-sided thalamus (R-Thal), frontothalamic fibers (R-FT), and medial forebrain bundle (R-MFB). (Part b of the image is reproduced with permission from Meola et al 2016.21)

cortex. There is otherwise limited information regarding the trajectory of these projections within humans.20

Extrathalamic Cholinergic Projections Extrathalamic cholinergic projections arise from the nucleus basalis of Meynert (Ch4) in the basal forebrain. They project via two pathways: lateral and medial. The medial pathway originates from Ch4 nuclei and travels via the gyrus rectus and medial orbital gyrus to the callosal rostrum, before entering the cingulum. As part of the cingulum course, it gives off radiations to orbitofrontal, subcallosal, cingulate, pericingulate, and retrosplenial cortex. The lateral projection pathway is bipartite. A capsular division ascends from Ch4 via the medial external capsule adjacent to the putamen, terminating within the dorsal frontoparietal cortex, middle and inferior temporal gyri, inferotemporal cortex, and the parahippocampal gyri. Individual capsular fibers travel via the uncinate fasciculus to the amygdala (this is potentially disputed, see below). The second lateral component is a perisylvian division also traveling within the claustrum, before turning laterally to terminate within the inferior frontal and superior temporal gyri. Individual perisylvian fibers radiate to the frontoparietal operculum,

superior temporal gyrus, and the insula.23 Perisylvian fibers can be tractographically visualized among others in their anatomical plane, that is, the claustrocortical fiber system (Fig. 24.5).24

Optic Radiations The optic radiations arise from the lateral geniculate nucleus of the ventral thalamus. Fibers first travel anteriorly within the temporal lobe, medial to the medial wall of the anterior horn of the lateral ventricle, before turning acutely and continuously fanning out around the supero-anterior aspect of the anterior horn. The radiations, now lateral to the lateral ventricle, continue posteriorly to the occipital lobe. They course within the ­sagittal stratum, lying superficial to the tapetum, which forms the lateral wall of the lateral ventricle. The optic radiations terminate at the superior and inferior cortical margins of the ­ calcarine fissure.25 Based on the gradated arrangement at which optic radiation fibers course around the anterior horn of the l­ateral ventricle, the variable trajectories of which certain groups of fibers exit the lateral geniculate nucleus, and the ­ different components of visual field subserved by each group, ­ discrete anatomo-functional subdivisions of the optic radiation into “anterior-superior” (passing around the tip of the anterior horn,

231

232

III  Surgical Anatomy and Mapping Techniques

Fig. 24.5  (a) Lateral hemispheric surface of a post mortem specimen. Arcuate fasciculus (AF) and superior longitudinal fasciculus (SLF) have been removed to reveal the inferior longitudinal (ILF) and middle longitudinal (MdLF) fasciculi lying superficially to the claustrum (Claust.) and uncinate (UF) fasciculi. (b) Corresponding tractographic picture to 5a. The AF and SLF have been “virtually dissected” to reveal the inferior (ILF) and middle (MdLF) longitudinal fasciculi. (c) Lateral hemispheric surface of a post mortem specimen. Dissection continued from 5a. The ILF and MdLF have been removed to reveal structures of the ventral external capsule (uncinate [UF] and inferior fronto-occipital fasciculi [IFOF]) and the claustrum (Claust.). (d) Corresponding tractographic picture to 5c. The ILF and MdLF have been removed to reveal the uncinate (UF) and inferior longitudinal fasciculi (ILF), and the claustrum (Claust.).

i.e., Meyer’s loop), “central-macular” (passing over the roof, between anterior and posterior bundles), and “posterior-­inferior” (passing over the roof and anterior atrium) ­components have been suggested.26,​27 This anatomo-functional subdivision may potentially be useful during temporal lobe surgery, where preoperative tractographic visualization of the optic radiations can be used to determine resection margins and/or predict postoperative deficits. Despite potential clinical benefit, accurate tractographic representation of the optic radiations proves difficult for most tractographic modalities. Algorithms may be unable to trace the gradated and acute trajectory of the optic radiations as they turn around the slanted superior aspect of the anterior horn. The close proximity of the optic radiations to the fluid-filled ventricles may also affect signal intensity, and close proximity to other sagittal stratum systems (i.e., inferior fronto-­occipital fasciculus [IFOF], inferior longitudinal fasciculus [ILF]) may result in false continuation artifacts.

Auditory Projections The auditory projections subserve transmission of auditory information from the medial geniculate nucleus to the primary auditory cortex. There is a relative paucity of published

i­nformation regarding their white matter course, which is described as a “tortuous-S.”28 They emanate from the medial geniculate nucleus at the posterior aspect of the thalamus, initially passing underneath the optic radiations and within the posterior aspect of the internal capsule, before curving around the inferior sulcus of the insula and terminating within primary auditory cortex at Heschl’s gyrus.26,​29 Their anatomical course is further convoluted by passage through several perpendicularly oriented fiber tracts, particularly white matter of the temporal stem.30 As previously noted, a limitation of DTI is its inability to trace crossing pathways. Tractography of the auditory radiations is preferably suited to, and has been successfully implemented with nontensor techniques.30,​31

Corticocerebellar Projections There are two major descending corticocerebellar projection pathways, both subserving motor coordination.32 They utilize the middle cerebellar peduncle as a conduit between the frontal cortical areas and white matter of the cerebellar hemispheres. Within the middle cerebellar peduncle, the corticopontine projections are arranged superiorly, with respect to the extrapontine projections, and project to all areas of the c­ erebellar cortex,

24  Tractographic Anatomy of White Matter

Fig. 24.6  (a) Right-sided, oblique posterior cutaway view demonstrating left thalamus (L-Thal.), right and left red nuclei (R-RN; L-RN), left dentate nucleus (L-Dent.), frontopontine (FPP), temporoparietopontine (TPP) internal capsular fibers, right and left superior cerebellar peduncles (R-SCP; L-SCP), and dorsal (Dors. TF), intermediate (Int. TF), and ventral (Vent. TF) transverse peduncular fibers. (b) Infero-oblique view of dissected brainstem structures. Visible are the ventral transverse peduncular fibers (Vent. TF), intermediate transverse peduncular fibers (Int. TF), dorsal transverse peduncular fibers (Dors. TF) of the middle cerebellar peduncle. Other tracts visible are frontopontine fibers (FPP) and descending motor fibers (DMF). (c) Right-sided, oblique anterior view demonstrating the middle cerebellar peduncular fibers (MCP) crossing transversely at the pons region. Passing deep and perpendicular to these fibers are the frontopontine fibers (FPP; anteromedial), descending motor fibers (DMF; anterolateral), temporopontine fibers (TPP; posterolateral), and thalamocortical sensory projections (TCSP; deep to aforementioned). (d) Right-sided, oblique anterior view of a dissected cadaveric specimen. Middle cerebellar peduncular fibers (MCP) have been removed to reveal the underlying perpendicular tracts. Visible are the frontopontine fibers (FPP; anteromedial), descending motor fibers (DMF; anterolateral), temporopontine fibers (TPP; posterolateral). Other structures visible are the superior cerebellar peduncle (SCP), the medullary pyramids (MP), and olivary bodies (OB). (Part a is reproduced with permission from Meola et al 2016.21)

passing around the dentate nucleus. The inferiorly arranged extrapontine fibers project to the posterior cerebellum.33 The “corticopontocerebellar” projections include the pontine nuclei as a waypoint on their course between cortex and the cerebellum. At this level the pathway decussates and continues to the contralateral cerebellum.34 DTI studies have suggested that corticopontine projections descend in a topographically organized arrangement, from all major cortical areas with strongest contribution from the prefrontal, frontal, and parietal areas.35 The second group of corticocerebellar pathways are postulated to pass from cortex to cerebellum via extrapontine routes. Though they have been visualized using tensor-based methods,33 these pathways have received negligible attention in the literature and their hemispheric white matter trajectory, including decussation, remains obscure.32 The descending components can be tractographically visualized in a compartmentalized manner; as the pathways involve multiple synaptic stages, region-of-interest placement at various known legs of the corticopontocerebellar

projections can enable their v ­ isualization (including their decussation) using both tensor33,​34,​36 and nontensor-based methods.21 Ascending pathways, also subserving motor coordination, project from the dentate nucleus to the thalamus via the superior cerebellar peduncle. They ascend through the ipsilateral pons, adjacent to the dorsolateral wall of the fourth ventricle and decussate to pass through the contralateral red nucleus within the midbrain. From here, they terminate within the ventrolateral nucleus of the contralateral thalamus.21,​33,​37 Because of the red nucleus waypoint, this pathway is sometimes referred to as the dentatorubrothalamic tract. The thalamocortical projections are to all cortical areas via the internal capsule, however, specifically concerned with movement are those to the precentral cortical areas, which are postulated to be somatotopically arranged, mirroring the organization of the cortex,15 and which can be tractographically visualized.38 The subthalamic portion of this system has been visualized (including the decussation) using both tensor-37 and nontensor-based approaches (Fig. 24.6).

233

234

III  Surgical Anatomy and Mapping Techniques

Descending Motor Projections The corticospinal fibers descend from the precentral gyrus and travel via the posterior limb of the internal capsule and the cerebral peduncle. They consist of two functional groups: corticospinal and corticobulbar. Corticospinal fibers subserve somatic motor function of the contralateral body half, following their decussation in the upper one-third of the medullary pyramids. A somatotopic arrangement of the corticospinal projections, mirroring the functional topography of the precentral gyrus was historically conjectured. Tractography has confirmed this organization down to the midbrain level.37,​39,​40 For the corticobulbar projections, the upper motor neuron components of the contralateral somatic cranial nerve pathways (with exceptions) travel via the corona radiata-genu-cerebral peduncular route, anteriorly to their corticospinal counterparts before decussating and synapsing upon the cranial nerve motor nuclei. These pathways have also been visualized and differentiated using tractographic means.8 Most tractographic modalities are able to demonstrate vertical portions of the corticospinal projections; however, a common shortcoming in many studies has been the incomplete representation of portions arising from the ventrolateral aspects of the precentral gyrus. The more radially oriented precentral fibers must traverse the centrum semiovale, a white matter area superolateral to the lateral ventricle, consisting of crossing projection, association, and commissural fibers. The high density of variably oriented crossing fibers within this area may prove difficult for some tractographic algorithms to resolve adequately (Fig. 24.6c, d).38

„„ Commissural Fibers Commissural fibers are telencephalic white matter structures permitting interhemispheric communication. There are three generally accepted commissural systems: the anterior commissure, corpus callosum, and psalterium (see 24.2.1). The posterior commissure has not been extensively studied in humans and is therefore not classically included in descriptions.

Anterior and Posterior Commissures The anterior and posterior commissures are compact telencephalic fiber pathways, crossing the midline, connecting gray- and white-matter structures within the hemispheres. They are readily visualized by most tractographic algorithms. The multifunctional anterior commissure (Fig. 24.4a) is a dense collection of white matter, which assumes a “handlebar” arrangement as it passes between hemispheres.6 The body is a dense collection of transversely oriented white matter, passing interhemispherically within the anterior wall of the third ventricle. Within each hemisphere, the lateral body bifurcates into anterior and posterior crura. The anterior crus passes within the medial orbitofrontal white matter to terminate within the olfactory bulb, anterior perforated substance, and anterior olfactory nucleus. The posterior crus continues superolaterally toward the temporal and occipital lobes. The temporal extension continues toward the temporal pole and terminates within the hippocampus and amygdala. The occipital extension merges with the sagittal stratum to terminate within the occipital cortex.6,​14 The posterior

commissure is located within the posterior aspect of the third ventricle. It subserves the consensual light reflex, connecting nuclei within the bilateral pretectal areas. Despite recent animal and human DTI studies regarding this fascicle, its anatomy and function are still not fully defined.41

Corpus Callosum The corpus callosum is a histologically heterogeneous interhemispheric conduit connecting homologous and h ­ eterotopic cortical areas across the midline. Its multifunctional roles include cognition, perception, decision making, and learning. It can be grossly divided into three regions: the genu, body, and splenium; however, contemporary tractographic research has suggested alternative subdivisional arrangements based on its connectivity.42 As the cadaveric corpus callosum appears to be of uniform morphological consistency upon gross visual inspection, classification schemes were derived in an attempt to segment it anatomically. Witelson’s43 prototypical scheme utilized fractional subdivision of its maximal sagittal length; within each callosal segment were anatomical areas with assumed bihemispheric cortical homologues.43,​44 This scheme did not incorporate any anatomically derived parameters as criteria for callosal subdivision, such as cortical connectivity patterns or histological differentiation. The tensor-based connectivity study by Hofer and Frahm44 elaborated upon, and modified Witelson’s description, demonstrating that white matter projections from the prefrontal, premotor, supplementary motor, primary motor, primary sensory, parietal, temporal, and occipital cortical homologues were parcellated within five discrete transhemispheric callosal segments. As the bulk of callosal white matter is arranged perpendicularly to the majority of projection (superoinferior) and association (anteroposterior) bundles, tractography of the corpus callosum may be preferentially suited to tractographic methods unconstrained by crossing fibers; the study of Jarbo et al42 used diffusion-spectrum imaging to define callosal connectivity, finding predominant homotopic connection within the frontal, parietal, and occipital association areas, with minimal temporal connectivity. The authors also found corticosubcortical callosal connectivity to the basal ganglia and thalamus (Fig. 24.4a and Fig. 24.7).

„„ Association System Association fibers subserve higher-order cognitive functions including language and visual integration. In concordance with evolutionary theory, the human association systems demonstrate a greater degree of differentiation relative to nonhuman species. Studies into the cortical connectivity of association tracts are therefore paramount to anatomo-functional understanding. The “gold standard” technique of immunohistochemical tracing to define connectivity necessitates sacrifice of the study subject and is therefore limited to nonhuman species. Likewise, the limitations of human post mortem white matter dissection have been discussed above. Tractography potentially reconciles these differences, allowing human connectivity to be studied in vivo. Nevertheless, differences in data acquisition parameters, tractographic modality, and tracking methods have introduced significant variability between studies. Because of these issues, controversies persist regarding their anatomy and function.

24  Tractographic Anatomy of White Matter

Fig. 24.7  (a) Tractographic rendering of the left corpus callosum connections, viewed from a lateral sagittal perspective. Cerebral isosurface has been overlaid for orientation. (b) Tractographic rendering of the corpus callosum connections, viewed from a superior axial perspective. Cerebral isosurface has been overlaid for orientation.

Fig. 24.8  (a) Lateral hemispheric surface of a cadaveric specimen. Overlying gray matter and U-fiber system has been removed to reveal the arcuate fasciculus (AF) and superior longitudinal fasciculus (SLF). (b) Corresponding tractographic picture to 8a. The most superficial large white matter systems are the arcuate (AF) and superior longitudinal (SLF) fasciculi.

Arcuate Fasciculus

Superior Longitudinal Fasciculus

The arcuate fasciculus (AF) is a “C-shaped” tract connecting the temporal lobe with the frontal lobe via the Sylvian fissure. It lies deep to the cortico-cortical U-fibers and is the most superficial large white matter system. Introduction of tractography has provided insight into morphology and connectivity of the arcuate previously obscured from post mortem observations: Both tensor- and nontensor-based methods have demonstrated the left-sided volumetric and connective lateralization of the AF in humans. This lateralization is thought to underpin the evolution of complex language functionality unique to humans, ­particularly in the domains of phonological and semantic processing.45,​46 The left arcuate consists of two components, most recently described by Fernandez-Miranda et al:46 a larger dorsal segment and a smaller ventral segment. The dorsal segment interconnects ventral precentral and caudal middle frontal gyri with caudal middle and inferior temporal gyri. The ventral segment connects the superior and rostral middle temporal gyri with the pars opercularis. The right-sided arcuate does not demonstrate the same extent of volumetric or connective differentiation; the ventral segment connects superior temporal gyrus with the pars triangularis. The dorsal segment connects rostral middle and inferior temporal gyri with the pars opercularis and caudal middle frontal gyrus (Fig. 24.8 and Fig. 24.9).

The superior longitudinal fasciculus (SLF) is a collection of dorsally running association fiber tracts. It lies deep to the corticocortical U-fibers and within close proximity to the dorsal leg of the AF. Historically, the AF has been considered as part of the larger SLF system;6 however, more recent evidence regarding evolutionary anatomo-functional differentiation serves to refute this concept. The SLF consists of two distinct components: SLF-II (larger, lying dorsally) and the SLF-III (lying ventrally). In general, the SLF-II connects the supramarginal and angular gyri with the caudal middle-frontal gyrus, and the SLF-III connects the supramarginal gyrus with the pars opercularis. The connectivity pattern of each SLF component varies hemispherically, with additional connectivity of the SLF-II to the caudal dorsal precentral gyrus on the left, and SLF-III showing exclusive connectivity to pars triangularis on the right. As the SLF is volumetrically lateralized to the right hemisphere, this lends further credence to the notion of the AF and SLF as separate association systems. (Fig. 24.8 and Fig. 24.10).47

Inferior Fronto-Occipital Fasciculus The inferior fronto-occipital fasciculus is a large, ventrally lying fiber tract connecting the ventral frontal cortices with the

235

236

III  Surgical Anatomy and Mapping Techniques occipital cortex. It assumes a “bow-tie” arrangement at its anterior and posterior extremities, interconnected by a compact stem. The inferior fronto-occipital fasciculus lies deep to the arcuate fasciculus within the white matter of the external and extreme capsules, with the external capsular portion traversing within the white matter of the temporal stem. At its frontal origins, the inferior fronto-occipital fasciculus lies dorsal to the uncinate fasciculus, with both systems traveling into the temporal stem together.6 At its posterior extent, it blends with the white matter of the sagittal stratum. Relevant to human evolution, this fascicle is not found within the subhuman primates, and along with other ventral fiber tracts is thought to subserve lexicosemantic processing among other complex functions. Its

exact subdivision, connectivity, and lateralization in the literature have been subject to considerable debate; however, both tensor-48 and nontensor-based49 studies generally concur that it consists of superficial and deep layers, which can be further subdivided. Consensus regarding ­subdivisions within the layers has not been reached, however. Volumetrically, it is not significantly leftward-lateralized, like the arcuate. Finally, though most groups agree upon its terminations within the frontal pole and occipital cortex, separate tractographic studies have demonstrated varying connectivity to dorsomedial and dorsolateral (i.e., Broca’s area) frontal areas, parietal cortex, and temporal lobe (i.e., the fusiform gyrus).48,​49,​ 50 Regarding its posterior terminations, due to blending with the ­sagittal stratum, tractographic algorithms may not be able to differentiate the IFOF from other tracts within this layer, ­potentially resulting in spurious connective observations (Fig. 24.5c, d and Fig. 24.11).

Uncinate Fasciculus

Fig. 24.9  Tractographic rendering of the left arcuate fasciculus, viewed from a lateral sagittal perspective. Cerebral isosurface has been overlaid for orientation.

Fig. 24.10  Tractographic rendering of the right superior longitudinal fasciculus, viewed from a lateral sagittal perspective. Cerebral isosurface has been overlaid for orientation.

The uncinate fasciculus (UF) is a fiber tract running between the ventral prefrontal areas and the temporal pole. It originates ventral to the inferior fronto-occipital fasciculus within the orbitofrontal cortex and frontal pole (see below) and terminates within the temporal pole.6 Along with the other ventral association fasciculi, it is thought to subserve lexicosemantic processing. There is general consensus ­ regarding its connectivity, with some exception; the majority of tractographic research has demonstrated bifurcating frontal connectivity with the medial and lateral orbitofrontal, rectus, and ventrolateral frontal gyri. Temporal connectivity is with the temporal pole, uncus, and parahippocampal gyrus.50,​51 ­Controversy exists regarding its connectivity with the amygdala, a limbic gray matter nucleus. Furthermore, there is inconsistency across tractographic studies ­regarding volumetric ­lateralization of the UF (Fig. 24.5c, d and Fig. 24.12).

Fig. 24.11  Tractographic rendering of the left inferior fronto-occipital fasciculus, viewed from a lateral sagittal perspective. Cerebral isosurface has been overlaid for orientation.

24  Tractographic Anatomy of White Matter

Inferior Longitudinal Fasciculus The inferior longitudinal fasciculus is the most ventrally lying association tract. It connects the superior, middle, inferior, and fusiform temporal gyri to the cuneus, lingula, and occipital poles. It is thought to lie between the caudal extension of the arcuate fasciculus and the postcapsular portion of the inferior fronto-occipital fasciculus, approximating the lateral contour of the lateral ventricle during its posterior course. The existence of the inferior longitudinal fasciculus has been called into ­question, with suggestions that ventral temporo-occipital ­connectivity is in fact subserved by corticocortical U-fibers.52 Nevertheless, certain tensor-based studies suggest a bicomponent structure, consisting of indirect and direct segments. The indirect component is proposed to con-

sist of serial corticocortical U-fibers and lies laterally to the direct component, connecting anterior portions of the superior, middle, and inferior temporal gyri, the medial hippocampal structures, and the amygdala to the occipital lobe (Fig. 24.5a, b and Fig. 24.13).7,​53

Middle Longitudinal Fasciculus The middle longitudinal fasciculus is a recently described white matter fasciculus. It has been studied using both tensor and nontensor tractography, which have demonstrated it as lying within the white matter between the arcuate fasciculus and inferior fronto-occipital fasciculus, running obliquely, in a posterosuperior course to the latter. It is generally agreed that the middle longitudinal fasciculus connects superior temporal gyrus to the precuneus and cuneus.3,​ 54 Original descriptions included the angular gyrus as a dorsal termination; however, the study of Wang et al3 demonstrated that these were likely aberrant findings secondary to tractographic variability. ­ Furthermore, the volumetric lateralization of the middle longitudinal fasciculus cannot be reliably concluded upon, with conflicting findings across tractographic studies (Fig. 24.5a, b and Fig. 24.14).

„„ Cranial Nerves

Fig. 24.12  Tractographic rendering of the right uncinate fasciculus, viewed from a lateral sagittal perspective. Cerebral isosurface has been overlaid for orientation.

Fig. 24.13  Tractographic rendering of the right inferior longitudinal fasciculus, viewed from a lateral sagittal perspective. Cerebral isosurface has been overlaid for orientation.

Several studies attempting to visualize the intracranial portions of the cranial nerves have been conducted with varying results.55 Of particular interest to the neurosurgeon are the pregeniculate optic tract, the sensorimotor cranial nerves, and the vestibulocochlear nerve, all of which are susceptible to displacement by space-occupying lesions, and which produce readily assessable clinical signs. As such, in vivo tractography has demonstrated its clinical usefulness in preoperative planning. Tensor- and ­nontensor-based studies have partially visualized intracranial portions of both healthy and pathologically ­displaced cranial nerves.

Fig. 24.14  Tractographic rendering of the right middle longitudinal fasciculus, viewed from a lateral sagittal perspective. Cerebral isosurface has been overlaid for orientation.

237

238

III  Surgical Anatomy and Mapping Techniques

„„ Conclusion Tractography has provided insight into the structure, function, and connectivity of almost all components of intracranial white matter. Though cadaveric dissection and immunohistochemical methods provided the basis from which the ­anatomical

­knowledge of intracranial white matter was founded, the potential insights of in vivo tractography in both research and clinical settings offer unparalleled opportunity to gather and integrate anatomical and functional insights. Despite these benefits, a consensus on optimal tractographic modality and method has yet to be reached, and it is still very much a work in progress.

References 1. Farquharson S, Tournier JD, Calamante F, et al. White matter fiber tractography: why we need to move beyond DTI. J Neurosurg 2013;118(6):1367–1377

superolateral and inferomedial medial forebrain bundles, and newly identified septum pellucidum tract. World Neurosurg 2015;83(1):54–61.e32

2. Abhinav K, Yeh FC, Pathak S, et al. Advanced diffusion MRI fiber tracking in neurosurgical and neurodegenerative disorders and neuroanatomical studies: a review. Biochim Biophys Acta 2014;1842(11):2286–2297

20. Coenen VA, Panksepp J, Hurwitz TA, Urbach H, Mädler B. Human medial forebrain bundle (MFB) and anterior thalamic radiation (ATR): imaging of two major subcortical pathways and the dynamic balance of opposite affects in understanding depression. J Neuropsychiatry Clin Neurosci 2012;24(2):223–236

3. Wang Y, Fernández-Miranda JC, Verstynen T, Pathak S, Schneider W, Yeh FC. Rethinking the role of the middle longitudinal fascicle in language and auditory pathways. Cereb Cortex 2013;23(10):2347–2356 4. Yeh FC, Wedeen VJ, Tseng WY. Generalized q-sampling imaging. IEEE Trans Med Imaging 2010;29(9):1626–1635 5. Fernández-Miranda JC, Rhoton AL Jr, Alvarez-Linera J, Kakizawa Y, Choi C, de Oliveira EP. Three-dimensional microsurgical and tractographic anatomy of the white matter of the human brain. Neurosurgery 2008;62(6, Suppl 3):989–1026, discussion 1026–1028 6. Catani M, Howard RJ, Pajevic S, Jones DK. Virtual in vivo interactive dissection of white matter fasciculi in the human brain. Neuroimage 2002;17(1):77–94 7. Catani M, Thiebaut de Schotten M. A diffusion tensor imaging tractography atlas for virtual in vivo dissections. Cortex 2008;44(8):1105–1132 8. Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S. Fiber tract-based atlas of human white matter anatomy. Radiology 2004;230(1):77–87 9. Shah A, Jhawar SS, Goel A. Analysis of the anatomy of the Papez circuit and adjoining limbic system by fiber dissection techniques. J Clin Neurosci 2012;19(2):289–298 10. Concha L, Gross DW, Beaulieu C. Diffusion tensor tractography of the limbic system. AJNR Am J Neuroradiol 2005;26(9):2267–2274 11. Catani M, Dell’acqua F, Thiebaut de Schotten M. A revised limbic system model for memory, emotion and behaviour. Neurosci Biobehav Rev 2013;37(8):1724–1737 12. Thiebaut de Schotten M, Dell’Acqua F, Valabregue R, Catani M. Monkey to human comparative anatomy of the frontal lobe association tracts. Cortex 2012;48(1):82–96 13. Jones DK, Christiansen KF, Chapman RJ, Aggleton JP. Distinct subdivisions of the cingulum bundle revealed by diffusion MRI fibre tracking: implications for neuropsychological investigations. Neuropsychologia 2013;51(1):67–78 14. Gungor A, Baydin S, Middlebrooks EH, Tanriover N, Isler C, Rhoton AL Jr. The white matter tracts of the cerebrum in ventricular surgery and hydrocephalus. J Neurosurg 2017;126(3):945–971 15. Yamada K, Nagakane Y, Yoshikawa K, et al. Somatotopic organization of thalamocortical projection fibers as assessed with MR tractography. Radiology 2007;242(3):840–845 16. Zhang D, Snyder AZ, Shimony JS, Fox MD, Raichle ME. Noninvasive functional and structural connectivity mapping of the human thalamocortical system. Cereb Cortex 2010;20(5):1187–1194 17. Foote SL, Morrison JH. Extrathalamic modulation of cortical function. Annu Rev Neurosci 1987;10:67–95 18. Hornung JP. The human raphe nuclei and the serotonergic system. J Chem Neuroanat 2003;26(4):331–343 19. Cho ZH, Law M, Chi JG, et al. An anatomic review of thalamolimbic fiber tractography: ultra-high resolution direct visualization of thalamolimbic fibers anterior thalamic radiation,

21. Meola A, Yeh FC, Fellows-Mayle W, Weed J, FernandezMiranda JC. Human connectome-based tractographic atlas of the brainstem connections and surgical approaches. Neurosurgery 2016;79(3):437–455 22. Jones BE, Moore RY. Ascending projections of the locus coeruleus in the rat. II. Autoradiographic study. Brain Res 1977;127(1):25–53 23. Selden NR, Gitelman DR, Salamon-Murayama N, Parrish TB, Mesulam MM. Trajectories of cholinergic pathways within the cerebral hemispheres of the human brain. Brain 1998;121 (Pt 12):2249–2257 24. Fernández-Miranda JC, Rhoton AL Jr, Kakizawa Y, Choi C, Alvarez-Linera J. The claustrum and its projection system in the human brain: a microsurgical and tractographic anatomical study. J Neurosurg 2008;108(4):764–774 25. Sincoff EH, Tan Y, Abdulrauf SI. White matter fiber dissection of the optic radiations of the temporal lobe and implications for surgical approaches to the temporal horn. J Neurosurg 2004;101(5):739–746 26. Párraga RG, Ribas GC, Welling LC, Alves RV, de Oliveira E. Microsurgical anatomy of the optic radiation and related fibers in 3-dimensional images. Neurosurgery 2012;71(1, Suppl Operative):160–171, discussion 171–172 27. Ebeling U, Reulen HJ. Neurosurgical topography of the optic radiation in the temporal lobe. Acta Neurochir (Wien) 1988;92(1–4):29–36 28. Maffei C, Soria G, Prats-Galino A, Catani M. Imaging white-matter pathways of the auditory system with diffusion imaging tractography. Handb Clin Neurol 2015;129:277–288 29. Rademacher J, Bürgel U, Zilles K. Stereotaxic localization, intersubject variability, and interhemispheric differences of the human auditory thalamocortical system. Neuroimage 2002;17(1):142–160 30. Berman JI, Lanza MR, Blaskey L, Edgar JC, Roberts TP. High angular resolution diffusion imaging probabilistic tractography of the auditory radiation. AJNR Am J Neuroradiol 2013;34(8):1573–1578 31. Behrens TE, Berg HJ, Jbabdi S, Rushworth MF, Woolrich MW. Probabilistic diffusion tractography with multiple fibre orientations: what can we gain? Neuroimage 2007;34(1):144–155 32. Ramnani N. The primate cortico-cerebellar system: anatomy and function. Nat Rev Neurosci 2006;7(7):511–522 33. Akakin A, Peris-Celda M, Kilic T, Seker A, Gutierrez-Martin A, Rhoton A Jr. The dentate nucleus and its projection system in the human cerebellum: the dentate nucleus microsurgical anatomical study. Neurosurgery 2014;74(4):401–424, discussion 424–425 34. Kamali A, Kramer LA, Frye RE, Butler IJ, Hasan KM. Diffusion tensor tractography of the human brain cortico-ponto-cerebellar pathways: a quantitative preliminary study. J Magn Reson Imaging 2010;32(4):809–817

24  Tractographic Anatomy of White Matter 35. Ramnani N, Behrens TE, Johansen-Berg H, et al. The evolution of prefrontal inputs to the cortico-pontine system: diffusion imaging evidence from Macaque monkeys and humans. Cereb Cortex 2006;16(6):811–818 36. Thiebaut de Schotten M, Ffytche DH, Bizzi A, et al. Atlasing location, asymmetry and inter-subject variability of white matter tracts in the human brain with MR diffusion tractography. Neuroimage 2011;54(1):49–59 37. Hong JH, Kwon HG, Jang SH. Probabilistic somatotopy of the spinothalamic pathway at the ventroposterolateral nucleus of the thalamus in the human brain. AJNR Am J Neuroradiol 2011;32(7):1358–1362 38. Fernandez-Miranda JC, Pathak S, Engh J, et al. High-­definition fiber tractography of the human brain: neuroanatomical validation and neurosurgical applications. Neurosurgery 2012;71(2):430–453 39. Holodny AI, Gor DM, Watts R, Gutin PH, Ulug AM. Diffusion-­ tensor MR tractography of somatotopic organization of corticospinal tracts in the internal capsule: initial anatomic results in c­ontradistinction to prior reports. Radiology 2005;234(3): 649–653 40. Kwon HG, Hong JH, Lee MY, Kwon YH, Jang SH. Somatotopic arrangement of the corticospinal tract at the medullary pyramid in the human brain. Eur Neurol 2011;65(1):46–49 41. Ozdemir NG. The anatomy of the posterior commissure. Turk ­Neurosurg 2015;25(6):837–843 42. Jarbo K, Verstynen T, Schneider W. In vivo quantification of global connectivity in the human corpus callosum. Neuroimage 2012;59(3):1988–1996 43. Witelson SF. Hand and sex differences in the isthmus and genu of the human corpus callosum. A postmortem morphological study. Brain 1989;112(Pt 3):799–835 44. Hofer S, Frahm J. Topography of the human corpus callosum ­ revisited—comprehensive fiber tractography using diffusion tensor magnetic resonance imaging. Neuroimage 2006;32(3):989–994

45. Rilling JK, Glasser MF, Preuss TM, et al. The evolution of the arcuate fasciculus revealed with comparative DTI. Nat Neurosci 2008;11(4):426–428 46. Fernandez-Miranda JC, Wang Y, Pathak S, Stefaneau L, Verstynen T, Yeh FC. Asymmetry, connectivity, and segmentation of the arcuate fascicle in the human brain. Brain Struct Funct 2015;220(3):1665–1680 47. Wang X, Pathak S, Stefaneanu L, Yeh FC, Li S, FernandezMiranda JC. Subcomponents and connectivity of the superior longitudinal fasciculus in the human brain. Brain Struct Funct 2016;221(4):2075–2092 48. Sarubbo S, De Benedictis A, Maldonado IL, Basso G, Duffau H. Frontal terminations for the inferior fronto-occipital fascicle: anatomical dissection, DTI study and functional considerations on a multi-component bundle. Brain Struct Funct 2013;218(1):21–37 49. Caverzasi E, Papinutto N, Amirbekian B, Berger MS, Henry RG. Q-ball of inferior fronto-occipital fasciculus and beyond. PLoS One 2014;9(6):e100274 50. Hau J, Sarubbo S, Perchey G, et al. Cortical terminations of the inferior fronto-occipital and uncinate fasciculi: anatomical stembased virtual dissection. Front Neuroanat 2016;10:58 51. Von Der Heide RJ, Skipper LM, Klobusicky E, Olson IR. Dissecting the uncinate fasciculus: disorders, controversies and a hypothesis. Brain 2013;136(Pt 6):1692–1707 52. Tusa RJ, Ungerleider LG. The inferior longitudinal f­asciculus: a reexamination in humans and monkeys. Ann Neurol 1985;18(5):583–591 53. Martino J, De Lucas EM. Subcortical anatomy of the lateral association fascicles of the brain: a review. Clin Anat 2014;27(4):563–569 54. Menjot de Champfleur N, Lima Maldonado I, Moritz-Gasser S, et al. Middle longitudinal fasciculus delineation within language pathways: a diffusion tensor imaging study in human. Eur J Radiol 2013;82(1):151–157 55. Yoshino M, Abhinav K, Yeh FC, et al. Visualization of cranial nerves using high-definition fiber tractography. Neurosurgery 2016;79(1):146–165

239

25

  Localization of Motor Cortex and Subcortical Pathways Using Functional Magnetic Resonance Imaging and Diffusion Tensor Imaging Sathish Kumar Dundamadappa and Mohit Maheshwari

Summary Presurgical localization of sensorimotor cortex and subcortical pathways is critical for preoperative risk assessment, planning of surgical trajectory, optimizing intraoperative mapping, and to obtain a maximal resection while preserving eloquent brain function. Whenever precise localization on structural imaging alone is not possible, functional magnetic resonance imaging (fMRI) plays a key role. It also increases surgeon’s confidence in localization. Functional anatomy of sensorimotor system, basic anatomy of corticospinal tract, common paradigms for sensorimotor mapping, and technique of diffusion tensor imaging are discussed in this chapter. Special considerations and challenges of functional imaging in young children are also discussed. Keywords:  presurgical localization, sensorimotor system, sensorimotor mapping, blood oxygen level dependent (BOLD) fMRI, presurgical white matter mapping, diffusion tensor imaging

„„ Introduction Precise localization of motor cortex and subcortical white matter pathways in relation to the epileptogenic zone or a lesion is critical part of the surgical planning when the surgical target is adjacent to the motor cortex. The most common reasons for the localization of motor cortex and subcortical white matter pathways are as follows: • yy yy yy

Preoperative risk assessment. Planning for intraoperative cortical or subcortical mapping. Planning operative approach and trajectory. Obtaining a maximal resection while preserving the ­eloquent brain function.

If there is a high degree of confidence in structural anatomical localization and the surgical lesion is away from the motor cortex and corticospinal tracts, then localization based on structural imaging alone is frequently good enough for surgical planning. However, localization on structural imaging is not always straightforward. fMRI localization is helpful in these cases.

„„ Functional MRI for Localization of the Motor Cortex Indications fMRI localization of the motor cortex is needed for the following reasons1: • There can be an anatomical variation (e.g., truncation of the precentral gyrus).2 yy Space-occupying lesions frequently result in mass effect and distortion of anatomical landmarks. yy Developmental malformations including malformations of cortical development can result in loss of anatomical landmarks and homotopic or heterotopic reorganization. yy Reorganization of cortical function in the setting of chronic or slow-growing abnormalities especially in pediatric patients. yy Need for homunculus localization, especially in patients who are not candidates for intraoperative cortical mapping (like younger pediatric patients). yy Multiplanar localization (e.g., no reliable structural landmarks for localization in coronal plane). yy Analyzing the significance of white matter bundles identified on diffusion tensor imaging (DTI) especially in cases with perinatal or prenatal insults where cortical reorganization is possible. yy In cases of small lesions that are not evident visually or with intraoperative ultrasound, their relation to preoperative fMRI activation can be correlated with intraoperative cortical mapping for better localization. yy If a lesion is located close to supplementary motor area (SMA), resection of those areas would result in inability to test motor function intraoperatively. With fMRI lesion’s relation to SMA can be illustrated preoperatively, and the order of lesion border dissections can be altered so that intraoperative testing is not affected. yy If there is discrepancy between the tumor morphology or location and the functional deficit, fMRI will help in assessing for reorganization.

25  Localization of Motor Cortex and Subcortical Pathways Using FMRI and DTI yy Within the motor cortex, only hand area can be identified with some certainty on structural images (hand knob). Otherwise, functional topographic assessment on s­ tructural imaging is inadequate. yy Intraoperative cortical stimulation of foot motor region is difficult secondary to its proximity to superior sagittal sinus.3 fMRI localization of this region is hence useful for surgical planning. yy Surgical planning and intraoperative navigation for the placement of percutaneous stereoelectroencephalography lead. fMRI is being considered a standard test in preoperative workup of patients with surgical lesions near motor cortex or ­corticospinal tract. It is the most commonly used noninvasive method to map brain cortical function. It has a high correlation with intraoperative cortical mapping, which is considered the gold standard.

Blood Oxygen Level Dependent fMRI Most fMRI studies use BOLD contrast for mapping. BOLD contrast depends on “neurovascular coupling,” which refers to regional increase in blood flow as a response to neural activity. The increase in local blood flow exceeds the increased oxygen demand by several times. It has been shown that increase in blood flow is about 30%, whereas increased oxygen consumption was only 5%.4 This results in higher venous oxygen saturation and affects the balance between oxyhemoglobin and deoxyhemoglobin. Deoxyhemoglobin is paramagnetic, and oxyhemoglobin is diamagnetic. Relative increase in oxyhemoglobin results in reduction in magnetic field inhomogeneity and hence a subtle increase in MRI signal intensity. The magnitude of this signal increase is small, typically about 2% at 1.5 Tesla (T) and about 4% at 3 T magnet. This small signal change cannot be directly seen in raw images and requires further processing and statistical analysis using the general linear model to produce “signal map.”

Stimulus Presentation It is preferable that fMRI studies are performed on a ­higher-strength magnet (3 T) as they provide higher signal to

Fig. 25.1  BOLD fMRI block design paradigm.

noise, which helps in capturing low BOLD signal. “Box-car block design” is the most commonly used paradigm design in clinical practice (Fig. 25.1). This involves alternating periods of task (typically involves motion in fingers or hand, toe, and tongue or lip) and control (typically rest for motor paradigms). The paradigms are presented to patient using visual stimulus projection. Statistical analysis of the data is typically performed on a separate workstation and then registered to the structural data.

Functional Anatomy of Sensorimotor System Sensorimotor system is complex and involves many cortical and subcortical areas with a precise location and specialized function.5 The complete extent is yet unknown. Overall, large frontal and parietal areas, cingulum, basal ganglia, and cerebellum contribute to different aspects of motor performance. However, from fMRI and surgical point of view, primary motor cortex, primary sensory cortex, premotor cortex (PMC), and supplemented motor area are the main regions subserving motor function.3 On fMRI motor tasks, activation is commonly seen in both motor and sensory gyri, and hence these are often together referred to as primary sensorimotor cortex.

Primary Motor Cortex Primary motor cortex (M1) is located in precentral gyrus (Brodmann’s area 4), lining its posterior bank and extending deep into the sulcus. It shows somatotopic organization, famously represented by “homunculus.” Corticobulbar areas are located inferolaterally (approximately at the level of superior longitudinal fasciculus), upper extremity located superolaterally, and the lower extremity located superomedially. The hand motor area is at the redundant cortex forming inverted omega or horizontal epsilon-shaped knob. This cortical redundancy provides the necessary larger representation for complex hand movements. Primary motor cortex is responsible for movements, which are planned and initiated by premotor areas. Bidirectional inputs to M1 include PMC, sensory cortex, and thalamus. Basal ganglia and cerebellum have indirect inputs via thalamus.6 Outputs from M1 are to bulbar motor neurons and spinal cord. Injury to M1 results in contralateral motor deficits, as well gait

241

242

III  Surgical Anatomy and Mapping Techniques ataxia with involvement of paracentral lobule.7 Corticospinal functions do not recover well, as opposed to ­corticobulbar deficits which tend to recover within several weeks to months.6

Secondary Motor Areas The secondary motor areas of interest from presurgical fMRI point of view are the premotor areas (located in Brodmann’s area 6), which primarily include SMA and PMC. The role of premotor areas is to plan, select, and initiate complex movements, thus guiding M1 function.6 These regions get activated temporally before the M1 activation. fMRI activation in these areas appear to increase with increased complexity of motor movements. SMA is located in medial Brodmann’s area 6 in medial bank of the superior frontal gyrus, just anterior to foot area in M1 and just superior to cingulate sulcus. Somatotopic organization is also seen in SMA, with facial region represented ­anteriorly.3 Another area of interest is the cingulate motor area. The ­caudal SMA is considered to primarily involved in movement execution,8 and the rostral portion is shown to have a role in action selection, initiation, motivation, and goal-directed behaviors.9,​10 PMC (lateral area 6) is located in posterior aspect of middle frontal gyrus, anterior bank of the precentral gyrus, and cortex surrounding precentral sulcus and posterior aspect of superior frontal sulcus. PMC is divided into dorsal and ventral cytoarchitectural regions. Dorsal PMC is centered adjacent to corticospinal representation of M1; ventral PMC is centered adjacent to corticobulbar M1.6 Unilateral motor tasks usually generate bilateral premotor activation on fMRI. Deficits of SMA produce apraxia, contralateral akinesia, alien hand syndrome, and perseveration.7 SMA deficits tend to recover almost completely.

Primary Somatosensory Cortex (S1) Primary somatosensory cortex is located in postcentral gyrus, along its dome, anterior bank, and posterior bank (Brodmann’s area 1, 2, 3a, and 3b). It has somatotopic organization. It also shows “hand knob” but not as prominent as the motor knob. Primary inputs are from somatosensory thalamus and outputs are to M1, premotor areas, and adjacent higher-order somatosensory cortex.6

Secondary Somatosensory Cortex Parietal cortex has multiple areas, which deal with specific aspects of sensory information. Among these regions, lateral parietal sensory cortex in inferior parietal lobule, known as SII, is considered to function in high-order processing of sensory information and play a role in attention, manual dexterity, coordination, temporal processing of somatic sensations, sensorimotor integration, tactile recognition, and tactile learning and memory.11,​12,​13,​14 This area receives inputs from SI and sensory thalamus. It projects to M1, PMC, insula, association areas, and spinal cord. Variable BOLD activation seen in this region depending on the stimulus shows a stronger contralateral than ipsilateral activation.15 SII may play a role in recovery from SI deficits. Numerous areas along the intraparietal sulcus have a role in integration of somatosensory and visuospatial information.

Motor tasks generate BOLD activation in both M1 and S1, typically centered in the central sulcus, in part secondary to sensory stimulation associated with the movement. The movement rate, force, and complexity correlate with the degree of BOLD activation. There is also ipsilateral deactivation, with transcallosal connections, the effect is greater for dominant versus nondominant hemispheres.16,​17 Peripheral sensory stimulation also activates both primary sensory and motor cortices, secondary to M1 activation from strong corticocortical connections.18,​19 Unilateral tasks can result in bilateral S1 activation secondary to interhemispheric connectivity, and contralateral activation is seen stronger than the ipsilateral activation.15 Passive movements may be used as an alternative or as a complimentary task to map sensorimotor homunculus.20

Common Sensorimotor Paradigms Start–Stop Motor Paradigm During the task block, the patient is instructed to perform a motor action (finger movement or toe movement or puckering of lips or back and forth movement of the tongue) whenever he or she sees auditory or visual cue which indicates “Start.” This alternates with control block “Stop” when the patient stays still. These blocks are generally 20 to 30 seconds each in duration and are repeated multiple times. Typical scan duration is 4 to 5 minutes. This is a simple technique and gives robust activation in primary sensorimotor areas (Fig. 25.2). Activation in secondary and association areas is less robust, given simplistic nature of the task.

Bilateral Complex Finger Tapping During the task block, patients repeatedly tap bilateral fingers on their thighs in the sequential order of 1, 3, 5, 2, and 4. This alternates with a control block which is rest. This paradigm given its complex architecture provides robust activation in bilateral sensorimotor cortex, bilateral basal ganglia, bilateral SMA, and superior cerebellum. Premotor activation is often noted.

Passive Movement or Sensory Paradigms These are utilized when the patient is unable to voluntarily perform motor function or follow directions (such as patients with developmental delay or limited understanding) or when the study is done under anesthesia or sedation (younger children typically less than 6 years old). During the task block, fingers or toes are passively moved (like flexion and extension of fingers or toes) or stroked with a brush (for sensory stimulation). This, like in other paradigms, is alternated with blocks of rest. Passive movements have shown to provide more robust BOLD activation compared to sensory tactile stimulation (Fig. 25.3). fMRI motor tasks are chosen depending on the location of the lesion in question. If entire motor strip needs to be mapped, it would prudent to get activation maps of fingers or hand, toes or foot, and tongue or lips (Fig. 25.4). Anatomical extrapolation using one specific regional activation (e.g., using finger movement activation to identify tongue region) is not always accurate. The inferior aspect of the central sulcus shortens and

25  Localization of Motor Cortex and Subcortical Pathways Using FMRI and DTI

Fig. 25.2  Sensorimotor mapping using “startstop” block design paradigm, demonstrating topographical representation. Color coding of the activation indicated by the font color.

Fig. 25.3  Passive sensorimotor mapping in a 4-year-old boy with intractable epilepsy.

moves anteriorly. It is common to have anatomical variations in this region. These factors make precise localization of facial region by anatomical features alone and by extrapolation of activation elsewhere in motor cortex difficult.

paradigms, such as young children, patients with altered mental status, sedated patients, and those who are paretic or ­aphasic,23,​24,​25 especially because BOLD activity seen in rs-fMRI has also been seen during sleep26 and anesthesia.27

Resting State fMRI

Validation in Literature

Resting state fMRI (rs-fMRI) measures the spontaneous lowfrequency fluctuations in the BOLD signal during rest. It helps in the study of brain functional architecture. With this technique, several resting state functional networks (consisting of spatially distinct areas of the brain with synchronous BOLD signal fluctuations at rest) can be identified.21 One such network is the sensorimotor network (Fig. 25.5), which consists of bilateral primary sensorimotor cortex, supplementary motor cortex, and secondary somatosensory cortex. Primary function of this network is detection and processing of sensory input and preparation and execution of motor functions.22 rs-fMRI is less demanding and can be performed on patients who may not otherwise be able to cooperate with task-based

Electric cortical stimulation (ECS) remains a gold standard for mapping eloquent cortex (including motor cortex). However, fMRI provides a relatively low cost and noninvasive alternative technique for presurgical mapping. Having this information prior to the surgery helps in prognostication and surgical planning including selection of the cases that would require intraoperative ECS. Therefore, fMRI is complementary to ECS is some cases and has essentially replaced it in many cases. Bartos et al28 compared the results of fMRI and ECS in primary hand motor area in 18 patients with tumor in the vicinity of precentral gyrus. Fifteen of 18 patients demonstrated high concordance between these two techniques. Another study by Pirotte and colleagues29 demonstrated 95% concordance (in

243

244

III  Surgical Anatomy and Mapping Techniques

Fig. 25.4  Passive sensorimotor mapping and DTI in a 10-month-old infant with intractable seizures. Extensive malformations of the cortical development (polymicrogyria, gray matter heterotopia—indicated by dots in a). Reorganization of the right motor strip (b), left foot is represented more laterally (activation denoted by the color green) (c). Corticospinal tractography (d, e) and color FA map (f), confirming the location of motor strip.

Fig. 25.5  Sensorimotor mapping with taskbased fMRI (on the left, smaller activations near the midline representing supplementary motor area) and with resting state fMRI (on the right) performed on the same day in a 12-year-old boy with epilepsy, showing nice correlation.

20/21 patients) between fMRI and ECS in patients with neuropathic pain. Petrella et al30 studied the effect of preoperative fMRI on therapeutic decision making in patients with resectable brain tumors. In 19 of 39 patients, there was a difference in treatment planning before and after fMRI with a more aggressive approach in 18 of them. They stated that because of fMRI, “in certain patients, surgical time may be shortened, the extent of resection increased, and craniotomy size decreased.” Hirsch et al31 found the overall sensitivity of fMRI was 100% for identification of sensorimotor area. For patients with tumors in these

eloquent areas, sensitivity was 97% for identification of sensorimotor area. Tiège et al32 investigated influence of the motor fMRI on surgical management of children and adolescent with symptomatic focal epilepsy. Success rate for fMRI motor mapping in this study was 93%. fMRI results contributed to surgical decisions in 74% (32/43) of patients including withholding surgery in nine patients. Passive and active sensorimotor mapping was compared by Kocak et al.20 They collected fMRI data from 11 healthy volunteers during active and passive movements of hand, elbow,

25  Localization of Motor Cortex and Subcortical Pathways Using FMRI and DTI shoulder, ankle, knee, and hip. It was found that primary motor cortex activation tended to increase with active compared with passive movements, although in the precentral gyrus, hand, elbow, and shoulder movements showed no statistically significant difference in mean number of activated voxels. In the postcentral gyrus, only the shoulder revealed a significant difference. In addition, 90% of the tasks employed, there was more than 50% proportional overlap of passive-on-active movement activation locations. Ogg et al33 showed that passive fMRI can accurately detect functional hand, leg, and face regions of the s­ ensorimotor cortex in the sedated child. In three cases, the active and passive fMRI colocalized the motor cortex. In four patients, ECS was also used to identify motor cortex, and in all four motor ECS yielded results consistent with the p ­ assive fMRI localization. Thirteen of 16 children have undergone resection based on passive fMRI findings with no unanticipated deficits.

„„ Diffusion Tensor Imaging Rationale for Presurgical White Matter Mapping Injury to an eloquent white matter tract could have similar deficits as surgical injury to corresponding eloquent cortex. In the areas where there are compact fibers (like corona radiata, internal capsule), deficits from even a small injury can be profound. Hence, presurgical white matter mapping is as crucial as cortical mapping. In clinical practice, white matter mapping is accomplished a technique called DTI. This is a much less demanding test in terms of patient cooperation compared to BOLD imaging. Wu et al34 analyzed the effect of DTI-based neuronavigation in surgery of cerebral gliomas with pyramidal one-tract involvement. With high-grade gliomas, there was ­ higher chance of gross total resection (74.4 vs. 33.3%) and lower chance of postoperative motor deterioration (15.3 vs. 32.8%). The 6-month Karnofsky performance scale score was 86 ± 24 in the study group compared to 74 ± 28 in the control group. With high-grade gliomas, the median survival for study cases was 21.2 months versus compared to 14 months for the control group. Correlation of diffusion tensor fiber tracking and intraoperative stimulation is validated in several studies.35,​36

Imaging Technique Diffusion-weighted imaging and DTI use thermal motion of water molecules to probe microstructural properties of the brain. Diffusion sequence is constructed by adding a pair of diffusion-sensitizing gradients to a T2-weighted spin-echo MR sequence. Water molecular motion due to diffusion results in loss of signal intensity due to incomplete rephasing of proton spins; conversely, loss of motion results in high signal intensity. In isotropic diffusion, motion is equal in all directions, for example, cerebrospinal fluid spaces (Fig. 25.6). Gray matter also demonstrates near isotropic diffusion. In white matter tracts, diffusion tends to occur more readily along the length of fiber tracts as opposed to orthogonal fibers. This is known as anisotropic diffusion. This anisotropy is exploited in DTI technique.

Several diffusion tensor parameters are used to characterize microstructural properties of brain: the three eigenvectors (lambda 1, lambda 2, and lambda 3). The primary eigenvector (lambda 1) indicates the direction and magnitude of greatest water diffusion, which is parallel to fiber direction. Lambda 1 is known as “longitudinal diffusivity.” The mean of lambda 2 and lambda 3 is known as “radial diffusivity.” Mean diffusivity refers to the mean of three eigenvector values. Fractional anisotropy (FA), the value ranging from a minimum of 0 (which would be equivalent to isotropic diffusion) to a maximum of 1 (indicating a strong anisotropy and diffusion occurring only along the primary eigenvector).37 Color FA maps are the most commonly used tensor parameter in clinical imaging. There are standard color conventions for these FA maps such that they depict the direction of the principle eigenvector in each voxel which in turn depicts direction of the bulk of axons or white matter tract. Red color indicates right–left (or left to right) direction; green indicates anteroposterior (or posteroanterior) direction; and blue indicates craniocaudal (or caudocranial) direction. Color intensity depicts the degree of anisotropy. From DTI raw data white matter tractography may also be performed. In clinical practice, fiber assignment by continuous tracking is the most commonly used method. In this method, the primary eigenvector is considered to be the fiber direction within the voxel at the seed point, and when it reaches the edge of the voxel, the trajectory is changed to match the primary eigenvector of the next voxel. A minimum FA (0.1–0.3) is required for continuation of this tracking into the subsequent voxel. A constraint is placed on the maximum angulation (40– 70 degrees), beyond which the tracking is terminated. Tractography helps to better visualize the relationship between the lesion and a particular tract. When more than one tract is bundled together in FA color map, tractography may help in delineating individual components. In clinical practice, radiologists generally assess the white matter tracts on FA maps and utilize tractography to better illustrate the relationship in three dimensions.

Basic Anatomy of Corticospinal Tract Corticofugal fibers descend through internal capsule and pass into the brainstem. Of these, corticobulbar fibers terminate in the brainstem. Corticospinal fibers descend along the entire length of the brainstem, majority crossing to the contralateral side in the medullary decussations, continuing in the spinal cord as lateral corticospinal tract.38 The main origin of corticospinal system is the motor cortex in precentral gyrus and middle part of the paracentral lobule. About 80% of the fibers come from precentral gyrus.39 Smaller number of fibers coming from somatosensory areas, which end synaptically at the afferent nerve cells in the spinal posterior gray horns, influencing the efferent signals.40 Then they traverse through corona radiata, internal capsule, cerebral crus, anterior aspect of the pons, and medulla oblongata. In the upper part of the diencephalon, corticospinal tract is located approximately in the middle third of posterior limb of internal capsule and in the lower part of the diencephalon, in the posterior third with the exception of posterior most region. In the midbrain, it is in middle third of cerebral crus. In the pons, it is located in the anterior aspect with pontine nuclei located between

245

246

III  Surgical Anatomy and Mapping Techniques

Fig. 25.6  Schematic representation of isotropic and anisotropic diffusion. Schematic representation of gray matter in (a), water molecular movement represented by red, equal in all directions (b). Schematic representation of white matter bundle in (c), preferential water molecular movement along the bundle direction represented in (d).

the fibers. In the medulla, pyramidal pathways form bilateral anterior protrusions. In lower medulla, 80 to 90% fibers cross at decussations and form lateral corticospinal tract. The smaller uncrossed portions of the pyramidal pathway descend as anterior corticospinal tract. Corticospinal tract anatomy is depicted in Fig. 25.7 and Fig. 25.8.

„„ Essential Considerations and Limitations with BOLD fMRI and DTI Techniques Various preoperative and perioperative techniques including clinical presentation or deficits, functional anatomy at standard imaging, bold fMRI, DTI, and intraoperative cortical and subcortical mapping need to be used together to assess functional network proximity risk. Each of these when considered alone is probably imperfect indicator of risk, but together can provide better risk assessment.41 As far as preoperative MRI is concerned, combined use of BOLD fMRI and DTI is superior to fMRI alone for preoperative risk assessment (Fig. 25.9). ­Negative ­predictive value for postoperative deficit can approach 100% using these techniques together.42

DTI as well as fMRI are best used as qualitative assessment of spatial relationships in presurgical mapping. Restrain should be used when measuring the actual distances from the lesion to a millimeter level. Phenomenon of intraoperative “brain shifting” should also be kept in mind while using the preoperative data during surgery. In infiltrating tumors, intratumoral activation can be seen, which would be displaced and scattered and correlates with the degree of infiltration.43 In noninfiltrating tumors, there is extratumoral shift of activation. Significant reduction in BOLD signal can be seen adjacent to glial tumors in comparison to the contralateral side, most pronounced in glioblastoma.44 This presumably is related to tumor-induced changes in cerebral hemodynamics, direct loss of cortical neurons, and to some extent, neurovascular ­uncoupling. In clinical practice, color-coded FA maps of DTI superimposed on anatomical imaging are the mainstay to assess the relationships of white matter tracts to the lesion. Lack of visualization of a tract does not necessarily mean involvement by the disease process. Anatomical distortion and displacement may obscure or reorient tracts, making them difficult to discern on DTI techniques. Tumor or edema can change the anisotropy, affecting the visibility of tracts. If a tract is visible, it has good

25  Localization of Motor Cortex and Subcortical Pathways Using FMRI and DTI

Fig. 25.7  Color FA maps depicting corticospinal tract (notated by diamond shape on the right side). Note “RGB” color convention depicting direction.

Fig. 25.8  (a) Coronal FA map showing corticospinal tract (right corticospinal tract indicated by the arrow). (b) Tractography by FACT method depicting corticospinal tract.

negative predictive value. If it is not, it can mean either involvement or distortion or displacement by the lesion. Functionally distinct tracts running in parallel within the same voxel cannot be distinguished. With tractography, beautifully depicted colored structures are not actual fiber bundles, they are just visual representation of fiber direction or density or fractional isotropy as assessed by DTI techniques. This technique assumes a single direction within a voxel, which is not always true. If there are complex crossing fibers within the voxel, the current clinical DTI techniques fail to depict all the fibers. Acute fiber angulations at the interface between cortex and white matter can also result in spurious lack of their revisualization (e.g., good portion of corticobulbar fibers are not depicted in pyramidal tractography). Correlation of multiple techniques (structural imaging, cortical

mapping using BOLD fMRI and DTI) become extremely crucial in such instances. BOLD fMRI is a statistical method. The area of activation can vary considerably based on the “selected threshold.” The measured signal changes are related to changes in relative oxyhemoglobin concentration in the vessels and not directly to neuronal activity. Hence, caution should be exercised while measuring the distance between the activity and the lesion margin. The surgery is generally considered safe if the lesion margin is more than 10 mm45 or is separated by more than a gyral width. fMRI has poor temporal resolution, in the order of seconds, as opposed to some other neurodiagnostic techniques (electroencephalograph and magnetoencephalography in ­milliseconds). fMRI indirectly measures neural activity by assessing its hemodynamic response. If there is neurovascular uncoupling

247

248

III  Surgical Anatomy and Mapping Techniques (­common causes are arteriovenous [AV] malformations and AV shunting), it results in spurious reduction in fMRI signal. The echoplanar imaging technique itself is prone for artifacts especially at tissue-air and tissue-bone junctions and near paramagnetic or ferromagnetic objects (hemorrhage, surgical ­ clips, etc.) (Fig. 25.10). Image noise, crossing fibers, divergent or conversion trajectories can alter orientation of major eigenvector

within the voxel and can falsely terminate a fiber tract. Negative fiber tracking does not rule out the presence of fiber tracts, especially when invaded by the tumor or in the vicinity of the tumor. Motion artifacts and other MRI artifacts can affect the study. The term “bordering the lesion” or “immediate proximity” is used when a tract or BOLD activation is contacting or within a few millimeters from the lesion border. The term “remote”

Fig. 25.9  BOLD sensorimotor mapping and corticospinal tractography (arrow in d) showing close relationship of the mass lesion (metastasis from osteogenic sarcoma) with the motor system (a–c, e).

Fig. 25.10  An 8-year-old patient with prior epilepsy surgery. No BOLD activation seen on the expected location in the primary motor cortex (right precentral gyrus) during left finger tapping. Activation was seen in supplementary motor area (shown as red). This spurious lack of activation is related to susceptibility artifact caused by the metallic vascular clip over right posterior frontal convexity.

25  Localization of Motor Cortex and Subcortical Pathways Using FMRI and DTI is used when the functional area is more than a centimeter from the tumor or surrounding signal intensity. The term “relative proximity” is used to describe the location somewhere between “immediate” and “remote.”41 Though there is general agreement on the utility of fMRI and it is generally considered the standard of care, it is still not fully validated.

Special Considerations and Challenges in Young Children Immature networks and differences in vascular reactivity may contribute to negative BOLD response during visual and sensorimotor tasks in neonates and infants.46,​47,​48 Beyond first weeks of life, BOLD hemodynamic response is comparable to adult activation with possible increase in amplitude until adulthood.49 It is crucial to assess the performance status of the patient prior to the test. Tasks need to be modified to the performance level of the patient. In general, cooperation can be obtained from children with a developmental age of 5 to 6 years or IQs of around 60, provided there are no behavioral disorders.50 Study acquisition can be split into two or more sessions for better patient cooperation. Images from different sessions can be coregistered during analysis. Careful protective measures must be taken to prevent side effects of acoustic noise of echoplanar imaging of fMRI, specifically in infants and sedated patients. Coarse real-time fMRI analysis may be of value, as it provides continuous monitoring of quality of the acquisition. Head motion is a critical factor. During

preprocessing of fMRI data, special attention needs to be paid for the degree of motion and the quality of correction. Passive sensorimotor mapping (with or without sedation) can be considered in patients who are unable to perform active tasks. Though still an evolving technique, resting state fMRI provides a good depiction of sensorimotor system. Bernal et al51 have compared various anesthetic agents and their effects on fMRI examination results and patient tolerance. In their institute, combination of dexmedetomidine and propofol is t­ypically used for sedation for passive fMRI studies. In our experience also, these agents work well. Plasticity of cortical organization tends to be seen with higher frequency in children. Studies have shown contralateral motor area taking over impaired function in patients with congenital hemiplegia.52

„„ Conclusion Sensorimotor mapping with BOLD imaging and diffusion tensor imaging are extremely useful techniques in clinical practice, especially when there is distortion of anatomical landmarks, congenital malformations, and for multiplanar localization. In conjunction with standard anatomical imaging, they offer valuable information for preoperative risk assessment, surgical planning, and reduce the time required for intraoperative cortical or subcortical mapping. It is equally vital to be mindful of limitations of these techniques.

References 1. Ulmer JL, DeYoe E, Gaggl W, et al. Half A Dozen Reasons to Map the Motor Cortex. American Society Neuroradiology Conference. Montreal, Canada; 2014 2. Salvan CV, Ulmer JL, Mueller WM, Krouwer HGJ, Prost RW, Stroe GO. Presurgical and intraoperative mapping of the motor system in congenital truncation of the precentral gyrus. AJNR Am J Neuroradiol 2006;27(3):493–497 3. Gabriel M, Brennan NP, Peck KK, Holodny AI. Blood oxygen level dependent functional magnetic resonance imaging for presurgical planning. Neuroimaging Clin N Am 2014;24(4):557–571 4. Fox PT, Raichle ME. Focal physiological uncoupling of cerebral blood flow and oxidative metabolism during somatosensory stimulation in human subjects. Proc Natl Acad Sci USA 1986;83(4):1140–1144 5. Filippi M, ed. FMRI Techniques and Protocols. New York, NY: Humana; 2009 6. Ulmer JL, Klein AP, Mark LP, Tuna I, Agarwal M, DeYoe E. Functional and dysfunctional sensorimotor anatomy and imaging. Semin Ultrasound CT MR 2015;36(3):220–233 7. Brazis PW, Masdeu JC, Biller J. Localization in Clinical Neurology. 6th ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2011:657 8. Fink GR, Frackowiak RS, Pietrzyk U, Passingham RE. Multiple nonprimary motor areas in the human cortex. J Neurophysiol 1997;77(4):2164–2174 9. Devinsky O, Morrell MJ, Vogt BA. Contributions of anterior cingulate cortex to behaviour. Brain 1995;118(Pt 1):279–306 10. Deiber MP, Passingham RE, Colebatch JG, Friston KJ, Nixon PD, Frackowiak RSJ. Cortical areas and the selection of movement: a study with positron emission tomography. Exp Brain Res 1991;84(2):393–402

11. Karhu J, Tesche CD. Simultaneous early processing of sensory input in human primary (SI) and secondary (SII) somatosensory cortices. J Neurophysiol 1999;81(5):2017–2025 12. Huttunen J, Wikström H, Korvenoja A, Seppäläinen AM, Aronen H, Ilmoniemi RJ. Significance of the second somatosensory cortex in sensorimotor integration: enhancement of sensory responses during finger movements. Neuroreport 1996;7(5):1009–1012 13. Hämäläinen H, Hiltunen J, Titievskaja I. fMRI activations of SI and SII cortices during tactile stimulation depend on attention. Neuroreport 2000;11(8):1673–1676 14. Mima T, Nagamine T, Nakamura K, Shibasaki H. Attention modulates both primary and second somatosensory cortical activities in humans: a magnetoencephalographic study. J Neurophysiol 1998;80(4):2215–2221 15. Blatow M, Nennig E, Durst A, Sartor K, Stippich C. fMRI reflects functional connectivity of human somatosensory cortex. Neuroimage 2007;37(3):927–936 16. Tzourio-Mazoyer N, Petit L, Zago L, et al. Between-hand difference in ipsilateral deactivation is associated with hand lateralization: fMRI mapping of 284 volunteers balanced for handedness. Front Hum Neurosci 2015;9:5 17. Hayashi MJ, Saito DN, Aramaki Y, Asai T, Fujibayashi Y, Sadato N. Hemispheric asymmetry of frequency-dependent suppression in the ipsilateral primary motor cortex during finger movement: a functional magnetic resonance imaging study. Cereb Cortex 2008;18(12):2932–2940 18. Wannier TM, Maier MA, Hepp-Reymond MC. Contrasting properties of monkey somatosensory and motor cortex neurons activated during the control of force in precision grip. J Neurophysiol 1991;65(3):572–589

249

250

III  Surgical Anatomy and Mapping Techniques 19. Rosén I, Asanuma H. Peripheral afferent inputs to the forelimb area of the monkey motor cortex: input-output relations. Exp Brain Res 1972;14(3):257–273 20. Kocak M, Ulmer JL, Sahin Ugurel M, Gaggl W, Prost RW. Motor homunculus: passive mapping in healthy volunteers by using functional MR imaging—initial results. Radiology 2009;251(2):485–492 21. Lee MH, Smyser CD, Shimony JS. Resting-state fMRI: a review of methods and clinical applications. AJNR Am J Neuroradiol 2013;34(10):1866–1872 22. Barkhof F, Haller S, Rombouts SARB. Resting-state functional MR imaging: a new window to the brain. Radiology 2014;272(1):29–49

35. Bello L, Gambini A, Castellano A, et al. Motor and language DTI Fiber Tracking combined with intraoperative subcortical mapping for surgical removal of gliomas. Neuroimage 2008;39(1):369–382 36. Coenen VA, Krings T, Axer H, et al. Intraoperative threedimensional visualization of the pyramidal tract in a neuronavigation system (PTV) reliably predicts true position of principal motor pathways. Surg Neurol 2003;60(5):381–390, discussion 390 37. Mukherjee P, Berman JI, Chung SW, Hess CP, Henry RG. Diffusion tensor MR imaging and fiber tractography: theoretic underpinnings. AJNR Am J Neuroradiol 2008;29(4):632–641 38. Standring S, ed. Gray’s Anatomy: The Anatomical Basis of Clinical Practice. 41st ed. New York, NY: Elsevier Limited; 2016

23. Liu H, Buckner RL, Talukdar T, Tanaka N, Madsen JR, Stufflebeam SM. Task-free presurgical mapping using functional magnetic resonance imaging intrinsic activity. J Neurosurg 2009;111(4):746–754

39. FitzGerald MJT, Folan-Curran J. Clinical Neuroanatomy and Related Neuroscience. 4th ed. London: Saunders; 2002

24. Kokkonen SM, Nikkinen J, Remes J, et al. Preoperative localization of the sensorimotor area using independent component analysis of resting-state fMRI. Magn Reson Imaging 2009;27(6):733–740

41. Ulmer JL, Klein AP, Mueller WM, DeYoe EA, Mark LP. Preoperative diffusion tensor imaging: improving neurosurgical outcomes in brain tumor patients. Neuroimaging Clin N Am 2014;24(4):599–617

25. Shimony JS, Zhang D, Johnston JM, Fox MD, Roy A, Leuthardt EC. Resting-state spontaneous fluctuations in brain activity: a new paradigm for presurgical planning using fMRI. Acad Radiol 2009;16(5):578–583 26. Fukunaga M, Horovitz SG, van Gelderen P, et al. Large-­ amplitude, spatially correlated fluctuations in BOLD fMRI signals during extended rest and early sleep stages. Magn Reson Imaging 2006;24(8):979–992 27. Peltier SJ, Kerssens C, Hamann SB, Sebel PS, Byas-Smith M, Hu X. Functional connectivity changes with concentration of sevoflurane anesthesia. Neuroreport 2005;16(3):285–288 28. Bartos R, Jech R, Vymazal J, et al. Validity of primary motor area localization with fMRI versus electric cortical stimulation: a comparative study. Acta Neurochir (Wien) 2009;151(9):1071–1080 29. Pirotte B, Neugroschl C, Metens T, et al. Comparison of functional MR imaging guidance to electrical cortical mapping for targeting selective motor cortex areas in neuropathic pain: a study based on intraoperative stereotactic navigation. AJNR Am J Neuroradiol 2005;26(9):2256–2266 30. Petrella JR, Shah LM, Harris KM, et al. Preoperative functional MR imaging localization of language and motor areas: effect on therapeutic decision making in patients with potentially resectable brain tumors. Radiology 2006;240(3):793–802 31. Hirsch J, Ruge MI, Kim KH, et al. An integrated functional magnetic resonance imaging procedure for preoperative mapping of cortical areas associated with tactile, motor, language, and visual functions. Neurosurgery 2000;47(3):711–721, discussion 721–722 32. De Tiège X, Connelly A, Liégeois F, et al. Influence of motor functional magnetic resonance imaging on the surgical management of children and adolescents with symptomatic focal epilepsy. Neurosurgery 2009;64(5):856–864, discussion 864 33. Ogg RJ, Laningham FH, Clarke D, et al. Passive range of motion functional magnetic resonance imaging localizing sensorimotor cortex in sedated children. J Neurosurg Pediatr 2009;4(4):317–322 34. Wu JS, Zhou LF, Tang WJ, et al. Clinical evaluation and follow-up outcome of diffusion tensor imaging-based functional neuronavigation: a prospective, controlled study in patients with gliomas involving pyramidal tracts. Neurosurgery 2007;61(5):935–948, discussion 948–949

40. Martin J. Neuroanatomy: Text and Atlas. 3rd ed. New York, NY: McGraw-Hill Medical; 2003

42. Ulmer JL, Berman JI, Mueller WM, et al. Issues in translating imaging technology and presurgical diffusion tensor imaging. In: Functional Neuroradiology. Boston, MA: Springer; 2011:731–765 43. Roux FE, Ranjeva JP, Boulanouar K, et al. Motor functional MRI for presurgical evaluation of cerebral tumors. Stereotact Funct Neurosurg 1997;68(1–4 Pt 1):106–111 44. Krings T, Töpper R, Willmes K, Reinges MHT, Gilsbach JM, Thron A. Activation in primary and secondary motor areas in patients with CNS neoplasms and weakness. Neurology 2002;58(3):381–390 45. Håberg A, Kvistad KA, Unsgård G, Haraldseth O. Preoperative blood oxygen level-dependent functional magnetic resonance imaging in patients with primary brain tumors: clinical application and outcome. Neurosurgery 2004;54(4):902–914, discussion 914–915 46. Marcar VL, Strässle AE, Loenneker T, Schwarz U, Martin E. The influence of cortical maturation on the BOLD response: an fMRI study of visual cortex in children. Pediatr Res 2004;56(6):967–974 47. Morita T, Kochiyama T, Yamada H, et al. Difference in the metabolic response to photic stimulation of the lateral geniculate nucleus and the primary visual cortex of infants: a fMRI study. Neurosci Res 2000;38(1):63–70 48. Heep A, Scheef L, Jankowski J, et al. Functional magnetic resonance imaging of the sensorimotor system in preterm infants. Pediatrics 2009;123(1):294–300 49. Shapiro KL, Johnston SJ, Vogels W, Zaman A, Roberts N. Increased functional magnetic resonance imaging activity during nonconscious perception in the attentional blink. Neuroreport 2007;18(4):341–345 50. Hertz-Pannier L, Noulhiane M, Rodrigo S, Chiron C. Pretherapeutic functional magnetic resonance imaging in children. Neuroimaging Clin N Am 2014;24(4):639–653 51. Bernal B, Grossman S, Gonzalez R, Altman N. FMRI under sedation: what is the best choice in children? J Clin Med Res 2012;4(6):363–370 52. Staudt M, Pavlova M, Böhm S, Grodd W, Krägeloh-Mann I. Pyramidal tract damage correlates with motor dysfunction in bilateral periventricular leukomalacia (PVL). Neuropediatrics 2003;34(4):182–188

26

  The Wada Test: Lateralization of Language and Memory David W. Loring and Gregory P. Lee

Summary The Wada test remains an essential component of the preoperative evaluation of epilepsy patients at many epilepsy surgery centers. The role of Wada memory testing has evolved since its introduction. In addition to establishing cerebral language representation preoperatively, Wada memory results may be used to establish risk for postoperative memory decline and to assist in identification of focal functional deficits associated with a unilateral seizure focus. The majority of clinical experience with Wada testing has been derived from adults. However, because neurodevelopment in pediatric patients is incomplete, patterns of expected performance in adults cannot necessarily be generalized to pediatric groups, and the predictive ability of Wada testing to accurately forecast long-term cognitive outcomes may be altered by neuroplasticity and cognitive maturation. Wada memory results are used to counsel patients regarding the likelihood of memory decline, which, even if not frank amnesia, is of sufficient severity to interfere with quality of life, cognitive development, and other factors that may affect school performance. Although Wada memory testing is less likely to yield helpful lateralizing information compared with adult studies, Wada memory asymmetries (WMAs) are able to indicate which patients may be at higher risk for postoperative memory decline. Keywords:  Wada test, intracarotid injection of amobarbital, memory testing, language lateralization

„„ Introduction While functional magnetic resonance imaging is now used frequently in epilepsy surgery cases, the Wada test remains an essential component of the preoperative evaluation of epilepsy patients at many epilepsy surgery centers. In addition to establishing cerebral language representation preoperatively, Wada memory results may be used to establish risk for postoperative memory decline and to assist in identification of focal functional deficits associated with a unilateral seizure focus.1–​4 Although a substantial Wada clinical literature in adults exists, there have been relatively few reports describing Wada test experience in pediatric epilepsy surgery candidates. Juhn Wada introduced his technique of intracarotid injection of amobarbital in the 1950s to establish cerebral language

representation in adult patients who were undergoing evaluation for epilepsy surgery.5 The procedure relies on a short-­ acting barbiturate introduced into the internal carotid artery that temporarily anesthetizes the anterior two-thirds of a cerebral hemisphere during which language testing is conducted. The most common anesthetic agent is amobarbital, although other drugs are successfully used, including etomidate,6 methohexital,7 and propofol.8 These newer agents have a shorter duration of action than amobarbital, and, in the case of etomidate, a constant infusion of drug is necessary to produce a sufficient length of anesthesia to permit language and memory testing. After observing several cases of unanticipated significant decline in memory function after temporal lobectomy, a memory component was introduced as part of the Wada test to provide a reversible model of temporal lobe surgery in which the risk of developing severe anterograde amnesia could be estimated.9 The Wada test thus creates a reversible pharmacological lesion in which induced behavioral deficits are thought to reflect surgical risk of including these areas in a surgical resection. At most centers performing the Wada, both language and memory functions are assessed. Language results often guide specific clinical decision-­ making. When surgery is planned in a hemisphere dominant for language, generally more conservative surgical ­approaches are used, and additional measures to protect eloquent language cortex such as electrocortical stimulation mapping are performed. Although the goals of Wada memory testing vary across epilepsy centers, the test’s primary purpose is to establish the risk for postoperative memory decline. In addition, however, interhemispheric memory asymmetry scores may be used to help confirm seizure-onset laterality in patients with less clearly established seizure onset. Patients in whom both structural and functional measures of unilateral mesial temporal lobe dysfunction are in agreement tend to have superior surgical efficacy, as well as decreased cognitive morbidity, in comparison with patients in whom there is incomplete agreement regarding lateralized impairment. The majority of clinical experience with Wada testing has been derived from adults. However, because neurodevelopment in pediatric patients is incomplete, patterns of expected performance in adults cannot necessarily be generalized to pediatric groups, and the predictive ability of Wada testing to accurately forecast long-term cognitive outcomes may be altered by ­neuroplasticity and cognitive maturation.

252

III  Surgical Anatomy and Mapping Techniques

„„ Special Considerations for Pediatric Wada Testing Testing of pediatric epilepsy patients presents unique challenges compared with adults. The Wada test is often a physically uncomfortable and emotionally frightening procedure for younger children to undergo, and they sometimes lack the appropriate maturity to participate fully in all aspects of the evaluation. Even before the potential effects of drug-induced behavioral deficits, which themselves may be frightening, the surgical aspects of catheter placement for medication delivery often exceed a child’s capacity to tolerate and cooperate with the procedure. Coaching and developing behavioral intervention techniques may be useful in decreasing the anxiety associated with the procedure, but these approaches are often of limited utility, are labor and time intensive, and are less useful in younger patients or in children with decreased cognitive abilities and less insight. Pre-Wada baseline assessment is more extensive and carefully constructed in children than adults because the stimulus materials selected for use during the Wada must be tailored to the developmental and cognitive level of each child. Because of these constraints, children younger than 7 or 8 years old are generally considered unsuitable candidates for the procedure. In youngsters who are unusually sensitive to pain or who may have difficulties cooperating for other reasons, sedation may be administered by an anesthesiologist to assist with catheter placement.10 Propofol, for example, is sufficiently short acting that anesthesia recovery is rapid, and Wada testing can be performed within 15 to 25 minutes after propofol cessation. Thus, anxiety and discomfort associated with catheter insertion can be avoided using propofol in appropriate cases, thereby maximizing the likelihood of obtaining valid Wada behavioral results. Unfortunately, some children awaken from the anesthesia disoriented or overly emotional and require much soothing before they are capable of proceeding with Wada testing. In some cases, a parent may be an asset in helping to calm the child in the immediate postanesthesia period. Parents may then be escorted out of the angiography suite after they have quieted their children but before amobarbital injection.

„„ Wada Language Testing in Pediatrics Because children who are evaluated for epilepsy surgery have extratemporal lobe epilepsy more often than adults, the critical information derived from Wada testing is often language laterality and representation, and Wada memory findings assume less importance. The need for valid determination of language lateralization by Wada testing is clear, if the gold standard of language localization is electrocortical stimulation mapping. Although stimulation mapping has reportedly been successfully performed in children as young as 4 years of age,11 in addition to the greater difficulties involved in stimulation language mapping with children, language cortex is less likely to be identified with mapping in children under 10 years of age.12 In one recent study, the presence of positive language results from stimulation language mapping was no different in children 10 years or older than it was in adults. Although these

authors also described less reliable Wada language results for children younger than 10 years compared with older children, the magnitude of this difference was not as marked. Thus, Wada language testing appears more likely to succeed in identifying language representation in children than electrocortical stimulation mapping does. Overall level of cognitive ability has been associated with the likelihood of obtaining useful Wada information in children.13 In a small series of 22 pediatric patients (ages 5–12 years), language testing was successful in all of the children with intellectual abilities (IQs) of at least 70, whereas only 57% of the children studied satisfactorily completed Wada language testing if their IQs were below 70. A similar pattern was seen with Wada memory results. Children with IQs of 70 or higher had good retention scores after injection ipsilateral to seizure onset but impaired retention after contralateral injection, whereas children with IQs of at least 70 were much less likely to show this lateralized discrepancy. In another pediatric series, the Wada procedure successfully established hemispheric language dominance and memory representation in fewer than twothirds of 42 preadolescent candidates for epilepsy surgery.14 Risk factors for unsuccessful testing included low full-scale IQ (especially < 80), young age (especially < 10 years), and left hemisphere seizure onset. The symptoms of aphasia seen in children during Wada testing also differ from those typically seen in adults, which may complicate interpretation. Many times, children simply become mute during Wada testing, and hence, there are often no positive signs of aphasia, such as paraphasic substitution errors or circumlocutions, to help confirm that language has been affected.

„„ Wada Memory Testing in Pediatrics The role of Wada memory testing has evolved since its introduction. The need to identify patients at risk for the development of a persistent frank amnesia by identifying significant contralateral mesial temporal lobe damage to the proposed temporal lobe surgery has been a significant topic in preoperative assessment of epilepsy surgery patients. Wada memory results are used to counsel patients regarding the likelihood of memory decline, which, even if not frank amnesia, is of sufficient severity to interfere with quality of life, cognitive development, and other factors that may affect school performance.15 Wada memory testing is intended to assess the functional integrity of the mesial temporal lobe and, to a smaller degree, the entire hemisphere being perfused with the anesthetic. This procedure differs significantly from other functional assessments in that it assesses each hemisphere in isolation, thereby helping to disentangle the effects of parallel distributed brain networks. The functional reserve capacity of the contralateral temporal lobe to sustain memory function in isolation is assessed when the hemisphere ipsilateral to a mesial temporal lobe focus is anesthetized16 and was the original goal of Wada memory testing when the procedure was first designed to avoid postoperative global amnesia. Because there are varying degrees of residual function in the diseased temporal lobe, the potential mnemonic contributions of the mesial temporal lobe structures ipsilateral to the seizure focus also must be assessed. This evaluates the functional adequacy of the diseased t­emporal lobe.

26  The Wada Test: Lateralization of Language and Memory Functional adequacy is assessed during injection contralateral to seizure onset. The relative differences between the memory performance of each hemisphere are termed WMA. Although Wada memory results are not a primary measure to lateralize seizure onset, WMAs may help clinical decision making when interpreted in the context of other clinical findings. The complex process of clinical determination of seizure-onset laterality often relies on the convergence of findings from multiple sources. Thus, WMAs have clinical implications that can either increase or decrease the confidence of unilateral seizure onset in cases with difficult-to-lateralize seizures. Patients with clinical findings that are not in complete agreement regarding seizure-onset laterality or location are often thought to be less ideal candidates, with decreased likelihood of becoming seizure free and an increased risk of postoperative cognitive morbidity. WMAs in pediatric populations correspond to seizure-onset laterality both on the group level and on individual patient basis.17 In a series of 87 children from three different institutions, Wada memory testing was able to accurately lateralize seizure onset in 69% of the sample, a rate that is somewhat lower than the rate of 70 to 88% correct classification in studies of adult surgical candidates. This slightly worse rate of seizure lateralization prediction in children is probably caused by the lower prevalence of temporal lobe seizures in the pediatric ­population. Because Wada protocols are not standardized, it is sometimes difficult to estimate to what degree method variance contributes to some of the reported variability in memory outcome prediction. We have shown that factors such as stimulus type (pictures versus real objects),18 timing of stimulus ­presentation,19 mixed stimuli requiring a verbal response, and amobarbital dose20 are related to Wada memory correlations with seizure-onset laterality. The potential confound of aphasia on certain verbal memory stimuli is well recognized,21 and generalizations of specific results to other Wada memory protocols must necessarily be made cautiously.22 As is the case in the adult Wada experience, a specific Wada protocol is used for assessment involves the likelihood of obtaining lateralized findings.23 Wada memory testing involved the presentation of information to be remembered after the introduction of medication, and this material may include line drawings, pictures, real objects, words, or designs. Compared with real objects, however, mixed stimulus Wada memory testing appears less sensitive to unilateral seizures in children and adolescents. This method discrepancy is greater in children with left-sided seizure onset and is seen on both the group and individual patient level. Further, the risk of incorrect classification based on WMAs is greater in the mixed stimulus method compared with real object. For example, when using the mixed stimulus Wada memory testing, approximately onethird of children in both the temporal and nontemporal groups had seizure-onset laterality incorrectly lateralized. The mixed stimulus method also incorrectly predicted the side of seizure onset in 25% of right hemisphere seizure patients. This finding contrasts with WMAs obtained using real objects in which 18% of children with focal seizures arising in the left hemisphere, and no (0%) child with left temporal lobe seizures had his or her seizure-onset laterality incorrectly classified. Several approaches have been used to validate Wada memory testing in adults. Although memory outcome might be considered the ideal variable for validation, Wada memory findings are used to establish surgical candidacy, which confounds the ­predicative

and outcome variables. Although there are reports of successful memory outcomes after Wada memory failure,24 there are also cases of amnesia in which Wada memory results appeared to predict that outcome.25 In addition to memory outcome studies, there have been numerous reports suggesting a relationship between Wada memory scores and hippocampal volume or cell counts.1,​26,​27,​28,​29 Both hippocampal volumes and WMAs are ­related to postoperative verbal memory decline.1,​3,​4,​30,​31,​32,​33 Prediction of postoperative memory change remains among the most important aspects of Wada memory testing. Postoperative risk, whether for cognitive change or efficacy in treating seizures, depends in part on concurrence of preoperative clinical findings. Children with or without WMAs in the direction opposite of that predicted based on clinical semiology are considered to be at higher risk of postoperative memory decline compared with children with WMAs in the predicted direction. In a retrospective review of 132 children, who received resective epilepsy surgery, approximately 70% had WMAs corresponding to the side of surgery. Children without WMAs corresponding to seizure-onset laterality demonstrated significant postoperative verbal memory decline, whereas children with appropriate WMAs showed significantly improved verbal memory scores after surgery. When examined on the individual patient level, 77% of children with WMAs in predicted direction showed no verbal memory decline after surgery, whereas 80% of children without correct WMAs had lower postoperative verbal memory for story recall tasks. WMAs had no value in predicting postoperative changes in visual–spatial memory. Our series found greater sensitivity of story memory for assessing verbal decline in children, whereas in adults, the decline in verbal memory is best captured with word list tasks.3,​31 Whether these findings for story memory result from a greater number of nontemporal cases is unknown. However, because the neural and cognitive systems are in the process of development in children, the brain memory test associations established for adults may be less applicable to children. It has been suggested that children recruit more neural tissue to perform certain linguistic tasks on functional imaging, for example, than do adults. This difference may be reflected in a different pattern of memory test failure after focal cortical resection in children.

„„ Conclusion Wada testing remains a valuable tool to establish language representation and memory function in children. Although the procedure is often technically more difficult to perform in children than adults (e.g., need for anesthesia during angiography), and the results are more difficult to interpret (e.g., mutism, behavior problems, wider range of skill levels), Wada testing has clearly been validated for use in the preoperative evaluation of pediatric epilepsy surgery candidates. The likelihood of obtaining satisfactory results depends on many factors such as the maturity of the individual child. Factors that have been associated with the likelihood of obtaining valid Wada results include age (Wada testing of children younger than 10 years of age is much less likely to provide useful results) and general cognitive function (children with IQs less than 70–80 are less likely to be good candidates). Although Wada memory testing is less likely to yield helpful lateralizing information compared with adult studies, WMAs

253

254

III  Surgical Anatomy and Mapping Techniques are able to indicate which patients may be at higher risk for postoperative memory decline. Whether Wada testing provides incremental prediction of memory decline “above and beyond,” other readily available clinical data (e.g., laterality of seizure

focus, age at onset, mesial temporal sclerosis in ­ magnetic resonance imaging, and preoperative neuropsychological ­ assessment, especially memory and language test scores) have not been proven yet.

References 1. Cohen-Gadol AA, Westerveld M, Alvarez-Carilles J, Spencer DD. Intracarotid Amytal memory test and hippocampal magnetic resonance imaging volumetry: validity of the Wada test as an indicator of hippocampal integrity among candidates for epilepsy surgery. J Neurosurg 2004;101(6):926–931 2. Loring DW, Meador KJ. Wada and fMRI testing. In: Fisch B, ed. Principles and Practices of Electrophysiological and Video ­Monitoring in Epilepsy and Intensive Care. New York, NY: Demos Medical Publishing; 2008 3. Sabsevitz DS, Swanson SJ, Morris GL, Mueller WM, Seidenberg M. Memory outcome after left anterior temporal lobectomy in patients with expected and reversed Wada memory asymmetry scores. Epilepsia 2001;42(11):1408–1415 4. Stroup E, Langfitt J, Berg M, McDermott M, Pilcher W, Como P. Predicting verbal memory decline following anterior temporal lobectomy (ATL). Neurology 2003;60(8):1266–1273 5. Wada JA. Youthful season revisited. Brain Cogn 1997;33(1):7–10 6. Jones-Gotman M, Sziklas V, Djordjevic J, et al. Etomidate speech and memory test (eSAM): a new drug and improved intracarotid procedure. Neurology 2005;65(11):1723–1729 7. Buchtel HA, Passaro EA, Selwa LM, Deveikis J, Gomez-Hassan D. Sodium methohexital (brevital) as an anesthetic in the Wada test. Epilepsia 2002;43(9):1056–1061 8. Takayama M, Miyamoto S, Ikeda A, et al. Intracarotid propofol test for speech and memory dominance in man. Neurology 2004;63(3):510–515 9. Milner B, Branch C, Rasmussen T. Study of short-term memory after intracarotid injection of sodium Amytal. Trans Am Neurol Assoc 1962;87:224–226 10. Masters LT, Perrine K, Devinsky O, Nelson PK. Wada testing in pediatric patients by use of propofol anesthesia. AJNR Am J Neuroradiol 2000;21(7):1302–1305 11. Chitoku S, Otsubo H, Harada Y, et al. Extraoperative ­ cortical stimulation of motor function in children. Pediatr Neurol 2001;24(5):344–350 12. Schevon CA, Carlson C, Zaroff CM, et al. Pediatric language mapping: sensitivity of neurostimulation and Wada testing in epilepsy surgery. Epilepsia 2007;48(3):539–545 13. Szabó CA, Wyllie E. Intracarotid amobarbital testing for language and memory dominance in children. Epilepsy Res 1993;15(3):239–246 14. Hamer HM, Wyllie E, Stanford L, Mascha E, Kotagal P, Wolgamuth B. Risk factors for unsuccessful testing during the intracarotid amobarbital procedure in preadolescent children. Epilepsia 2000;41(5):554–563 15. Loring DW, Meador KJ, Lee GP, Smith JR. Structural versus functional prediction of memory change following anterior temporal lobectomy. Epilepsy Behav 2004;5(2):264–268 16. Chelune GJ. Hippocampal adequacy versus functional reserve: predicting memory functions following temporal lobectomy. Arch Clin Neuropsychol 1995;10(5):413–432 17. Lee GP, Park YD, Hempel A, Westerveld M, Loring DW. Prediction of seizure-onset laterality by using Wada memory

asymmetries in pediatric epilepsy surgery candidates. Epilepsia 2002a;43(9):1049–1055 18. Loring DW, Hermann BP, Perrine K, Plenger PM, Lee GP, Meador KJ. Effect of Wada memory stimulus type in discriminating lateralized temporal lobe impairment. Epilepsia 1997;38(2):219–224 19. Loring DW, Meador KJ, Lee GP, et al. Stimulus timing effects on Wada memory testing. Arch Neurol 1994b;51(8):806–810 20. Loring DW, Meador KJ, Lee GP. Amobarbital dose effects on Wada memory testing. J Epilepsy 1992;5(3):171–174 21. Kirsch HE, Walker JA, Winstanley FS, et al. Limitations of Wada memory asymmetry as a predictor of outcomes after temporal lobectomy. Neurology 2005;65(5):676–680 22. Meador KJ, Loring DW. The Wada test for language and memory lateralization. Neurology 2005;65(5):659 23. Lee GP, Park YD, Westerveld M, Hempel A, Loring DW. Effect of Wada methodology in predicting lateralized memory impairment in pediatric epilepsy surgery candidates. Epilepsy Behav 2002b;3(5):439–447 24. Loring DW, Lee GP, Meador KJ, et al. The intracarotid amobarbital procedure as a predictor of memory failure following unilateral temporal lobectomy. Neurology 1990;40(4):605–610 25. Loring DW, Hermann BP, Meador KJ, et al. Amnesia after unilateral temporal lobectomy: a case report. Epilepsia 1994a;35(4):757–763 26. Baxendale SA, Van Paesschen W, Thompson PJ, Duncan JS, Shorvon SD, Connelly A. The relation between quantitative MRI measures of hippocampal structure and the intracarotid amobarbital test. Epilepsia 1997;38(9):998–1007 27. Davies KG, Hermann BP, Foley KT. Relation between intracarotid amobarbital memory asymmetry scores and hippocampal sclerosis in patients undergoing anterior temporal lobe resections. Epilepsia 1996;37(6):522–525 28. Loring DW, Murro AM, Meador KJ, et al. Wada memory testing and hippocampal volume measurements in the evaluation for temporal lobectomy. Neurology 1993;43(9):1789–1793 29. Sass KJ, Lencz T, Westerveld M, Novelly RA, Spencer DD, Kim JH. The neural substrate of memory impairment demonstrated by the intracarotid amobarbital procedure. Arch Neurol 1991;48(1):48–52 30. Lee GP, Westerveld M, Blackburn LB, Park YD, Loring DW. Prediction of verbal memory decline after epilepsy surgery in children: effectiveness of Wada memory asymmetries. Epilepsia 2005;46(1):97–103 31. Loring DW, Meador KJ, Lee GP, et al. Wada memory asymmetries predict verbal memory decline after anterior temporal lobectomy. Neurology 1995;45(7):1329–1333 32. Perrine K, Westerveld M, Sass KJ, et al. Wada memory disparities predict seizure laterality and postoperative seizure control. Epilepsia 1995;36(9):851–856 33. Sperling MR, Saykin AJ, Glosser G, et al. Predictors of outcome after anterior temporal lobectomy: the intracarotid amobarbital test. Neurology 1994;44(12):2325–2330

27

  Language Lateralization and Localization: Functional Magnetic Resonance Imaging Cristina Go and Elizabeth Donner

Summary Accurate localization of cortical language centers is crucial for children undergoing epilepsy surgery to avoid postoperative language deficits. Functional magnetic resonance imaging (fMRI), a noninvasive imaging test utilizing blood oxygen level dependent (BOLD) signal from regional neuronal activity during performance of specific tasks, provides good spatial resolution to identify areas involved in expressive and receptive language tasks. While it may be challenging in young children, this technique is now the primary tool for determination of language lateralization in many pediatric epilepsy centers, with invasive language mapping reserved for cases in which fMRI is inconclusive. With the advent of new techniques, including resting state fMRI (rs-fMRI), fMRI is likely to become even more valuable in our armamentarium of brain mapping techniques. Keywords:  functional neuroimaging, fMRI, language

„„ Introduction Epilepsy surgery can provide seizure freedom for carefully selected children with drug-resistant epilepsy (DRE). Accurate lateralization and localization of cortical language centers is crucial for children undergoing epilepsy surgery to avoid postoperative language deficits. fMRI is a noninvasive imaging test utilizing BOLD signal from regional neuronal activity during performance of specific tasks and can provide good localization and spatial resolution to identify areas involved in expressive and receptive language tasks. The 2017 Practice Guideline of the American Academy of Neurology regarding the use of fMRI in the presurgical evaluation of patients with epilepsy concluded that fMRI may be considered an alternative option to the intracarotid amobarbital procedure, also known as the Wada test (see Chapter 26), for lateralizing language functions in patients with medial temporal lobe epilepsy, temporal lobe epilepsy in general, or extratemporal epilepsy.1 fMRI can also aid in predicting postsurgical language deficits after anterior temporal lobe resection.2,​3 Basic principles of fMRI are reviewed in Chapter 20. This chapter will specifically focus on the application of fMRI in the presurgical lateralization and localization of language function in children with DRE, the challenges and limitations of fMRI in children due to age and compliance, and emerging fMRI applications that may help overcome these.

„„ fMRI in Children: Challenges and Limitations Direct electrocortical stimulation (ECS) and Wada testing have long been considered the gold standards for identifying essential language areas and pathways through disruption of language functions during direct brain stimulation of intracranial electrodes with ECS, or suppression of hemispheric brain activity through intracarotid injection of anesthetic agents such as sodium amytal or etomidate during Wada testing. Both ECS and Wada tests, however, can be associated with significant risks due to their invasive nature. ECS is also limited by the fact that only areas covered by implanted electrodes can be tested, and ECS carries an inherent risk of stimulation-induced seizures. fMRI is increasingly being used for presurgical language lateralization in children with DRE, because it is a noninvasive procedure that is not limited to a specific brain region and can be repeated over time as a follow-up tool to assess postoperative deficits and potential shifts or reorganization of language functions to contralateral hemisphere.4 fMRI using active language tasks can be done in children as young as 5 to 6 years old, although even with adequate preparation, practice, and orientation in typically developed children, the failure rate can still be high especially at younger ages.5 Not surprisingly, it has been demonstrated that children with epilepsy, and also children with the common comorbidities of epilepsy, attention deficit or hyperactivity disorder, and autism spectrum disorders, are less successful with fMRI than typically developed children. While up to 80% of children with epilepsy will successfully complete a single fMRI scan, the success rate of a complete battery of fMRI scans required to lateralize language function varies from about 30 to 60%.6 Reasons for failed scans include excessive head motion, refusal to complete the task or even to enter the MRI scanner, and inattention resulting in not completing the task and falling asleep. These issues can be compounded by anxiety, claustrophobia, restlessness, fatigue, and difficulty in understanding instructions, in addition to the potential contributions of the epilepsy itself, side effects of anticonvulsant medications, and the behavioral and cognitive comorbidities of the underlying epilepsy syndrome. To counteract the challenges of fMRI in children, it is necessary to screen subjects before the test is booked. Once it is determined that a child is a good candidate for fMRI, it is important to

256

III  Surgical Anatomy and Mapping Techniques review the procedure with parents and children. Many centers use a mock MRI scanner to allow children to practice the language tasks in an environment mirroring the MRI. Children may also be provided with sample tasks in advance of the fMRI to practice and gain familiarity with the tasks. It should also be noted that novel tasks may produce stronger BOLD activation, and as such, there are task practice-related changes that can affect the magnitude and pattern of activation, which needed to be taken into consideration.7

„„ fMRI Language Task Selection The language tasks used during fMRI should be tailored to the child’s developmental and educational level. Tasks that are too easy or too difficult can result in poorer or lower BOLD response.8,​9 To ensure adequate assessment of the full extent of a child’s language network, including frontal and temporal language areas, most centers use a combination of visual and

auditory tasks to obtain activations. A panel of tasks is more likely to demonstrate the full extent of language activation.10 A mixture of expressive language tasks, such as word generation tasks, and receptive language tasks, such as listening or decision tasks will help to ensure that both frontal and temporal language centers are assessed (Fig. 27.1). Language tasks in a block design are usually easier for children to follow than tasks presented in event-related paradigms. In a block design paradigm, the child is asked to perform the language task for 20 to 30 seconds, alternated with a control task for a total duration of 5 or 6 minutes. The selection of a control task is important; ideally, the control task will vary from the language task only in the function of interest. For example, if the language task is a word reading task, the control task may be to look at a “word-like” string of symbols, thereby controlling for the visual activations associated with looking at the word. Likewise, if the language task is to listen to a sentence, the control task may be to listen to a jumbled nonsensical “sentence,” controlling for the auditory stimulation

Fig. 27.1  fMRI in two children demonstrates the need for a comprehensive panel of fMRI language tasks to obtain complete language mapping. In (a), the letter fluency task demonstrates bilateral frontal activations, the verb generation task demonstrates left frontal activations, and the auditory decision task demonstrates left frontal and temporal activations. In (b), letter fluency and verb generation demonstrate bilateral frontal activations, and the addition of the auditory decision task demonstrated left temporal activations.

27  Language Lateralization and Localization: Functional Magnetic Resonance Imaging of listening to the sentence. Although it is less scientifically rigorous, we have also found some benefit to using a simple motor task (repeated finger-thumb tapping) alternated with a language task in a block design. The advantage of this is the hand motion of the motor blocks is easily visualized from outside the scanner to ensure the child is following the task. It also allows us to obtain motor mapping at the same time. We pair this motor task with letter fluency and verb generation tasks and have found them to provide excellent frontal lobe language region activations. To customize each fMRI session, we maintain a flexible battery of tasks to allow fMRI tasks to be tailored according to the child’s cognitive abilities and level of cooperation. For most of our patients, we utilize the following covert tasks with task duration between 5 and 6 minutes: covert visual letter fluency, covert visual verb generation paired with an overt motor control task and covert auditory description decision task, and auditory category decision task.11 For children who are unable to read, auditory verb generation and object picture naming tasks are used. Visual stimuli are projected through MR-compatible goggles and presented in the center of the visual field, whereas digitalized auditory stimuli are presented via pneumatic earphones. For covert tasks, response monitoring can be done using button presses or by instructing patients to tap their upper leg with a single hand to indicate a choice between different stimuli. Other centers have used overt tasks in presurgical assessments for language, with the advantage of producing higher-­ quality scans and in-scanner performance monitoring, which may improve the yield of fMRI.12

Fig. 27.2  Mixed language dominance in a child with left temporal lobe epilepsy; fMRI demonstrates left frontal and right temporal language activations.

„„ Language Laterality by fMRI Determination of language hemisphere dominance by fMRI can be done by visual inspection or by a calculated laterality index. Visual inspection relies on qualitative clinical visual interpretation of fMRI maps by an experienced neurologist, neuroradiologist, or fMRI scientist and is widely accepted as an appropriate method of determining laterality by fMRI.10 A laterality index (LI) is a more quantitative approach in which the volume of fMRI activation in each hemisphere is calculated, either for the whole brain and for language-based region of interest (ROI). The LI is then calculated to determine which hemisphere has the greater activations and thereby concluded to be the hemisphere of language dominance. Lateralization of language function is dependent on handedness, family history of handedness, and structural and functional brain pathology. It has also been shown that laterality of language function changes with age, with increasing left-sided dominance as children age, reaching a peak around 20 years of age, and decreasing left-sided dominance in the later adult years.13 Atypical language patterns can be found in up to 25% of people with epilepsy.14 Atypical patterns can take three forms: right-sided language dominance; left-sided dominance, with altered localization of language activations; and mixed dominance, with frontal language dominance in one hemisphere and temporal language dominance in the opposite hemisphere.15 Examples are shown in Fig. 27.2 and Fig. 27.3. Atypical patterns are more likely to be found in left-handed children, and those with left hemisphere

Fig. 27.3  Right-sided frontal lobe language activation in a child with seizures arising from a left frontal lobe tumor.

257

258

III  Surgical Anatomy and Mapping Techniques epilepsy or early acquired brain lesions.16,​17 Children with epilepsy and atypical language patterns have been shown to have more widespread neural network disruption suggesting that atypical language patterns may be a marker for more extensive brain dysfunction.18

„„ New Applications of fMRI in Language Mapping in Children Successful fMRI is dependent on a cooperative subject, able to complete language tasks during scanning. Obviously, the challenges of fMRI are particularly relevant to the pediatric population. In 1995, Biswal et al reported that activations of the sensorimotor area could be seen when subjects were at rest in the MR scanner.19 Termed rs-fMRI, this technique evaluates low-frequency fluctuations in BOLD signal during times when the subject is not performing a task, or at rest. Coherent fluctuations in BOLD signal during rest have been used to identify language, visual and auditory networks. It is hypothesized that ongoing cognitive processes during rest produce

these fluctuations and allow for network identification.20,​21 With these observations, there has been a growing interest in the use of rs-fMRI for presurgical mapping. Some centers have begun to use this technique to identify language networks.20,​21,​22 While more study is required to demonstrate reliability of language mapping using rs-fMRI, it shows excellent promise and will likely improve our ability to perform accurate presurgical mapping in younger and less cooperative children.

„„ Conclusion Over the last two decades, fMRI has gradually become an accepted and reliable tool for presurgical language mapping in children being evaluated for epilepsy surgery. In many centers, including our own, fMRI is now the primary tool for determination of language lateralization. Invasive language mapping techniques are now reserved for cases in which fMRI is inconclusive. With the advent of new techniques, including rs-fMRI, fMRI is likely to become even more valuable in our armamentarium of brain mapping techniques.

References 1. Szaflarski JP, Gloss D, Binder JR, et al. Practice guideline summary: use of fMRI in the presurgical evaluation of patients with epilepsy: Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Neurology 2017;88(4):395–402 2. Bonelli SB, Thompson PJ, Yogarajah M, et al. Imaging language networks before and after anterior temporal lobe resection: results of a longitudinal fMRI study. Epilepsia 2012;53(4):639–650 3. Sabsevitz DS, Swanson SJ, Hammeke TA, et al. Use of preoperative functional neuroimaging to predict language deficits from epilepsy surgery. Neurology 2003;60(11):1788–1792 4. Hertz-Pannier L, Chiron C, Jambaqué I, et al. Late plasticity for language in a child’s non-dominant hemisphere: a pre- and post-surgery fMRI study. Brain 2002;125(Pt 2):361–372 5. Byars AW, Holland SK, Strawsburg RH, et al. Practical aspects of conducting large-scale functional magnetic resonance imaging studies in children. J Child Neurol 2002;17(12):885–890 6. Yerys BE, Jankowski KF, Shook D, et al. The fMRI success rate of children and adolescents: typical development, epilepsy, attention deficit/hyperactivity disorder, and autism spectrum disorders. Hum Brain Mapp 2009;30(10):3426–3435 7. Raichle ME, Fiez JA, Videen TO, et al. Practice-related changes in human brain functional anatomy during nonmotor learning. Cereb Cortex 1994;4(1):8–26 8. Bookheimer SY. Functional MRI applications in clinical epilepsy. Neuroimage 1996;4(3, Pt 3):S139–S146 9. Nagel BJ, Barlett VC, Schweinsburg AD, Tapert SF. Neuropsychological predictors of BOLD response during a spatial working memory task in adolescents: what can performance tell us about fMRI response patterns? J Clin Exp Neuropsychol 2005;27(7):823–839 10. Gaillard WD, Balsamo L, Xu B, et al. fMRI language task panel improves determination of language dominance. Neurology 2004;63(8):1403–1408 11. Rodin D, Bar-Yosef O, Smith ML, Kerr E, Morris D, Donner EJ. Language dominance in children with epilepsy: concordance of

fMRI with intracarotid amytal testing and cortical stimulation. Epilepsy Behav 2013;29(1):7–12 12. Croft LJ, Rankin PM, Liégeois F, et al. To speak, or not to speak? The feasibility of imaging overt speech in children with epilepsy. Epilepsy Res 2013;107(1–2):195–199 13. Szaflarski JP, Holland SK, Schmithorst VJ, Byars AW. fMRI study of language lateralization in children and adults. Hum Brain Mapp 2006;27(3):202–212 14. Berl MM, Zimmaro LA, Khan OI, et al. Characterization of atypical language activation patterns in focal epilepsy. Ann Neurol 2014;75(1):33–42 15. Dijkstra KK, Ferrier CH. Patterns and predictors of atypical language representation in epilepsy. J Neurol Neurosurg Psychiatry 2013;84(4):379–385 16. Anderson DP, Harvey AS, Saling MM, et al. FMRI lateralization of expressive language in children with cerebral lesions. Epilepsia 2006;47(6):998–1008 17. Yuan W, Szaflarski JP, Schmithorst VJ, et al. fMRI shows atypical language lateralization in pediatric epilepsy patients. Epilepsia 2006;47(3):593–600 18. Ibrahim GM, Morgan BR, Doesburg SM, et al. Atypical language laterality is associated with large-scale disruption of network integration in children with intractable focal epilepsy. Cortex 2015;65:83–88 19. Biswal B, Yetkin FZ, Haughton VM, Hyde JS. Functional connectivity in the motor cortex of resting human brain using echo-­ planar MRI. Magn Reson Med 1995;34(4):537–541 20. Lee MH, Miller-Thomas MM, Benzinger TL, et al. Clinical resting-state fMRI in the preoperative setting: are we ready for prime time? Top Magn Reson Imaging 2016;25(1):11–18 21. Tie Y, Rigolo L, Norton IH, et al. Defining language networks from resting-state fMRI for surgical planning—a feasibility study. Hum Brain Mapp 2014;35(3):1018–1030 22. Roland JL, Griffin N, Hacker CD, et al. Resting-state functional magnetic resonance imaging for surgical planning in pediatric patients: a preliminary experience. J Neurosurg Pediatr 2017;20(6):583–590

28

  Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia James L. Stone and Bartosz Grobelny

Summary The goal of epilepsy surgery is the removal or disconnection of seizure foci with maximal preservation of neurological function, and particularly those functions considered eloquent, within the realm of fine motor and related sensory functions. Children, more often than adults, require general anesthesia for such surgery, and the avoidance of motor or tactile sensory ­deficits requires techniques that identify and protect the R ­ olandic ­ cortical sulci. This chapter emphasizes the use of somatosensory-evoked potential (SSEP) and motor evoked potential (MEP) methods, direct cortical mapping, and ­monitoring techniques including cortical, subcortical, and transcranial electrical ­stimulation (TES). The recent preoperative use of navigated transcranial magnetic motor stimulation is ­presented, along with additional monitoring ideas for future. Keywords:  cerebral cortical surface recorded afterdischarges (ADs), bipolar cortical stimulation—Penfield’s/Ojemann’s technique, corticospinal tract (CST), D-waves—direct CST waves, compound muscle action potentials (CMAPs), direct cortical stimulation (DCS-MEP), direct subcortical stimulation (DSS-MEP), electrocorticography—cortical surface recorded EEG (ECoG), electromyogram (EMG), I waves—indirect CST waves, motor evoked potential (MEP), multi-pulse train/ high-frequency MEP technique (MPT), neuromuscular blocking agents (NMBAs), navigated transcranial magnetic motor stimulation (nTMS), short latency somatosensory-evoked potential (SSEP), SSEP phase reversal (PR) technique, train-of-four (TOF) NMBA decrement test, transcranial electrical stimulation (TES) “The rapidly expanding field of pediatric epilepsy surgery is perhaps the most dramatic recent advance in surgical treatment, offering opportunities to arrest or reverse inevitable developmental delay and lifelong institutionalization in children whose condition previously would have been considered hopeless.” —Jerome Engel Jr.1

„„ Introduction As in adults, the goals of epilepsy surgery in children are to remove or disrupt seizure foci while preserving neurological function to the greatest extent possible. To achieve this goal,

multiple adjuncts are used to try to identify both the surgical target of resection and protect areas of eloquent cortex, defined as a region whose resection or disconnection can result in a permanent neurological deficit within the realm of motor, sensory, or language functioning.1,​2 A critical goal is the preservation of functional eloquent cortex and related subcortical connectivity. Focal surgical resections for seizure control in the area around the central sulcus of Rolando carry a permanent deficit rate of about 28%, which we believe can be lessened by cortical or subcortical mapping and monitoring techniques.3,​4 “Mapping” is the identification of eloquent cortex, while “monitoring” is the continuous or near continuous ongoing evaluation of such cortical functioning. Metabolically based neuroimaging such as functional MRI and radioactive tagged imaging linked to local cerebral cortical blood flow or oxygenation have improved our ability to lateralize and identify cortex “involved” in specific motor, ­ ­sensory, or language/behavioral tasks. However, these modalities cannot tell us if such involved cortex is primarily eloquent or secondary to hyperemic spread.5,​6,​7 Epileptogenic as well as nonepileptogenic lesions also induce functional plasticity resulting in r­ edistribution of motor, or other eloquent functions to less affected regions making identification and protection of potentially eloquent areas difficult.8,​9,​10,​11 As clearly demonstrated in brain tumor surgery, awake craniotomy can minimize permanent eloquent cortex ­deficits, though some older children may be traumatized by the ­experience.12 Additionally, younger children and some adults will always require general anesthesia.13 These are the patients who can benefit from localization and protection of eloquent cortex utilizing SSEP and MEP mapping and monitoring under general anesthesia for epilepsy surgery. As direct intraoperative cortical mapping of eloquent nonmotor functions cannot be examined under general anesthesia and requires an awake, cooperative patient or child of about 12 years of age or greater, such mapping is beyond the scope of this chapter. In this chapter, we discuss the basic physiology and usage of SSEPs and MEPs under general anesthesia to map and protect eloquent cortex during the resection of epileptic foci in children with refractory seizures originating near or within the eloquent central/Rolandic region. We next expand on the usefulness of the cortically recorded SSEP to locate the postcentral gyrus, and anteriorly the central sulcus and precentral gyrus.

260

III  Surgical Anatomy and Mapping Techniques The recording of MEPs in response to direct cortical (DCS) and direct subcortical electrical stimulation (DSS) is presented and discussed. Following this, we review the use and physiology of MEPs elicited by TES and recent preoperative adjunctive use of navigated transcranial magnetic motor stimulation (nTMS). The practice and success of pediatric epilepsy surgery and a need for cortical mapping in children certainly pose additional demands. A better understanding of maturation related to childhood cortical development as well as monitoring modalities of cortical electrical or magnetic excitability borrowed from intraoperative brain and spinal cord monitoring have led to methodologies enabling localization of eloquent motor cortex with decreased delivery of potentially dangerous electrical energy. We discuss various pitfalls in the practical use of these modalities including critical anesthetic concerns, and in closing, present newer ideas and future suggested directions.

„„ Anatomy and Applied Physiology The traditional scalp recorded short (latency) SSEP is easily recorded with commercially available instrumentation and depends on adequate electrical stimulation of peripheral nerves. It should not be confused with longer latency event-related associational cortical responses in alert subjects used in psychiatry or psychology. The SSEP is generated by large-fiber afferent sensory fibers for proprioception and vibration. The selected nerves for the upper extremity are the median or ulnar nerves at the wrist; and for the lower extremity—posterior tibial at the ankle or popliteal fossa, or peroneal nerve at the ankle or knee. The signal enters the spinal cord through the posterior roots, and then travels to the ipsilateral dorsal (posterior) column. Nerve fibers from the cervical and thoracic region terminate in the medullary cuneate nucleus, and fibers from the lower body terminate in the gracilis nucleus of the medulla. These fibers then cross to the contralateral medulla and ascend in the medial lemniscus, which terminates in the somatosensory nuclei of the ventral posterolateral nucleus of the thalamus. The primary somatosensory cortex (postcentral gyrus, Brodmann’s areas 3, 1, and 2) receives input from the thalamus in a somatotopic distribution, the lower extremity closest to the midline, laterally followed by the trunk, upper extremities, fingers, and face.14,​15,​16,​17 The various SSEP waveform peaks are named by their negativity or positivity (N or P) and latencies in milliseconds at the various volume-conducted nerve bundle, synaptic relay, or terminating cortical points (i.e., negative deflection at about 20 ms—N20). For successful extremity SSEP recording or somatosensory cortical mapping, the patient must have normal or only a mild deficit in cortical sensory function regarding proprioception, light touch, and two-point discrimination.4 Interruption or damage anywhere in this pathway, including the vascular territory supporting these structures, if sufficient, will result in both a decrease of the amplitude and latency delay in one or several of the recorded waveforms. The middle cerebral artery, representing the terminal territory of the internal carotid artery, provides blood supply to the cortex representing the upper extremity SSEP cortical peak (N20). The corresponding lower extremity SSEP cortical peak (P37) is usually supplied by terminal branches of the anterior cerebral artery. The vertebral arteries supply the upper cervical cord and medulla, basilar artery largely supplies the pons and midbrain, and per-

forating arteries off proximal portions of the above arteries or communicating branches supply the deep paramedian diencephalon and deep cerebral hemispheres.14,​15,​16,​17 The MEP pathway originates with stimulation at the primary motor cortex, which is located predominately on the precentral gyrus (and less so the adjacent postcentral gyrus). The primary motor cortex is responsible for voluntary gross and fine movements in contralateral muscle groups with more cortical area (larger homunculi) devoted to the face, tongue, hand, and foot in humans. Similar to the somatosensory cortex, the primary motor cortex is somatotopically organized with the face, neck, and tongue motor regions, inferiorly near the sylvian fissure, hand and arm neurons in the middle posterior-directed convexity, proximal leg superiorly, knee near the superior crest of the gyrus, and distal lower extremity and foot inferiorly in the mesial parasagittal region. The face is bilaterally innervated in the fetus, but more so in the upper face after term. The tongue is bilaterally innervated. The human pyramidal or corticospinal tract (CST) largely originates in the precentral gyrus (Brodmann’s areas 4 and 6) and consists of pyramidal cell axons located in layer 5 of the cerebral cortex. Contrary to earlier views, only 2 to 3% of CST axons originate from the giant Betz cells, the remainder arise from smaller pyramidal cells, some unmyelinated. Traditionally, the area 4, cortical layer 5 Betz cells have been believed responsible for the low threshold for motor excitability in response to DCS as well as the discreteness of the muscle responses.18 The large myelinated axons in or just below the primary motor cortex are thought to be the predominately activated fibers, consisting of corticospinal and corticobulbar pathways.19,​20,​21,​22 The corticobulbar tract fibers originating in the motor cortex travel alongside the CST fibers, until they diverge into the brainstem and terminate on interneurons, and a smaller number directly synapse on lower motor neurons to generate cranial nerve muscle movements. There is a wide range of conduction speeds within the CST, but the electrophysiological recordings are likely dominated by the fastest conducting synchronous fibers with largest amplitude.18,​ 23 Only a small percentage of CST fibers directly synapse on anterior horn cells, as most terminate on interneurons in the intermediate zone between the dorsal and ventral horns. That small proportion of CST fibers directly synapsing on alpha motor neurons are primarily involved in fine movements of the distal extremities. At the medullary level, from 70 to 90% of the CST fibers decussate to the contralateral side, while the uncrossed fibers form the small ventral CST, which is variable in size, at times absent, often only extends as far as the cervical or thoracic region, and absent at times.18,​23 The middle and anterior cerebral arteries primarily supply blood to the motor cortex, the lenticulostriate perforators, and anterior choroidal artery supply the internal capsule, and vertebral and basilar artery branches supply the brainstem, all of which may produce distinct changes to MEP responses in the presence of ischemia.19,​24 The anterior spinal artery is primarily responsible for vascular supply to the CST in the spinal cord. The pre- and postcentral gyri are connected in three places within the depth of the central sulcus and considered a continuum of one region into another (pli de passage of Broca). These areas of apparent sensory or motor integration are just above the sylvian fissure, within the motor and sensory hand region, and superomedially near the interhemispheric fissure.25

28  Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia Phylogenetically, the CST is believed to have arisen from the postcentral (sensory) gyrus and in lower animals projects to sensory areas in the dorsal horn of the spinal cord.23 Direct electrical cortical stimulation of this central area in awake patients has proven overlap of sensory and motor function.26,​27,​28,​29,​30 This is in line with cytoarchitectonic evidence that pyramidal cells can be found in the pre- as well as postcentral gyrus.31 About two-thirds of the human CST arise from the precentral region, and one-third from adjacent postcentral areas of the anterior parietal lobe.32 Areas of motor hand activation are present in the anterior face of the corresponding postcentral gyrus and adjacent walls of the central sulcus. Thus, one must be aware that cortical function is not only represented in the visible crowns of the gyri but also in the depth of their walls. The parietal contribution to the human CST is believed to terminate in the dorsal column nuclei (gracilis and cuneatus) and substantia gelatinosa of the dorsal horn in the spinal cord, and likely plays a role in regulation or feedback of the sensory consequences of voluntary movement.23 Most authorities believe that direct electrical stimulation of the cerebral cortex activates upper motor neurons whose cell bodies are found in the lower cellular layers, and which form the CST (pyramidal). This activation likely occurs at the axon hillock just before its initial node of Ranvier myelin segment, near the cortical white matter junction. These larger CST fibers are “directly” activated at lower stimulus currents and have higher conduction velocities than smaller-diameter fibers, and are recorded as “D waves.”33,​34 Other CST neurons are secondarily or “indirectly” activated by cortical interneurons and produce “I waves,” which follow an initial, usually higher amplitude D wave, as recorded from an upper thoracic or lower cervical electrode in the spinal epidural space passed upward from the thoracic region. This can be done through a laminectomy opening if a spinal cord tumor is being operated upon, or passed upward in the epidural space from a percutaneous midline Tuohy needle insertion.35 For direct cortical or subcortical motor mapping to be successful, the patient generally requires antigravity (grade 3 or better) extremity strength. If a more severe hemiparesis is present, such motor mapping is usually not possible. Similarly, in healthy young children below the age of about 4 years, motor cortical and subcortical direct electrical excitability is much reduced. In a patient with a severe motor deficit or in a young child, however, SSEP phase-reversal (SSEP-PR) mapping can often be used to identify the primary sensory cortex, central sulcus, and secondarily the precentral gyrus.4 Increasing the intensity of electrical stimuli will also produce recordable extremity muscle responses if the proper anesthetic regime is utilized and the amount of neuromuscular blocking agents (NMBAs) allows such movement.36 Extremity movement during the surgical procedure may be observed, although recordings of compound muscle action potentials (CMAPs) with electromyogram (EMG) needles in target muscles allow for better quantification of the muscle response.23 There is recent concern that after the administration of NMBA under general anesthesia, different extremity muscles demonstrate variable recovery times, and decrement profiles, and can lead to false-positive MEP monitoring results.37 It is suggested that the train-of-four (TOF) NMBA decrement test, used by anesthesia or the monitoring team, be performed separately upon upper and lower extremities and in particular on the

same (or contralateral to) MEP muscles being ­ monitored.37 Consequently, a TOF from the forehead or facial muscles is not considered reliable for MEP from upper or lower extremity muscles.37 Both the selection of anesthetic agents and depth of anesthesia are critical for the successful recording of SSEPs and MEPs. To optimize the maintenance of plasma concentrations of intravenous agents, it is recommended that a target-­controlled infusion pump be utilized, as well as an electroencephalo­ graphic bispectral index (BIS) or similar device for depth of anesthesia.34

„„ Somatosensory-Evoked Potentials Short-Latency Somatosensory-Evoked Potentials Stimulation and Recording the SomatosensoryEvoked Potential The commercially available evoked potential units have multiple stimulus and recording channels to simultaneously monitor a number of evoked potential modalities concurrently with the SSEP during neurosurgical operations on the head and spine. It is recommended when recording the SSEP from both sides to interleave stimulation and recording from each limb individually. Disposable conductive solid gel surface or disposable subdermal needle electrodes are used to deliver cathode-generated rectangular pulses of constant current to stimulate peripheral nerves that generate the SSEP responses. If more proximal stimulation sites are used among shorter patients or children, the normative SSEP latencies are significantly shortened. Conversely, in tall- or long-limbed individuals, normative latencies are lengthened. By the late teens, adult normative values are present.14,​15,​16,​17

 timulus parameters: Pulse width: 200 to 300 μs (0.2–0.3 ms); S frequency: 1.5 to 5 Hz. Intensity: Supramaximal for distal extremity twitch; less than 60 mA for surface electrodes and less than 40 mA for subdermal needle electrodes (about 5–15 mA variances between stimulator used).

Supramaximal stimulation can be done in the presence of muscle relaxants and is recommended to minimize response variation. This can be verified by increasing the stimulus intensity to the point where further increments do not appreciably increase the amplitude of the recorded response. The SSEP stimulation frequency should not be a factor of 60 cycle (Hz), which is the standard or main current frequency in United States (50 Hz in Europe). This is a frequently encountered environmental noise in the operating room. Selecting a frequency that is not a factor of the 60-Hz frequency allows for computerized signal averaging (linking the time frame after stimulus to a specific finite analysis duration time or window for recording the response). This effectively removes out-of-phase noise from the averaged signal and is the basis of computerized signal averaging to summate these microvolt recorded responses to stimulation. Electrode impedances are ideally below 5 kOhms. Electrical artifact and interference from other equipment or electrical lines in or near the operating room make signal

261

262

III  Surgical Anatomy and Mapping Techniques averaging necessary to obtain an appropriate signal-to-noise ratio. Several hundred to one thousand responses are averaged before analyzing an SSEP waveform. Analysis time (window) is 100 ms. Cranial scalp recorded SSEP electrode positions (as well as MEP scalp stimulation electrode positions) are located according to the 10–20 International system of EEG electrode placement. Recording from various locations along the SSEP pathway is essential to give information regarding the level and integrity of the pathway. It is also recommended to record individual responses from bilateral upper extremity and lower extremity nerve stimulations. Bilateral upper and lower extremity nerve recording also allows for determination of global systemic (i.e., anoxia) or environmental (i.e., hypothermia) versus focal local signal changes. Normative SSEP latencies must be adjusted downward for children by age and in adolescents by height. Examples for adult scalp recorded SSEP upper extremity wrist stimulation are N13, a subcortical response, and N19/20, the postcentral gyrus cortical response. For lower extremity SSEP, the adult cortical postcentral response is the P37. Differing literature sources may label these peaks at slightly different latencies: for example, N19 is sometimes referred to as N20. However, these are both referring to the positive deflection in the cortical (postcentral gyrus) potential from median nerve stimulation. By knowing which waveform is being referred to, it becomes intuitive which waveform feature is being discussed.14,​15,​16,​17 SSEP recording utilizes high-gain amplifiers, and bandpass filtering. A low-cut filter between 10 and 30 Hz and a high-cut filter between 300 and 500 Hz is recommended for the cortical responses, while the recommendations for subcortical and peripheral nerve responses is between 10 and 30 and 1,000 and 3,000 Hz. False positives can arise from technical problems such as electrodes becoming displaced or equipment malfunction in the acquisition system and software. It is important for the neurophysiology monitoring team to be properly trained with the commercial system, as well as the availability of technical expertise with extensive computer troubleshooting skills to remedy issues which may arise. Utilizing checks such as verifying stimulus artifacts and measuring electrode impedance values can aid in identifying confounding technical issues.

SSEP Upper Extremity Scalp Recorded Waveforms: Adult (Adjusted Pediatric) Latency Values The ascending upper SSEP is first recorded at Erb’s point. This potential reflects brachial plexus activity and this negative peak, known as N9 (N7), is approximately 9 ms after median or ulnar nerve wrist stimulation in most adults. A subdermal needle or solid gel electrode can be placed approximately 1 to 2 cm superior to the midpoint of the clavicle. The Erb’s point potential is recorded referentially, with the reference electrode being the contralateral Erb’s point. Recall that bilateral stimuli are interleaved, meaning the stimulation rate is the same on each side, but the stimulation timing is offset so that stimuli are not delivered bilaterally at the same time. This is why one Erb’s point can be referentially referred to the other. The subcortical C5 spinal cord entry potential is known as N13 (P/N 8–9) (noncephalic reference such as the shoulder). This SSEP potential is recorded from the posterior neck spinous process of C5. Multiple generators are believed responsible for

this waveform and include the root entry zone, dorsal horns, or dorsal columns. A small potential P14/N18 is often recorded in the central region contralateral to the site of median or ulnar nerve stimulation. The believed sources are the cuneate nucleus and medial lemniscus. The major upper extremity cortical potential generated by the posterior central gyrus is called N19/20 (N16).15,​38,​39 The active electrode is contralateral to the median or ulnar nerve stimulus and placed at C3′ or C4′ (′ indicates “prime”), which in adults is 2 cm (less in children) posterior to C3 or C4, respectively, and referred to the other C3′ or C4′—ipsilateral to the stimulus (Fz or earlobe may also be used). This potential is believed to be generated by electrical volleys in the thalamocortical fibers, which synapse in the primary somatosensory parietal cortex, and possibly adjacent parietal associational cortex. N19–N20 is considered a near-field generated potential as its amplitude drops when recorded a short distance from the respective C3′ or C4′ (Fig. 28.1).15

SSEP Lower Extremity Scalp Recorded Waveforms: Adult (Adjusted Pediatric) Latency Values The popliteal fossa potential is the first ascending potential following posterior tibial nerve stimuli, in the lower extremity SSEP and is analogous to Erb’s point. The active electrode is in the posterior crease of the knee, with the reference electrode several centimeters superior. A recorded negative sensory nerve action potential comes at about 8 ms N8 (N5). An electrode over the T12–L1 spine can record the ascending volley as N22. A small lower extremity subcortical potential may be recorded at the C5 or C2 spinous process of the neck and referred to the midline frontal vertex scalp (Fz) or a noncephalic reference point and labeled N30 (N20). This is believed to reflect the subthalamic brainstem medial lemniscus with possible component of gracilis nucleus. The mesial sensory cortical potential P37 (P28) is recorded from Cz’ to Fz, or C’ipsi to C′contra (C3′–C4′ and C4′–C3′).15,​38,​39 These prime (′) positions are approximately 2 cm behind C3 and C4 in the adult (less in children). This potential is believed to be generated by electrical volleys in the thalamocortical fibers, which synapse in the primary somatosensory parietal cortex and possibly adjacent parietal associational cortex. It is believed that deeper within the medial aspect of the postcentral gyrus where the sensory foot area is located, this is a negativity (similar to the postcentral gyrus hand region), but the dipole projecting superiorly to the scalp electrode registers a positivity. P37 (P28) is thus considered a near-field potential in that its amplitude drops when recorded a short distance from Cz’ (Fig. 28.2).15

Effects of Maturation upon the SSEP Postnatal development of the somatosensory system is complex because of nonparallel changes in maturation parameters over a lengthy pathway, varying rates of peripheral and central myelination of portions of the pathway, and increasing number of synapses in the pathway. These developmental sequences are relatively simple compared with the maturation of brain, which includes the process of synaptogenesis as well as lengthening and myelination of the complex polysynaptic pathways of the

28  Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia

Fig. 28.1  Normal scalp recorded median nerve– generated somatosensory-evoked potential. N9 brachial plexus (Erb’s point) recorded over the ipsilateral clavicle. N13: Believed source is ipsilateral C5 spinal cord entry, recorded over the C5 cervical spinous process. P14/ N18: Believed sources are cuneate nucleus/ medial lemniscus, recorded at the contralateral scalp C3′ or C4′. N20: Primary upper extremity somatosensory cortical response, recorded at the contralateral scalp C3′ or C4′. EP, Erb’s point; Epi, ipsilateral; EPc, contralateral; Cc, contralateral somatosensory cortex; Fz, midline frontal vertex. (Reproduced with permission from Nuwer and Packwood 2008.15)

Fig. 28.2  Normal scalp recorded posterior tibial nerve–generated somatosensory-evoked potential. N8: Posterior tibial nerve, recorded over the popliteal fossa. N22: Lumbar spinal cord, recorded over the T12/L1 spinous processes. N30: Believed sources are gracilis nucleus/medial lemniscus, recorded over the C5 or upper cervical spinous processes. P37: Primary lower extremity somatosensory cortex, recorded over the midline vertex Cz’. PF-K, popliteal fossa to knee; lc, lumbar spinal cord; Ci, ipsilateral somatosensory cortex; Cc, contralateral somatosensory cortex; C’z, midline central, 2 cm posterior to Cz (in adults), vertex; Fz, midline frontal vertex. (Reproduced with permission from Nuwer and Packwood 2008.15)

263

264

III  Surgical Anatomy and Mapping Techniques thalamocortical system.38,​39 Nevertheless, the somatosensory system develops very early and SSEPs have been recorded in preterm infants between 27 and 32 weeks estimated gestational age, with latencies decreasing rapidly over the preterm period.40 After birth, by 1 to 3 months of age latency to the major SSEP cortical peak (N19–20 ms in adults) is slow due to immaturity. However, by age 1 year to adolescence, the latency to the SSEP N19–20 is shortened compared to that of adults, due to their shortened stature and limb length.38,​39,​40 This is in striking contrast to the delay in maturation of the corticospinal or pyramidal motor system.

Anesthetic Considerations Anesthetic agents can have varying effects on recorded SSEP responses depending on the combination of agents used.17,​41,​42,​43 Typically for SSEP recordings, inhalational agents should be largely avoided or minimized. Often the earliest sign of increasing depth of anesthetic is blunting or attenuation of all of the SSEP cortical peaks, with relative preservation of the subcortical peaks.15,​44 At increasing dosage levels, benzodiazepines, barbiturates, and propofol show progressive reduction of cortical amplitudes and increasing latencies. Opioids show a mild increase in cortical latencies.44,​45 Etomidate (at low doses) and ketamine increase cortical amplitudes.44,​46,​47,​48 Dexmedetomidine has no significant effect on the SSEP.49 It is crucial for the neurophysiology team to communicate with the anesthesia team in order to optimize the anesthetic plan for intraoperative neurophysiological recording.16,​17,​44,​50,​51,​52,​53 Muscle relaxants may be used in the presence of SSEP recordings, and may enhance the signal clarity by improving the signal-to-noise ratio. However, if MEPs are simultaneously recorded, use of muscle relaxants significantly decrease and may eliminate muscle responses (see below).

SSEP Interpretation and Application during Resection in Proximity to Eloquent Cortex The strategy differs if the precentral and postcentral gyri are exposed at surgery. It is always advantageous to obtain bilateral peripheral, subcortical, and cortical SSEP responses during intracranial surgery as opposed to cortical responses alone. When identifying a change in SSEP responses, it is important to determine if localized changes proximal to the operative site are present, or if observed changes are a result of an anesthetic, technical, or environmental temperature issue. This can be done by analyzing the signal change level (peripheral, subcortical, or cortical), and whether the change occurred unilaterally or bilaterally. Systemic signal changes, limb malpositioning, or large cerebral infarcts can potentially be identified when bilateral peripheral, subcortical, and cortical recording is performed. The generally accepted criterion for significant intraoperative SSEP changes during intracranial surgery is the 50/10 rule: an amplitude reduction equal or greater the 50 or 10% latency increase for a major waveform compared to its baseline must be reported immediately to the surgical team.14,​15,​16,​17,​54,​55 Systemic factors that can lead to signal changes are: temperature, hypotension, or hypoxia. Hypotension and hypoxia are associated with amplitude decease or loss. Decreases in temperature cause decreased nerve conduction velocities and thus increased SSEP latencies in affected limbs. Cold intravenous (IV)

fluid injection may increase the latency in one limb. Preexisting neurological deficits that are detectable with SSEP responses can be amplified such that the response in the affected limb is more sensitive to minor degrees of hypotension or cooling compared to unaffected limbs.14,​15,​16,​17 Limb malpositioning is highly suspected, by comparing postpositioning responses to baselines and identifying signal deterioration at a peripheral site and ascending recording sites. In this regard, SSEPs are very valuable any time prolonged surgical positioning will put the patient at risk for a postoperative nerve compression or stretch-related injury. SSEP recording is of particular value when neurovascular structures are at risk, as there is a linear correlation between cortical SSEP amplitudes and cerebral blood flow (CBF) when decreased below 15 mL per 100 g of brain parenchyma per minute. Similarly, cortical SSEP amplitude loss correlates with middle cerebral artery and carotid artery infarcts; however, cortical SSEP, like the EEG is relatively insensitive to subcortical ischemia. However, routine EEG is more sensitive to lowered CBF than SSEP, showing a loss of 50% amplitude or delta slow waves with CBF values in the mid to low 20s.14,​17,​24,​56,​57

Practical Limitations In order to identify changes in the SSEP during surgical identification of the primary sensory cortex, central sulcus, and by anatomical inference, the primary motor cortex, the SSEP recording must have a reliable and robust baseline. This requires normal, or only a mild neurological deficit in cortical sensory function as regard proprioception, two-point discrimination, and light touch. SSEP responses are averaged signals, so there is a limit as to how rapidly responses can be obtained and analyzed. Depending on the frequency of stimulation and number of averages being used, signals can generally be obtained between 1 and 5 minutes. Highly trained personnel are essential in optimizing stimulation and recording parameters such that an adequate signal-to-noise ratio is achieved with consistently adequate waveforms for comparison and interpretation. Effective and frequent communication between the surgical and neurophysiology team can significantly reduce this delay. The neurophysiology team must have a clear understanding of the critical structures in the nearby vicinity of the surgical field or dissection and correlate particular vigilance to the pertinent monitoring modalities. During a subcortical resection and after identification of the postcentral gyrus, central sulcus, and precentral gyrus, continuous SSEP monitoring can warn of damage to the ascending somatosensory pathway. The surgical team should announce when a critical manipulation or resection is about to occur, and the monitoring team will then immediately begin a new series of averaging responses. This is important such that a waveform collected before the critical maneuver will not be averaged into the interpreted response or sequence, which could potentially mask (minimize or nullify) any important changes secondary to the critical maneuver. The sensitivity of SSEP monitoring varies on the particular type of surgical condition and surgical procedure, but in most series, detection ranges about 80%. When only severe postoperative deficits were considered, the sensitivity and negative predictive values were 81 and 98%, respectively.58 In general, when using the 50% amplitude diminution or 10% latency increase

28  Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia rule regarding a major SSEP waveform, false-positive results occur in a small proportion of cases, and false-negative cases are rare.15 In a large study of over 51,000 spinal procedures, sensitivity was 92%, specificity was 98.9%, and negative predictive value was 99.93%.15 Careful attention to possible confounding intraoperative systemic and anesthetic events almost always will account for potential false-positive results.

Localization of the Primary Somatosensory and Motor Cortices with the SSEP Phase Reversal Technique In addition to the use of SSEP responses for monitoring functional integrity of the primary somatosensory cortex (postcentral gyrus) and subcortical ascending pathway, the

SSEP-PR N20–P20 technique can be used intraoperatively for mapping the location of the central sulcus (sulcus centralis) of Rolando as the anterior border of the postcentral gyrus (Fig. 28.3 and Fig. 28.4).59 The SSEP-PR technique was introduced nearly 50 years ago as an aid to pediatric epilepsy surgery by Goldring,60,​61,​62 and has since been frequently described to benefit the safe resection of brain lesions.59,​60,​61,​63,​64,​65,​66,​67,​68,​69,​ 70 By identification of the postcentral gyrus and the central sulcus anteriorly, the primary motor cortex or precentral cortex is identified and protected just anterior to the central sulcus. Under general anesthesia, a central convexity craniotomy is performed, and the dura opened widely. After preliminary identification of the postcentral gyrus, central sulcus, and precentral motor gyrus, SSEP monitoring with a strip electrode may be applied during a subcortical resection to warn of damage to the ascending somatosensory pathway. Not infrequently, the

Fig. 28.3  Recording setup for the direct, cortically recorded median nerve somatosensory evoked potential phase reversal (SSEP-PR) technique. Note the cortical electrode grid is placed at an angle of approximately 15 degrees across the central sulcus (also called sulcus centralis, SC), straddling the hand regions of the presumed postcentral (electrode #3) gyrus, and precentral (electrode #4) gyrus. The central sulcus (SC) is located between the electrode contacts #3 and #4. (Reproduced with permission from Kombos 2008.59)

Fig. 28.4  Somatosensory evoked potential phase reversal (SSEP-PR) technique: Either bipolar, referential (monopolar), or both forms of recordings may be used to demonstrate a phase reversal (mirror image change) of the N20/P25 waveform polarity as recorded from the postcentral gyrus (electrode number 3); N20 is directed upward, and P25 is directed downward. The mirror image opposite is recorded from the precentral gyrus (electrode number 4); N20*is directed downward, and P25* is directed upward. This clear change denotes the intervening central sulcus (SC) between the pre- and postcentral gyri. Such a response may be unobtainable or severely distorted if the regional cortical anatomy is pathological as in the proximity of a structural lesion (i.e., tumor, dysplasia) or the presence of a severe sensorimotor deficit. (Reproduced with permission from Kombos 2008.59)

265

266

III  Surgical Anatomy and Mapping Techniques central sulcus location can be difficult to appreciate as thickened arachnoid, surface veins, and arteries may occupy the sulcus and obscure its clear definition. No attempt should be made to dissect or move these structures from their position, and they uncommonly interfere with the electrical recordings. Contralateral median nerve stimulation is performed with similar settings compared to SSEP for scalp recordings. The cortical SSEP has an increased amplitude compared to the scalp recorded potential, due to a more favorable signal-to-noise ratio.67 Both scalp and direct cortical surface recordings after standard contralateral median nerve SSEP stimulation disclose a recorded positivity at about 22 ms (P22, adult value) from the hand region of the precentral gyrus. This appears approximately 2 to 3 ms after the postcentral gyrus hand region negativity at N19–N20 (adult ­value).60,​61,​62,​63 This reversal of phase polarity is the basis of the SSEP “phase reversal (PR)” as recorded from the pre- and ­postcentral cortices, and secondarily marks the identity of the intervening central sulcus. Typically, grid or strip electrodes are placed on the cortical surface across what appears to be the central sulcus at the ­nticipated location of the primary hand motor and sensory cortex from 3 to 4 cm anterior to the central sulcus to a similar distance p ­ osterior to the central sulcus (~ 4–6 cm above the sylvian fissure, and between 3 and 8 cm from the midline). As a guiding point, the central sulcus is often sinusoidal consisting of three bends, ­usually the most prominent is the middle bend with convexity facing posteriorly. This feature distinguishes the motor hand area or “hand knob” of the corresponding precentral gyrus.71 Although the central sulcus hand knob region may be obscured by overlying arachnoid, surface bridging veins, and arteries, “the ‘precentral knob’ can be identified intraoperatively with ease as the structure opposite to the intersection of the superior frontal sulcus with the precentral sulcus”.30 If a mass lesion is present, the grid or strip electrode should be placed adjacent to the visible margins of the lesion and not fully over a presumed lesion. The placement of the grid or strip should then be rotated or displaced to maintain the area of maximal peak SSEP cortical amplitude with the lowest intensity of stimulation by moving the electrode array on the brain surface.17,​59,​69,​72 Optimal SSEP amplitude is often achieved by rotating the electrode to make a slightly superior slanted angle of roughly 15 degrees in the sagittal plane, the presumed precentral gyrus electrode placement slightly superior to the postcentral gyrus.59,​73 Communication with the neurophysiology team is essential, and patience is required once the central sulcus has been identified. The phase reversal between P20 and P22 (adult value), representing the precentral gyrus, and N19–N20 (adult value), representing the postcentral gyrus, is due to varying dipole sources in the postcentral and precentral gyri secondary to the median nerve (hand distribution stimulus) at the contralateral wrist. It is recommended to adjust the position of the electrode grid or strip electrode such that the localized central sulcus is situated between several different electrode positions, to verify the consistency of the hand area. Posterior tibial or popliteal SSEP-PR can also be performed, but the primary sensory cortical representation of the lower extremity is limited to a much smaller area in the superior medial aspect of the postcentral gyrus.74 Although a small recording strip electrode may be slipped into this narrow region, the presence of often engorged bridging veins in this tight region may bleed or obscure the anatomy in this area, resulting in difficulty

verifying correct electrode position and obtaining reliable recordings, especially in younger children.59 Each electrode site within the grid or strip serves as an active recording site, utilizing a common reference. The reference electrode is typically a subdermal needle or solid gel surface electrode placed on the contralateral mastoid or cephalic region, or a needle electrode in the exposed temporalis muscle. Impedance checks should routinely be performed to verify electrode contact. Saline irrigation may improve impedance, but excess irrigation can lead to shunting between electrodes.

Interpretation and Application during Lesion Resection in Proximity to Eloquent Cortex Following median nerve stimulation, the reason for the electropositive phase reversal is based on the principle that a perpendicular electrical dipole generated on the postcentral gyrus relative to the central sulcus results in dipole changes in polarity on the adjacent precentral gyrus.70 The phase reversal technique depends on the neurophysiology team identifying the electrode locations where the reversal of phase is observed. Either bipolar or referential (monopolar) recording of the phase reversal can be utilized. The precentral electrodes exhibit a peak positivity at a slightly increased latency when compared to the peak negativity (N19) recorded off the postcentral electrodes. In Fig. 28.3, a 1 × 6 electrode strip is placed across the central sulcus in the hand sensorimotor region.59 Fig. 28.4 shows examples of cortical median nerve SSEP-PR waveforms.59 PRs are observed between electrode positions 3 and 4 for the electrode strips pictured in Fig. 28.3.59 In situations, where a clear PR cannot be adequately identified, increasing the number of grid or strip electrode contacts, and utilizing a bipolar as opposed to referential montage, where adjacent electrode positions are more directly compared (differentially amplified between electrodes), can often provide a clearer phase reversal.

Practical Limitations Though the central sulcus may be identified using anatomical landmarks and MRI images, the SSEP-PR technique is regarded as one of the most reliable tools for identifying the central sulcus. Distorted anatomy resulting from displaced cortical structures in the presence of a lesion, individual variations in functional organization and anatomy, lesion-induced cortical plasticity changes, and limitations in preoperative imaging studies all support the complementary use of SSEP-PR intraoperatively for increased accuracy of central sulcus identification.17,​59,​69 Success rates for identification of the central sulcus with SSEP-PR technique are roughly 90% or higher.59,​68,​75,​76,​77 Situations in which SSEP-PR cannot identify the central sulcus include lesion-related displacement of the sulcus, preexisting marked sensorimotor deficits, dysplastic gyri, and anesthetic, or technical issues analogous to those relating to scalp SSEP recording.59 Proposed causes for absent or distorted cortical potentials include a tumor, cortical dysplasia, or a lesion that desynchronizes the propagated afferent electrical volleys along the thalamocortical pathway, distorts the spatiotemporal projection of cortical electrical dipoles to the brain surface, or the recording site may not be appropriate (obscuration by vascular structures or scar) for recording a potential generated in the hand area of the postcentral gyrus.9,​59

28  Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia Some authorities feel that although SSEP-PR is reliable for verifying the location of the central sulcus, it does not directly identify motor function, and when used alone, it is inadequate for preventing postoperative motor deficits. Many believe methods of motor mapping, if possible, are also indicated. However, motor mapping using DCS carries the risk of precipitating a seizure. Additionally, in small children it may not be possible to perform motor mapping under general anesthesia, and in some centers, no alternative may be present than to trust and respect the presence of the primary motor cortex anterior to central sulcus as determined by SSEP-PR.

„„ Motor Evoked Potentials Direct Cortical and Subcortical Electrical Motor Stimulation DCS is a mapping technique in which electrical current stimulation is applied to the exposed cerebral cortex at the time of craniotomy. It is preferably performed after central sulcus localization. A handheld monopolar, bipolar, or subdural strip or grid electrode may be used as stimulating electrodes. Although DCS is utilized during awake craniotomies with cooperative patient feedback, this chapter emphasizes methods of motor stimulation under general anesthesia with minimal or no muscle paralysis to facilitate CMAP responses for visual or preferably EMG recording of extremity movement.

Stimulation and Recording Under general anesthesia, two DCS-MEP techniques may be employed: bipolar cortical stimulation (Penfield’s/Ojemann’s technique) and multipulse train high-frequency mapping ­ technique (MPT) (Taniguchi’s method). A system check is first ­performed by stimulating the exposed temporalis muscle up to 10 mA to confirm a visual contraction. This confirms an intact stimulator generator, cable connections, battery, and handheld stimulator. It also establishes the paralytic level of the patient. Once the central sulcus of Rolando, precentral sulcus, and precentral gyrus are located (often confirmed with SSEP-PR technique), motor stimulation is usually begun in the suprasylvian precentral gyrus and advanced superiorly. Maximum stimulation efficiency often requires stimulation of cortex nearest the central sulcus.7 Current is gradually increased observing for a motor response. Cortical patches of about 1 cm are chosen for stimulating different nonadjacent areas, along with pauses of 10 to 15 seconds between stimulations to lessen the chance of an intraoperative seizure. Uncommonly, motor cortex electrical stimulation blocks ongoing motor activity rather than eliciting motor movement.78 The mechanism is uncertain but may presumably involve excitation of underlying inhibitory mechanisms or a depolarization blockade beneath the electrode.79,​80

Bipolar Cortex Stimulation (Penfield’s Technique) Bipolar rectangular pulses with 0.5- to 1-ms duration are delivered via a bipolar handheld probe with about 5-mm tip spacing, or a subdural grid is applied to the exposed motor cortex at 50 to 60 Hz for approximately 1 to 4 seconds.27 Threshold intensities for eliciting a motor response under general

­ nesthesia are determined by starting at 4 mA, and increasing a by increments of 0.5 to 2 mA.7 Threshold amplitudes for evoking a motor response are typically less than 10 mA.17,​19,​24,​27,​63,​65 The upper limit for Penfield’s stimulation is generally 15 mA. Ojemann further refined the Penfield technique by introducing a 60 Hz 1 ms biphasic current-constant pulse delivered with a bipolar probe.65,​81,​82,​83,​84,​85,​86 The motor responses to stimulation tend to be more tonic in nature.

The Multi-Pulse Train High-Frequency Technique: Taniguchi’s Method Trains of four to nine (typically five) monophasic anodal rectangular pulses 200 to 500 μs (0.2–0.5 ms) in duration with an interstimulus interval of 2 to 4 ms are delivered via a handheld probe or subdural grid to the exposed motor cortex.87 Threshold amplitudes are identified by increasing amplitude at increments of 0.5 to 2 mA, but generally not exceeding 25 mA. In a large study of 422 patients utilizing DCS with the MPT, current up to 30 mA was used for cortical mapping with a train of five to seven anodal stimuli of 0.5-ms pulse duration with a 4-ms interstimulus interval and a train repetition rate of 2 Hz. In cases in which the patient had a history of seizures, the train repetition rate was decreased to be as low as 0.4 Hz.88 The mean threshold for primary motor cortex stimulation was reported to be 6 to 12 mA.87 The MPHF stimulation technique can be applied in a bipolar, or a monopolar fashion with the return electrode in exposed temporalis muscle or scalp.17,​19,​24,​69 Motor responses tend to be finer than using the Penfield technique with more limited movement of a restricted number of muscles.34,​76 The authors prefer the MPT stimulation technique due to its decreased incidence of induced seizures, lower delivered total stimulation charge, and lower stimulus artifact during electrocorticography (ECoG) recording for afterdischarge (AD) potentials or epileptiform discharges. The MPT technique allows for a more quantitative analysis of elicited motor responses, both as a mapping and if the cortical stimulating electrodes do not interfere with the resection, then a continued motor monitoring technique during the subsequent cortical or subcortical resection.66,​87,​89,​90,​91,​92,​93,​94,​95 In both the MPT and Penfield’s/Ojemann’s DCS methods, it is recommended to electrically stimulate the entire cortical area of interest before incrementally increasing the stimulus intensity. Penfield’s technique is associated with a higher risk of induced seizures than the MPT technique. The incidence of seizures after MPT DCS has been reported to be 1.6%, and 5 to 24% after the Penfield DCS stimulation technique.88 With both DCS techniques, it is recommended that multiple surrounding electrodes or a subdural grid be placed on the exposed cortex for ECoG in areas adjacent and circumferentially to the stimuli, to check for the presence of prolonged ADs, a warning sign of possible forthcoming seizures. With either stimulating technique, it is advised to quickly respond to an intraoperative electrographic or clinical seizure with ice-cold saline or Ringers’ solution applied to the cortical surface.17,​24,​69,​96,​97,​98 The anesthesia team must have intravenous anticonvulsants or sedatives on hand to treat a possible generalized seizure, which could result in brain swelling. Administering bolus sedatives, however, can impair continued cortical mapping by directly decreasing neuronal excitability and synaptic activity.17,​53

267

268

III  Surgical Anatomy and Mapping Techniques

Recording Under general anesthesia, motor responses to DCS are documented by subdermal needle EMG electrodes placed in contralateral muscle groups in position to visibly observe extremity movement during stimulation, which can range from slight movement to a brisk clonic or tonic motion risking injury. Penfield’s stimulation typically elicits a more tonic muscle response, whereas the MPT technique elicits a single quantifiable (EMG capture), CMAP response.90

Application during Lesion Resection in Proximity to Eloquent Cortex Identification of the central sulcus may be difficult due to distorted anatomy secondary to a lesion or vascular channels in and about the sulcus. Although SSEP phase reversal (see above) and improved anatomical and functional imaging techniques may enable identification of the central sulcus and precentral gyrus, many feel intraoperative DCS, or extraoperative stimulation (usually in the intensive care unit following operative placement of cortical strip or grid cortical surface electrodes) remains the gold standard for verifying functional primary motor cortex. Under general anesthesia without muscle paralysis, we perform DCS of the presumed precentral, primary motor cortex after central sulcus localization by SSEP phase reversal. When using DCS for cortical mapping, under general anesthesia, threshold values of less than 10 mA are generally accepted as indicative of eloquent motor cortex localization, for either ­Penfield’s or multi-pulse simulation methods.17,​24,​69 The MPT technique has become increasing popular due to decreased incidence of induced seizure, lower delivered total charge during stimulation, and minimal stimulus artifact on ECoG recordings. The MPT technique has also allowed a more quantitative analysis of elicited motor responses, increasing the value of DCS as a monitoring technique. Once DCS has successfully localized motor function, a grid or small strip electrode array may then be placed to stimulate and evoke muscle responses utilizing the MPT or Penfield’s/Ojemann’s techniques as an ongoing motor monitoring tool during the subsequent cortical resection, if it does not get in the way.66,​87,​89,​90,​91,​92,​93,​94,​95 An increase in stimulus threshold of 4 mA necessary to evoke CMAP responses has been suggested as a criterion for significant change during such monitoring.94,​99

Practical Limitations of Direct Cortical Motor Mapping To maximize specificity of location, and the control of unwanted electrical stimulation spread, it is important to continually use threshold or near-threshold settings. Supramaximal DCS settings may uncontrollably activate motor fibers and even spread or jump further distal thus decreasing local sensitivity and specificity. By the use of threshold or near-threshold DCS as a monitoring tool during cortical resections, it is only the patients with significant signal deterioration or increased thresholds who exhibited continued motor deficits 3 months after surgery.92,​93,​94 DCS performed under general anesthesia is generally confined to the primary motor area, as stimulation of the supplementary motor area is not generally considered eloquent and produces more complex tonic movements.2 Because

of the great differences in stimulus amplitude measures between DCS and Transcranial electrical stimulation (TES) is utilized when motor monitoring is desired, but the primary motor cortex is not surgically exposed. Patient movement during DCS is more minimal than with TES, and especially with high-frequency train techniques more localized, and much less of a concern than with TES (see below). If CMAPs are absent during stimulation, then the function of the stimulating probe can be verified by checking for stimulation artifact on ECoG or stimulating the exposed temporalis ­muscle and observing for a twitch response. Systemic usage of particular anesthetic agents at higher dosages can result in deterioration of CMAPs resulting from DCS be it traditional direct cortical (Penfield’s/Ojemann’s) or MPT techniques (see “­Anesthetic Considerations” later). There are significant age-related differences and limitations on DCS in infants, toddlers, children, and adults.12 Not until approximately 4 years of age does the primary motor cortex begin to exhibit a sufficiently low electrical threshold to induce muscle responses,19 and the adult pattern does not usually appear until adolescence.12,​64 This initially high threshold is due to immature development of pyramidal or CST tract nerve cells, smaller axon diameters due to poor myelination, and likely immature supportive synaptic activity.18 DCS in the young child may be crudely mapped, but general anesthesia often additionally suppresses the ability to elicit consistent responses.100 Furthermore, young children differ from adults as they rarely demonstrate motor responses in the absence of electrical ADs.101,​102,​103 Motor responses secondary to traditional cortical stimulation are obtained in children 16 years or older, but unobtainable in roughly 50% age 8 to 9 years, 80% aged 4 to 5 years, and absent in children under 1 year of age.101,​104 In children under 2 to 4 years of age, motor responses in the absence of an AD is rarely found despite the use of maximal stimulation.102 Between roughly ages 4 and 6, motor responses without an associated AD begin to appear, and at age 6 years, tonic flexion or extension finger movement responses emerge.12 In patients under 4 years of age, only a novel “dual cortical stimulation paradigm” and not traditional Penfield’s stimulation reliably elicited both ADs and motor responses.103 By gradual stepwise increments of stimulus intensity, but in particular stimulus duration, the chronaxie of the strength–duration curve for poorly myelinated fibers is improved upon to facilitate ADs and allow motor cortical stimulation. This occurs with decreased overall energy expenditure, and an improved electrical safety profile.12,​101,​102,​103,​104 In children 5 to 10 years of age, both paradigms (traditional Penfield’s and the “dual cortical stimulation”) elicited functional responses.100,​101,​102,​103

Direct Subcortical Stimulation It may be preferable when working in or near eloquent c­ ortical areas such as a resection along the white matter margin of a determined epileptogenic zone or an adjacent structural lesion to perform DSS. Stimulation parameters for DSS are generally the same or lower as myelinated fibers are directly ­stimulated.7,​104,​105,​106 Subcortical and cortical motor (CST) stimulation can be performed under general anesthesia, ­ while sensory, speech, and language subcortical stimulation ­techniques all require the cooperation of a patient under local anesthesia.107 After DCS mapping of the primary motor cortical areas, by ­traditional Penfield’s or MPT technique, DSS may be

28  Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia used to detect corresponding descending motor pathways. Conversely, deep to the primary sensory cortex one may stimulate the ascending cortical fibers while recording ECoG from the primary sensory cortex. Again, beware that descending or ascending eloquent white matter fibers may not travel perpendicular to the gyral crown.7 Bipolar stimulation is felt to provide more precise results, as the current field produced is smaller than with monopolar stimulation, and limited between the probe tips. However, some prefer monopolar stimulation due to the homogeneous current field created by radial current spread, which can also be used to estimate the distance between the stimulation site and CST.90,​94,​108 The ratio of threshold current to distance from the CST during DSS is approximately 1 to 1.5 mA or 1 mm. It is suggested that resections should be stopped when subcortical stimulation thresholds are 2 mA, and that higher thresholds indicate a safe distance from the CST.99,​109,​110 DCS used along with DSS during resections has shown a combined sensitivity and specificity of about 67 and 97%, respectively, for prediction of iatrogenic injury in a study of 100 patients.94,​99 In summary, DCS loss or change greater than 4 mA is prognostically predictive and directly worrisome for neurological injury, as compared to DSS, which does not directly assess neurological function or CST integrity, only CST proximity.94,​99 The CST D- and I- wave MEP recordings (Fig. 28.5) are useful under general anesthesia as they are much less sensitive to inhalational and intravenous anesthetic agents, and not affected by muscle relaxants as are muscle recorded MEPs (mMEPs). However, the CST responses have much smaller amplitudes than mMEPs (microvolts versus millivolts), although signal averaging of the CST responses can improve their recognition. Currently, D- and I-wave CST-MEP recordings have found use in spinal cord tumor surgery. Stopping tumor resection when D-wave amplitude decreases by 50% correlates with the patient’s motor outcome.33,​111 Unfortunately, recorded D-wave amplitude progressively diminishes in that; about 50% of the CST is given off in the cervical cord enlargement region, 20% thoracic cord, and the final 30% of the CST terminates in the lumbosacral cord.18 CST D-waves are at times difficult to record from the lumbosacral cord and generally not recordable from the cauda equina.33,​111 The CST D-wave recordings after DCS or DSS, if further investigated, could be of significant benefit in cranial surgery, with perhaps exacting amplitude thresholds as noted in spinal cord tumor surgery. In recent decades, patients with very large cerebral lowgrade gliomas, bordering on eloquent cortical and subcortical motor and sensory pathways underwent complete excisions of their tumor, without permanent neurological deficits. These excisions have highlighted the advantages of microneurosurgical technique, neuronavigation, ultrasonic tumor aspiration, SSEP monitoring, as well as cortical and subcortical CST stimulation monitoring.86,​105 Unfortunately, in some instances functionally eloquent cortical or subcortical tissue may be present within the boundaries of tumor tissue, and in addition, slow-growing tumors may allow time for plasticity of eloquent functions to an alternate region or regions. Although slow-growing low-grade gliomas usually respect gyral and pial boundaries and tend to displace rather than invade functional sensory and motor pathways, large areas of cortical dysplasia and associated developmental tumorous masses as found in children with epilepsy may not.112 The use of ultrasonic aspiration has found use in ­epilepsy

surgery in the technique of gyral emptying associated with careful respect for pial boundaries and microvascular injury.113 In recent years, an effort has been made to develop an electrified handheld ultrasonic brain tumor aspirator tip that can intermittently or continuously provide direct, real-time subcortical CST monitoring although a number of technical challenges remain.114,​115,​116

„„ Transcranial Electrical Stimulation As mentioned earlier, TES has become the clinical method of choice for intraoperative CST monitoring when the primary motor cortex is not exposed or does not require exposure at ­surgery, but the descending motor pathways may be in jeopardy (Fig. 28.5).33,​117 However, TES-MEP is not routinely used in ­pediatric or adult epilepsy surgery, but may ­occasionally be used in children age 4 to 5 or older, at lowered current settings to monitor the motor pathway. TES and transcranial magnetic stimulation (TMS) have aided our understanding of basic motor cortex physiology as well as the critical anesthetic effects upon both MEPs and SSEPs.17,​41,​42,​43,​53,​118,​119,​120,​121,​122 In the mid-1990s, s­everal groups first demonstrated the above described pulse train, electrical motor stimulation technique under general anesthesia.123,​124,​125 MacDonald’s 2002 report alleviated safety concerns regarding the human use of TES, the first government-­ ­ approved commercial stimulators came on the market, and TES for intraoperative clinical usage under general anesthesia and research use expanded.19,​126 Currently, the MPT-MEP stimulation technique is the preferred intraoperative methodology in TES under general anesthesia for many neurosurgical operations. This includes many spinal surgeries and some craniotomies where motor system monitoring is desirable, but the primary motor cortex itself does not require exposure. TES requires significantly more energy (both current and voltage) than DCS; however, in a large series of over 4,000 monitored patients undergoing TES, a much less risk of seizure (0.7%) was found compared to 422 patients who underwent traditional Penfield’s DCS (5.4%).88

Physiology, Stimulation, and Recording The somatotopically organized primary motor cortex is selected for TES due to the low threshold necessary to induce muscle responses.19 The large myelinated, most rapidly conducting, CST axons in the deeper layers or just below the primary motor cortex are thought to be predominately activated fibers during TES stimulation. These conduct action potentials to lower motor neurons, some of which without an interneuron synapse.19,​20,​ 21,​22,​34 Studies in subhuman primates showed direct waves (D waves) recorded from the CST are produced from a single-pulse transcranial electrical stimulus. This was subsequently verified in humans undergoing intramedullary tumor surgery. Multipulse train stimuli were more effective under general anesthesia by eliciting D-wave volleys produced by direct axonal stimulation. Indirect waves (I waves) are produced by intracortical circuits that incite additional cortical motor neurons and follow the D waves when sufficiently strong pulse train stimuli are used. Stimuli with adequate D-wave activation along with some I-wave recruitment produces enough temporal summation of excitatory postsynaptic potentials (EPSPs)

269

270

III  Surgical Anatomy and Mapping Techniques

Fig. 28.5  General anesthesia, craniotomy performed, and the motor cortex exposed. The authors prefer the multi-pulse train high-frequency stimulation (MPT, Taniguchi’s method) for direct cortical motor stimulation (DCS-MEP), and MPT stimulation is generally essential for transcranial electrical stimulation (TES-MEP). (a) With TES-MEP, bilateral corticospinal tracts (CSTs) are stimulated resulting in spinal epidural recorded corticospinal tract recordings (C) of direct (D) and indirect (I) waves after progressively increased electrical voltage delivery (up to 750 V). (b) Direct cortical stimulation (DCS-MEP) is performed by electrical grid current stimulation (usually equal or less than 25 mA) to achieve single upper or lower extremity limb muscle recordings as seen in D. However, significantly deeper penetration of the electrical energy into the brain is noted with TES-MEP resulting in bilateral cortical stimulation, shortened latency, and extremity muscle responses as well. Note the sharper epidural recording C on the left proximal to a spinal tumor compared to distal to the tumor, with increased D-wave amplitude and shortened latency (C—upper tracing). (c) An increase in stimulation voltage is also accompanied by an increasing prominence and number of I waves. However, D- and I-wave recordings, which are much less affected by anesthetic and certainly muscle paralyzing agents have been a challenge to record with DCS. (Reproduced with permission from Deletis and Sala 2008.33)

to activate lower motor neurons, resulting in CMAP potentials (see Fig. 28.5) .17,​19,​20,​21,​24,​34 For TES, monophasic anodal trains of rectangular pulses are delivered through the scalp to motor cortex over each hemisphere. Optimal active scalp stimulation sites at C1–C2 for upper extremities and C3–C4 for lower extremities are selected as they minimize threshold current and maximize repeatable muscle responses. See Fig. 28.5. TES is routinely used for intraoperative spinal monitoring, or when motor tract monitoring is required in an open craniotomy that does not expose the peri-Rolandic region. For lower extremity TES responses, midline regions for TES stimulation, such as Cz to inion, have been recently reported to be more preferable than the standard C3–C4 or C4–C3.127 Spiral (corkscrew) needle scalp stimulating electrodes are optimal for TES, since they are secure and rarely displaced.19,​21,​97 Typical parameters for TES are 3 to 8 pulses, 50 to 1,000 μs pulse widths, and interstimulus intervals of 3 to 4 ms. Constant current and constant voltage stimulus generators are commercially

available with upper safety limits of 200 mA or 1,000 V. However, the exact combination of pulse width and number of pulses may limit current or voltage amplitudes, as overall delivered charge does not exceed maximum safety limits.17,​19,​126,​128 Because the skull has a very high impedance, only 10 to 20% of delivered current is believed to reach the motor cortex, which equates with a higher safety margin than that of DCS, and in accord with the decreased incidence of seizure with TES compared to DCS.129 TES stimulation is believed to occur axonally, and as stimulation intensity is increased, the D-wave latency shortens, a likely indication of stimulation occurring at an increased depth within the white matter. At higher levels approaching the maximal settings of commercially available equipment, stimulation may occur as deep as the pyramidal (CST) decussation at the foramen magnum associated with shortening of D-wave latency.130 Thus, it is critical to optimize (but not significantly overshoot) stimulation charge delivery to avoid stimulation jumping distal to the surgical site and dangerously confounding the monitoring results.24,​33,​130,​131 CMAPs are

28  Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia recorded referentially by two needle electrodes placed in the muscle belly 2 to 4 cm apart. Areas well represented in the primary motor cortical homunculus, such as hands and feet, are more easily activated. Commonly used muscle sites are brachioradialis, abductor pollicis brevis, abductor digiti minimi, tibialis anterior, and abductor hallucis. CMAPs are bandpass filtered between 10–100 Hz and 15,00–3,000 Hz.17,​19 Some stimulus artifact may be present but is useful to confirm stimulus delivery. Normal CMAPs are polyphasic with variable waveforms and amplitudes between stimuli, suggesting interstimulus variation in the activated anterior horn cell pool and their accompanying motor units. Interestingly, D-wave recordings (which require an electrode in proximity to the spinal cord) do not show such variations. CMAP amplitudes range between 10 and 1,000 μV, and signal averaging as used with the SSEP is not required. Cervical innervated muscle latencies are about 10 to 40 ms and lower extremity muscles with longer latency depending on leg length. The short pulses (50–1,000 μs duration) used during TES are considered safe, as electrochemical injury occurs only with greater than 1 ms (1,000 μs) pulse duration of prolonged monophasic train stimuli. The commercially available stimulators have safety limits, and thermal scalp burns are exceptionally rare. Seizures are very unlikely (0.03%) with the brief, high-frequency trains utilized in TES. The most common TES complications are tongue bite injuries caused by corticobulbar jaw stimulation and an improperly placed bite block. The soft bite blocks should be placed between both sets of molar teeth and kept in place.19,​132 Acting as a conditioning or priming stimulus double pulse train MEP stimulation can favorably augment mMEPs especially in neurologically impaired or young children with small or unobtainable responses.21,​34,​133 Similarly, stimulation of a peripheral nerve at the palm or sole of the foot just before DCS or TES mMEP has been found to facilitate the response.34,​134,​135 Others have adopted threshold TES-MEP methodology whereby an increase in the stimulation threshold to produce the same muscle response represents a worrisome change in the viability of CST response and resultant D waves and mMEPs. However, this technique assumes anesthetic agent concentrations and the level of muscular relaxation are constant.136,​137

Anesthetic Considerations As in the use of SSEP under general anesthesia, TES-MEP is altered by certain anesthetics.17,​34,​53,​138 The neurophysiology and anesthesia team must work closely to make sure that TES is feasible and avoid issues that make CMAP interpretation uncertain. TES responses are more sensitive to inhalational agents than SSEP responses. In some cases, administering 0.5% minimal alveolar concentration is tolerable for SSEP responses but may result in absent or nonreproducible TES responses. The widely recommended anesthetic combination for TES recording is IV propofol and opioid (such as remifentanil) known as total intravenous anesthesia.19,​21,​22,​33,​51,​96,​105,​136,​139,​140,​141 Remifentanil, in particular, allows rapid and reliable emergence from even prolonged surgery and a much wider dosage window for mMEP than other opoids.34 Ketamine can be considered in place of propofol in younger pediatric patients where neuromonitoring is problematic.34,​134 Muscle relaxants are generally not recommended during

TES, but low levels may be used if kept constant in conjunction with frequent train-of-four extremity responses eliciting four ­twitches.19,​36,​37 D-wave recordings are much less affected by ­anesthetic levels and not affected by muscle-paralyzing agents.

Applications and Interpretation during Lesion Resection in Proximity to Eloquent Cortex Anatomical distortion of the central sulcus of Rolando by adjacent mass lesions, localized cortical atrophy, dysplastic cortex, or the imposition of obscuring variations in the sulcal arterial and venous channels may render the central sulcus difficult to accurately localize. Also, in some patients, slow-growing cerebral lesions are known to displace eloquent cortical functions, from their classical cortical locations, a form of cortical plasticity. Although anatomical and functional neuroimaging may enable precentral gyrus identification, DCS is still considered the gold standard to verify the primary motor cortex. Intermittent checking of transcranial motor signals, particularly when operating in close proximity to possibly eloquent brain is recommended.

„„ Navigated Transcranial Magnetic Motor Stimulation A novel presurgical adjuvant is nTMS. nTMS can detect the presence of functional eloquent cortex within a tumor’s boundary and the presence of cortical plasticity or movement of particular cortical functions to alternative regions, secondary to neurological deficit, slow tumor growth, or previous operation. nTMS is a painless procedure that delivers a brief magnetic pulse generated and delivered by a high-strength, electrified copper coil held over the scalp. The resulting conical-shaped field of magnetic energy (based on Faraday’s law of electromagnetic induction) passes through the scalp and skull to induce an electrical current, which depolarizes localized neurons in the underlying cerebral cortex to generate action potentials.142,​143,​144 Single pulses will briefly stimulate the motor cortex, while repetitive TMS (rTMS) pulse trains can have either an inhibitory or stimulation effect on the cortex and are more often used in language testing.145,​146 Precise positioning of the stimulation coil at cortical sites of stimulation in the awake, preoperative patient is captured by the use of a standard sensor-based neuronavigational frameless stereotactic system and superimposed on the patient’s previously obtained high-resolution MRI scan. Thus, preoperatively and in real time, a relatively precise location of a motor point as well as its threshold (strength of the magnetic pulse to obtain the response) is obtained. In addition, eloquent cortical motor point accuracy derived from presurgical nTMS was found to be within the same gyrus and within 4 to 5 mm of that derived in the same patients following direct cortical motor stimulation at tumor or cortical resective epilepsy surgery.147,​148,​149,​150,​151 nTMS can also be used in cooperative children for language localization, identify regions of cortical plasticity related to previous neurological deficits or a prior cerebral resection, and significantly aid the safety of approach for reexploration tumor or epilepsy surgery.144,​152

271

272

III  Surgical Anatomy and Mapping Techniques Single-pulse TMS, as used in nTMS is well tolerated and associated with very minimal adverse effects such as machine noise, occasional headaches, and temporary hearing reduction despite use of earplugs. Seizures are extremely uncommon and only occurred in relatively few cases over a 30-year period with hundreds of thousands examined subjects.153 Most of these have occurred in conscious patients with a seizure history or on pro-epileptic medications.142,​144,​154 TMS delivered in trains or rTMS has been reported to elicit seizures in conscious patients with or without a seizure history.142,​155 rTMS usually given at a rate of 10 Hz is commonly used for the treatment of depression, and currently FDA approved for mapping of speech and language. As many language sites are below the temporalis muscle and facial nerve, they may produce annoying muscle spasms during testing which mimic dysarthria. This dysarthria from rTMS can usually be differentiated from aphasia as the majority of language sites produce speech arrest.144,​152 Absolute contraindications to the use of TMS include intracranial metallic devices such as aneurysm clips and deep brain stimulators, and other internal electronic devices or pulse generators such as cardiac pacemakers, defibrillators, or cochlear implants.142,​156

„„ Future Directions The challenging study and surgical treatment of ­intractable childhood epilepsy is among the most provocative and humbling endeavors a physician could undertake. As it stands, we have a number of surgical adjuncts to aid us in the

localization of eloquent extremity motor and sensory brain tissue under general anesthesia. We believe a direction toward improving this task is further perfection in the utility of MEP, including the possible capture of D waves secondary to direct cortical or subcortical stimulation. D waves are not subject to the anesthetic and muscle relaxant problems of muscle recorded MEPs and would appear to provide more direct information. In addition, the application of c­onditioning or priming stimuli before direct electrical cortical or subcortical stimulation could lower threshold levels for these modalities—similar to transcranial electrical monitoring techniques. Another advance would be the development of a handheld ultrasonic brain tissue aspirator that can deliver electrical current for mapping or monitoring, when combined with CST image-guided navigation, may more precisely determine the relationship between current density and distance to the activated CST.23 Finally, Goldring et al had hopes that the “direct (electrical) cortical response (DCR)”, which he found unique to histologically different cortical regions, would prove a means of recognizing healthy from unhealthy cortex.157 We believe methods like multiple subpial transection could be subjected to DCR before and after partial or complete disconnection to better objectify the level of disconnection, standardize the procedure, and study the surgical results. Surgery for intractable childhood epilepsy is a rapidly advancing discipline as improved diagnosis combined with newer forms of navigation, and novel intraoperative neurophysiological m ­ onitoring techniques will undoubtedly continue to result in both improved surgical safety and optimal results in seizure control.

References 1. Engel J Jr. Non-pharmacological therapy of seizures. Seizures and Epilepsy. 2nd ed. Vol. 83. UK: Oxford University Press; 2013:605 2. Lachhwani DK, Dinner DS. Cortical stimulation in the definition of eloquent cortical areas. Handbook of Clinical Neurophysiology. 2003;3:273–286 3. Delev D, Send K, Wagner J, et al. Epilepsy surgery of the rolandic and immediate perirolandic cortex: surgical outcome and prognostic factors. Epilepsia 2014;55(10):1585–1593 4. Celix JM, Silbergeld DL. Intraoperative Cortical Mapping Techniques and Limitations. Functional Mapping of the Cerebral Cortex. Wien, Austria: Springer; 2016:63–76 5. Kanner AM, Morcillo MCA. Mapping of Eloquent Cortex in Focal Epilepsy with Intracranial Electrodes. Functional Mapping of the Cerebral Cortex. Wien, Austria: Springer; 2016:105–113 6. Mueller WM, Yetkin FZ, Haughton VM. Functional magnetic resonance imaging of the somatosensory cortex. Neurosurg Clin N Am 1997;8(3):373–381 7. Schuster J, Silbergeld D. Motor, sensory, and language mapping and monitoring for cortical resections. Youmans Neurological Surgery. Philadelphia, PA: Saunders; 2004:2531–2539 8. Julkunen P, Karhu J. Brain Plasticity in Neurosurgery. Navigated Transcranial Magnetic Stimulation in Neurosurgery. Switzerland: Springer; 2017:267–285 9. Kessler S. Non-invasive Brain Stimulation in Pediatric E ­ pilepsy: Diagnostic and Therapeutic Uses. In: Kirton A, G ­ ilbert DL, eds. Pediatric Brain Stimulation: Mapping and Modulating the Developing Brain. London: Academic Press; 2016: 281–304

10. Mäkelä JP, Vitikainen AM, Lioumis P, et al. Functional plasticity of the motor cortical structures demonstrated by navigated TMS in two patients with epilepsy. Brain Stimul 2013;6(3):286–291 11. Huttenlocher PR. Neural plasticity: The Effects of Environment on the Development of the Cerebral Cortex. Cambridge, MA: Harvard University Press; 2009 12. Duchowny M, Jayakar P. Functional cortical mapping in children. In: Devinsky O, Beric A, Dogali M, eds. Electrical and Magnetic Stimulation of the Brain and Spinal Cord. Advances in Neurology Vol 63. NY: Raven Press; 1993: 149–154 13. Kim SS, McCutcheon IE, Suki D, et al. Awake craniotomy for brain tumors near eloquent cortex: correlation of intraoperative cortical mapping with neurological outcomes in 309 consecutive patients. Neurosurgery 2009;64(5):836–845, discussion 345–346 14. Moller AR. Intraoperative Neurophysiological Monitoring. 2nd Ed. Totowa, NJ: Humana Press Inc; 2006. 15. Nuwer MR, Packwood JW. Somatosensory evoked potential monitoring with scalp and cervical recording. In: Nuwer MR, ed. Intraoperative Monitoring of Neural Function. Handbook of Clinical Neurophysiology. Vol. 8. Amsterdam: Elsevier; 2008:180–189 16. Toleikis JR; American Society of Neurophysiological Monitoring. Intraoperative monitoring using somatosensory evoked potentials. A position statement by the American ­ Society of Neurophysiological Monitoring. J Clin Monit Comput 2005;19(3):241–258 17. Simon M. A Comprehensive Guide to Monitoring and Mapping. New York, NY: Demos Medical; 2010

28  Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia 18. Lassek AM. The Pyramidal Tract—Its Status in Medicine. Charles C. Thomas. IL USA: Springfield; 1954. 19. Macdonald DB, Skinner S, Shils J, Yingling C; American Society of Neurophysiological Monitoring. Intraoperative motor evoked potential monitoring—a position statement by the American Society of Neurophysiological Monitoring. Clin Neurophysiol 2013;124(12):2291–2316 20. Amassian VE. Animal and human motor system neurophysiology related to intraoperative monitoring. In Deletis V, Shils J, eds. Neurophysiology in Neurosurgery: A Modern Intraoperative Approach. London:Academic Press; 2002:3–23 21. Deletis V. Intraoperative neurophysiology and methodologies used to monitor the functional integrity of the motor system. New York, Amsterdam: Academic Press; 2002 22. Macdonald DB. Intraoperative motor evoked potential monitoring: overview and update. J Clin Monit Comput 2006;20(5):347–377 23. Yingling CD. Mapping the corticospinal tract. In: Nuwer MR, ed. Intraoperative Monitoring of Neural Function. Handbook of Clinical Neurophysiology. Vol. 8. Amsterdam: Elsevier; 2008:319–331 24. Szelenyi A. Intraoperative neurophysiological monitoring under general anesthesia. New York, NY: SpringerWien; 2011 25. Olivier A, Boling WW, Tanriverdi T. Surgical anatomy. Endopial resection (intervascular endopial gyral emptying). Surgical treatment of cortical dysplasias. Cambridge, UK: Cambridge University Press; 2012 26. Cushing H. A note upon the faradic stimulation of the postcentral gyrus in conscious patients. Brain 1909;32(1):44–53 27. Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain 1937;60(4):389–443 28. Uematsu S, Lesser RP, Gordon B. Localization of sensorimotor cortex: the influence of Sherrington and Cushing on the modern concept. Neurosurgery 1992;30(6):904–912, discussion 912–913 29. Nii Y, Uematsu S, Lesser RP, Gordon B. Does the central sulcus divide motor and sensory functions? Cortical mapping of human hand areas as revealed by electrical stimulation through subdural grid electrodes. Neurology 1996;46(2):360–367 30. Yousry TA, Schmid UD, Alkadhi H, et al. Localization of the motor hand area to a knob on the precentral gyrus. A new landmark. Brain 1997;120(Pt 1):141–157 31. Brodmann K. Vergleichende Lokalisationslehre der Grosshirnrinde in ihren Prinzipien dargestellt auf Grund des Zellenbaues. Barth; 1909 32. Kahle W, Frotscher M. Colour Atlas and Textbook of Human Anatomy: Nervous System and Sensory Organs. New York, NY: Thieme; 2003 33. Deletis V, Sala F. Corticospinal tract monitoring with D- and I-waves from the spinal cord and muscle MEPs from limb muscles. In: Nuwer MR, ed. Intraoperative Monitoring of Neural Function. Handbook of Clinical Neurophysiology. Vol. 8. Amsterdam: Elsevier; 2008:235–251 34. Journee HL. Motor EP physiology, risks and specific anesthetic effects. In: Nuwer MR, ed. Intraoperative Monitoring of ­Neural Function. Handbook of Clinical Neurophysiology. Vol. 8. ­Amsterdam: Elsevier; 2008;8:218–234 35. Deletis V, Rodi Z, Amassian VE. Neurophysiological mechanisms underlying motor evoked potentials in anesthetized humans. Part 2. Relationship between epidurally and muscle recorded MEPs in man. Clin Neurophysiol 2001;112(3):445–452 36. Sloan TB. Muscle relaxant use during intraoperative neurophysiologic monitoring. J Clin Monit Comput 2013;27(1):35–46 37. Gavrancic B, Lolis A, Beric A. Train-of-four test in intraoperative neurophysiologic monitoring: differences between hand and foot train-of-four. J Clin Neurophysiol 2014;31(6):575–579 38. Gilmore RL. Somatosensory Evoked Potentials in Pediatrics Normal. In: Holmes GL, Jones HR, Moshe SL. Clinical Neurophysiology of Infancy, Childhood and Adolescence. Oxford,UK: Butterworth-Heinemann; 2006:168–181

39. Aminoff MJ. Somatosensory Evoked Potentials in Infants and Children. In: Aminoff’s Electrodiagnosis in Clinical Neurology. 6th Ed. NY: Elsevier/Sanders; 2012: 603–614. 40. Taylor MJ, Boor R, Ekert PG. Preterm maturation of the somatosensory evoked potential. Electroencephalogr Clin Neurophysiol 1996;100(5):448–452 41. Ghaly R, Stone J, Levy W. Protocol for intraoperative SSEP-Myogenic MEP recordings. Neurosurgery 1991;29:480–482 42. Ghaly RF, Stone JL, Levy WJ. Intraoperative Motor Evoked Potential Monitoring. Amsterdam: Elsevier; 1991 43. Ghaly RF, Stone JL, Lee JJ, Ham JH, Levy WJ. Monitoring spinal cord motor and somatosensory evoked potentials in anesthetized primates. Neurol Res 1999;21(4):359–367 44. Sloan TB. Anesthetic effects on electrophysiologic recordings. J Clin Neurophysiol 1998;15(3):217–226 45. Pathak KS, Brown RH, Cascorbi HF, Nash CL Jr. Effects of fentanyl and morphine on intraoperative somatosensory cortical-evoked potentials. Anesth Analg 1984;63(9):833–837 46. Kochs E, Treede RD, Schulte am Esch J. [Increase in somatosensory evoked potentials during anesthesia induction with etomidate] Anaesthesist 1986;35(6):359–364 47. McPherson RW, Sell B, Traystman RJ. Effects of thiopental, fentanyl, and etomidate on upper extremity somatosensory evoked potentials in humans. Anesthesiology 1986;65(6):584–589 48. Schubert A, Licina MG, Lineberry PJ. The effect of ketamine on human somatosensory evoked potentials and its modification by nitrous oxide. Anesthesiology 1990;72(1):33–39 49. Tobias JD, Goble TJ, Bates G, Anderson JT, Hoernschemeyer DG. Effects of dexmedetomidine on intraoperative motor and ­ somatosensory evoked potential monitoring during spinal surgery in adolescents. Paediatr Anaesth 2008;18(11):1082–1088 50. Zouridakis G, Papanicolaou A. A concise guide to intraoperative neuromonitoring. Boca Raton, FL: CRC Press; 2001 51. Sloan T. Intraoperative neurophysiology and anesthesia managment. Vol. VII. San Diego, CA: Academic Press; 2002 52. Sloan TB, Fugina ML, Toleikis JR. Effects of midazolam on median nerve somatosensory evoked potentials. Br J Anaesth 1990;64(5):590–593 53. Nathanson M, Moppett IK, Wiles M. Neuroanaesthesia. England: Oxford University Press; 2011 54. Nuwer MR, Daube J, Fischer C, Schramm J, Yingling CD. Neuromonitoring during surgery. Report of an IFCN Committee. Electroencephalogr Clin Neurophysiol 1993;87(5):263–276 55. Society AE; American Electroencephalographic Society. Guideline eleven: guidelines for intraoperative monitoring of sensory evoked potentials. J Clin Neurophysiol 1994;11(1):77–87 56. Neuloh G, Schramm J. Evoked potential monitoring during ­surgery for intracranial aneurysms. In: Nuwer MR, ed. Intraoperative Monitoring of Neural Function. Handbook of Clinical Neurophysiology. Vol. 8. Amsterdam: Elsevier; 2008:801–814 57. Symon L. The relationship between CBF, evoked potentials and the clinical features in cerebral ischaemia. Acta Neurol Scand Suppl 1980;78:175–190 58. Wiedemayer H, Sandalcioglu IE, Armbruster W, Regel J, Schaefer H, Stolke D. False negative findings in intraoperative SEP monitoring: analysis of 658 consecutive neurosurgical cases and review of published reports. J Neurol Neurosurg Psychiatry 2004;75(2):280–286 59. Kombos T. Somatosensory evoked potentials for intraoperative mapping of the sensorimotor cortex. In: Nuwer MR, ed. Intraoperative Monitoring of Neural Function. Handbook of Clinical Neurophysiology. Vol. 8. Amsterdam: Elsevier; 2008:211–215 60. Goldring S. A method for surgical management of focal epilepsy, especially as it relates to children. J Neurosurg 1978;49(3):344–356 61. Goldring S, Gregorie EM. Surgical management of epilepsy using epidural recordings to localize the seizure focus. Review of 100 cases. J Neurosurg 1984;60(3):457–466

273

274

III  Surgical Anatomy and Mapping Techniques 62. Kelly DL Jr, Goldring S, O’Leary JL. Averaged evoked somatosensory responses from exposed cortex of man. Arch Neurol 1965;13(1):1–9 63. Woolsey CN, Erickson TC, Gilson WE. Localization in somatic sensory and motor areas of human cerebral cortex as determined by direct recording of evoked potentials and electrical stimulation. J Neurosurg 1979;51(4):476–506 64. Allison T. Scalp and cortical recordings of initial somatosensory cortex activity to median nerve stimulation in man. Ann N Y Acad Sci 1982;388:671–678 65. Berger MS, Kincaid J, Ojemann GA, Lettich E. Brain mapping techniques to maximize resection, safety, and seizure control in children with brain tumors. Neurosurgery 1989;25(5):786–792 66. Cedzich C, Taniguchi M, Schäfer S, Schramm J. Somatosensory evoked potential phase reversal and direct motor cortex stimulation during surgery in and around the central region. Neurosurgery 1996;38(5):962–970 67. Stone JL, Ghaly RF, Crowell RM, Hughes JR, Fino JJ Jr. A simplified method of somatosensory evoked potential recording from the cerebral cortical surface: technical note. Clin Electroencephalogr 1989;20(4):212–214 68. Wood CC, Spencer DD, Allison T, McCarthy G, Williamson PD, Goff WR. Localization of human sensorimotor cortex during surgery by cortical surface recording of somatosensory evoked potentials. J Neurosurg 1988;68(1):99–111 69. Neuloh G, Schramm J. Intraoperative neurophysiological mapping and monitoring for supratentorial procedures. San Diego, CA: Academic Press; 2002 70. Nuwer MR. Localization of motor cortex with median nerve somatosensory evoked potentials. Heidelberg, Berlin: Springer; 1991 71. Boling W, Olivier A. Anatomy of important functioning cortex. In: Byrne RW, ed. Functional Mapping of the Cerebral Cortex. Cham, Switzerland: Springer; 2016:23–40 72. Sanmillan JL, Fernández-Coello A, Fernández-Conejero I, Plans G, Gabarrós A. Functional approach using intraoperative brain mapping and neurophysiological monitoring for the surgical treatment of brain metastases in the central region. J Neurosurg 2017;126(3):698–707 73. Szelényi A, Hattingen E, Weidauer S, Seifert V, Ziemann U. Intraoperative motor evoked potential alteration in intra­ cranial tumor surgery and its relation to signal alteration in postoperative magnetic resonance imaging. Neurosurgery 2010;67(2):302–313 74. Kumabe T, Nakasato N, Nagamatsu K, Tominaga T. Intraoperative localisation of the lip sensory area by somatosensory evoked potentials. J Clin Neurosci 2005;12(1):66–70 75. King RB, Schell GR. Cortical localization and monitoring during cerebral operations. J Neurosurg 1987;67(2):210–219 76. Kombos T, Suess O, Funk T, Kern BC, Brock M. Intra-operative mapping of the motor cortex during surgery in and around the motor cortex. Acta Neurochir (Wien) 2000;142(3):263–268 77. Suess O, Ciklatekerlio Ö, Suess S, Da Silva C, Brock M, Kombos T. Klinische Studie zur Anwendung der hochfrequenten monopolaren Kortexstimulation (MKS) für die intraopera­ tive Ortung und Überwachung motorischer Hirnareale bei Eingriffen in der Nähe der Zentralregion. Klin Neurophysiol 2003;34(3):127–137 78. Lüders HO, Lesser RP, Dinner DS, et al. A negative motor response elicited by electrical stimulation of the human frontal cortex. Adv Neurol 1992;57:149–157 79. Gugino LD, Aglio LS, Raymond SA, et al. Intraoperative cortical function localization techniques. Tech Neurosurg 2001; 7(1):19–32 80. Ojemann G. Functional mapping of the cortical language areas in adults. Intrapoerative approaches. In: Devinsky O, Beric A, Dogali M, eds. Electrical and Magnetic Stimulation of the Brain and Spinal Cord. Advances in Neurology Vol 63. Raven Press, NY, NY, 1993:155–164

81. Berger MS. Functional mapping-guided resection of low-grade gliomas. Clin Neurosurg 1995;42:437–452 82. Yingling CD, Ojemann S, Dodson B, Harrington MJ, Berger MS. Identification of motor pathways during tumor surgery facilitated by multichannel electromyographic recording. J Neurosurg 1999;91(6):922–927 83. Berger MS. Minimalism through intraoperative functional mapping. Clin Neurosurg 1996;43:324–337 84. Berger MS, Rostomily RC. Low grade gliomas: functional mapping resection strategies, extent of resection, and outcome. J Neurooncol 1997;34(1):85–101 85. Whitaker HA, Ojemann GA. Graded localisation of naming from electrical stimulation mapping of left cerebral cortex. Nature 1977;270(5632):50–51 86. Berger MS, Ojemann GA. Intraoperative brain mapping techniques in neuro-oncology. Stereotact Funct Neurosurg 1992;58(1–4):153–161 87. Taniguchi M, Cedzich C, Schramm J. Modification of cortical stimulation for motor evoked potentials under general anesthesia: technical description. Neurosurgery 1993;32(2):219–226 88. Ulkatan S, Jaramillo AM, Téllez MJ, Kim J, Deletis V, Seidel K. Incidence of intraoperative seizures during motor evoked potential monitoring in a large cohort of patients undergoing different surgical procedures. J Neurosurg 2017;126(4):1296–1302 89. Sala F, Lanteri P. Brain surgery in motor areas: the invaluable assistance of intraoperative neurophysiological monitoring. J Neurosurg Sci 2003;47(2):79–88 90. Kombos T, Suess O, Kern BC, et al. Comparison between monopolar and bipolar electrical stimulation of the motor cortex. Acta Neurochir (Wien) 1999;141(12):1295–1301 91. Deletis V, Camargo AB. Transcranial electrical motor evoked potential monitoring for brain tumor resection. Neurosurgery 2001;49(6):1488–1489 92. Kombos T, Kopetsch O, Suess O, Brock M. Does preoperative paresis influence intraoperative monitoring of the motor cortex? J Clin Neurophysiol 2003;20(2):129–134 93. Kombos T, Suess O, Ciklatekerlio O, Brock M. Monitoring of intraoperative motor evoked potentials to increase the safety of surgery in and around the motor cortex. J Neurosurg 2001;95(4):608–614 94. Seidel K, Beck J, Stieglitz L, Schucht P, Raabe A. The warning-sign hierarchy between quantitative subcortical motor mapping and continuous motor evoked potential monitoring during resection of supratentorial brain tumors. J Neurosurg 2013;118(2):287–296 95. Seidel K, Beck J, Stieglitz L, Schucht P, Raabe A. Low-threshold monopolar motor mapping for resection of primary motor cortex tumors. Neurosurgery 2012;71(1, Suppl Operative):104–114, discussion 114–115 96. Szelényi A, Kothbauer K, de Camargo AB, Langer D, Flamm ES, Deletis V. Motor evoked potential monitoring during cerebral aneurysm surgery: technical aspects and comparison of transcranial and direct cortical stimulation. Neurosurgery 2005;57(4, Suppl):331–338, discussion 331–338 97. Szelényi A, Kothbauer KF, Deletis V. Transcranial electric stimulation for intraoperative motor evoked potential monitoring: stimulation parameters and electrode montages. Clin Neurophysiol 2007;118(7):1586–1595 98. Sartorius CJ, Berger MS. Rapid termination of intraoperative s­timulation-evoked seizures with application of cold Ringer’s lactate to the cortex. Technical note. J Neurosurg 1998;88(2):349–351 99. Landazuri P, Eccher M. Simultaneous direct cortical motor evoked potential monitoring and subcortical mapping for motor pathway preservation during brain tumor surgery: is it useful? J Clin Neurophysiol 2013;30(6):623–625 100. Cross JH, Jayakar P, Nordli D, et al; International League against Epilepsy, Subcommission for Paediatric Epilepsy Surgery. Commissions of Neurosurgery and Paediatrics. Proposed criteria

28  Localization of Eloquent Cortex and White Matter Tracts Under General Anesthesia for referral and evaluation of children for epilepsy surgery: ­recommendations of the Subcommission for Pediatric Epilepsy Surgery. Epilepsia 2006;47(6):952–959 101. Alvarez L. Cortical stimulation with subdural electrodes: special considerations in infancy and childhood. J Epilepsy 1990;3:125–130 102. Nespeca M, Wyllie E, Lüders H, et al. EEG recording and functional localization studies with subdural electrodes in infants and young children. J Epilepsy 1990;3:107–124 103. Jayakar P, Alvarez LA, Duchowny MS, Resnick TJ. A safe and effective paradigm to functionally map the cortex in childhood. J Clin Neurophysiol 1992;9(2):288–293 104. Riviello JJ, Kull L, Troup C, Holmes GL. Cortical stimulation in children: techniques and precautions. Tech Neurosurg 2001;7(1):12–18 105. Duffau H. Intraoperative cortico-subcortical stimulations in surgery of low-grade gliomas. Expert Rev Neurother 2005;5(4):473–485 106. Mandonnet E. Intraoperative Electrical Mapping: Advances, Limitations and Perspectives. Brain Mapping [Internet]. Vienna, Springer; 2011:101–108 107. Duffau H. Brain Mapping: From Neural Basis of Cognition to Surgical Applications. Vienna, Austria: Springer Science & Business Media; 2011 108. Jayakar P. Cortical electrical stimulation mapping: special considerations in children. J Clin Neurophysiol 2018;35(2):106–109 109. Kamada K, Todo T, Ota T, et al. The motor-evoked potential threshold evaluated by tractography and electrical stimulation. J Neurosurg 2009;111(4):785–795 110. Prabhu SS, Gasco J, Tummala S, Weinberg JS, Rao G. Intraoperative magnetic resonance imaging-guided tractography with integrated monopolar subcortical functional mapping for resection of brain tumors. Clinical article. J Neurosurg 2011;114(3):719–726 111. Costa P, Peretta P, Faccani G. Relevance of intraoperative D wave in spine and spinal cord surgeries. Eur Spine J 2013;22(4):840–848 112. Sarnat HB, Blümcke I. Malformations of cortical development. In: Blümcke I, Sarnat HB, Coras R, eds. Surgical Neuropathology of Focal Epilepsies: Textbook and Atlas. Paris: John Libbey Eurotext; 2015:18–53 113. Olivier A, Boling WW, Tanriverdi T. Techniques in Epilepsy Surgery: the MNI Approach. Cambridge, UK: Cambridge University Press; 2012 114. Carrabba G, Mandonnet E, Fava E, et al. Transient inhibition of motor function induced by the Cavitron ultrasonic surgical aspirator during brain mapping. Neurosurgery 2008;63(1): E178–E179, discussion E179 115. Raabe A, Beck J, Schucht P, Seidel K. Continuous dynamic mapping of the corticospinal tract during surgery of motor eloquent brain tumors: evaluation of a new method. J Neurosurg 2014;120(5):1015–1024 116. Shiban E, Krieg SM, Obermueller T, Wostrack M, Meyer B, Ringel F. Continuous subcortical motor evoked potential stimulation using the tip of an ultrasonic aspirator for the resection of motor eloquent lesions. J Neurosurg 2015;123(2):301–306 117. Thirumala PD, Crammond DJ, Loke YK, Cheng HL, Huang J, Balzer JR. Diagnostic accuracy of motor evoked potentials to detect neurological deficit during idiopathic scoliosis correction: a systematic review. J Neurosurg Spine 2017;26(3):374–383 118. Ghaly RF, Stone JL, Aldrete J. Motor evoked potentials (MEP) following transcranial magnetic stimulation in monkey anesthetized with Nitrous Oxide, Ketamine, and Thiamylal Sodium. (Abstract) Anesthesiology 1988;69:A606 119. Ghaly RF, Stone JL, Levy WJ, et al. The effect of neuroleptanalgesia (droperiodol-fentanyl) on motor potentials evoked by

transcranial magnetic stimulation in the monkey. J Neurosurg Anesthesiol 1991;3(2):117–123 120. Ghaly RF, Stone JL, Levy WJ, Kartha RK, Miles ML, Jaster HJ. The effect of etomidate or midazolam hypnotic dose on motor evoked potentials in the monkey. J Neurosurg Anesthesiol 1990;2:244 121. Ghaly RF, Stone JL, Aldrete JA, Levy WJ. Effects of incremental ketamine hydrochloride doses on motor evoked potentials (MEPs) following transcranial magnetic stimulation: a primate study. J Neurosurg Anesthesiol 1990;2(2):79–85 122. Stone JL, Ghaly RF, Levy WJ, Kartha R, Krinsky L, Roccaforte P. A comparative analysis of enflurane anesthesia on primate motor and somatosensory evoked potentials. Electroencephalogr Clin Neurophysiol 1992;84(2):180–187 123. Jones SJ, Harrison R, Koh KF, Mendoza N, Crockard HA. Motor evoked potential monitoring during spinal surgery: responses of distal limb muscles to transcranial cortical stimulation with pulse trains. Electroencephalogr Clin Neurophysiol 1996;100(5):375–383 124. Pechstein U, Cedzich C, Nadstawek J, Schramm J. Transcranial high-frequency repetitive electrical stimulation for recording myogenic motor evoked potentials with the patient under general anesthesia. Neurosurgery 1996;39(2):335–343, discussion 343–344 125. Rodi Z, Deletis V, Morota N, Vodušek DB. Motor evoked potentials during brain surgery. Pflugers Arch 1996;431(6, Suppl 2):R291–R292 126. MacDonald DB. Safety of intraoperative transcranial electrical stimulation motor evoked potential monitoring. J Clin Neurophysiol 2002;19(5):416–429 127. Tomio R, Akiyama T, Ohira T, Yoshida K. Effects of transcranial stimulating electrode montages over the head for lower-extremity transcranial motor evoked potential monitoring. J Neurosurg 2017;126(6):1951–1958 128. Mendiratta A, Emerson RG. Transcranial electrical MEP with muscle recording. Handbook of Clinical Neurophysiology. 2008;8:260–272 129. Agnew WF, McCreery DB. Considerations for safety in the use of extracranial stimulation for motor evoked potentials. Neurosurgery 1987;20(1):143–147 130. Rothwell J, Burke D, Hicks R, Stephen J, Woodforth I, Crawford M. Transcranial electrical stimulation of the motor cortex in man: further evidence for the site of activation. J Physiol 1994;481 (Pt 1):243–250 131. Katayama Y, Tsubokawa T, Maejima S, Hirayama T, Yamamoto T. Corticospinal direct response in humans: identification of the motor cortex during intracranial surgery under general anaesthesia. J Neurol Neurosurg Psychiatry 1988;51(1):50–59 132. MacDonald DB, Deletis V. Safety issues during surgical monitoring. In: Nuwer MR, ed. Intraoperative Monitoring of Neural Function. Handbook of Clinical Neurophysiology. Vol. 8. Amsterdam: Elsevier; 2008:882–898 133. Journee H, Hoving E, Mooij J. P27.4 Stimulation threshold–age relationship and improvement of muscle potentials by preconditioning transcranial stimulation in young children. Clin Neurophysiol 2006;117(Suppl 1):115 134. Erb TO, Ryhult SE, Duitmann E, Hasler C, Luetschg J, Frei FJ. Improvement of motor-evoked potentials by ketamine and spatial facilitation during spinal surgery in a young child. Anesth Analg 2005;100(6):1634–1636 135. Taniguchi M, Schramm J. Motor evoked potentials facilitated by an additional peripheral nerve stimulation. Electroencephalogr Clin Neurophysiol Suppl 1991;43:202–211 136. Calancie B, Harris W, Broton JG, Alexeeva N, Green BA. “Threshold-level” multipulse transcranial electrical stimulation of motor cortex for intraoperative monitoring of spinal motor tracts: description of method and comparison to

275

276

III  Surgical Anatomy and Mapping Techniques somatosensory evoked potential monitoring. J Neurosurg 1998;88(3):457–470

resection of tumors in eloquent motor areas. J Neurosurg 2012;116(5):994–1001

137. Abboud T, Schaper M, Dührsen L, et al. A novel threshold criterion in transcranial motor evoked potentials during surgery for gliomas close to the motor pathway. J Neurosurg 2016;125(4):795–802

148. Lefaucheur JP, Picht T. The value of preoperative functional cortical mapping using navigated TMS. Neurophysiol Clin 2016;46(2):125–133

138. Ghaly RF, Stone JL, Levy WJ. A protocol for intraoperative somatosensory (SEP) and motor evoked potentials (MEP) recordings. J Neurosurg Anesthesiol 1992;4(1):68–69 139. Chen Z. The effects of isoflurane and propofol on intraoperative neurophysiological monitoring during spinal surgery. J Clin Monit Comput 2004;18(4):303–308 140. Langeloo DD, Journée HL, De Kleuver M, Grotenhuis J. Criteria for transcranial electrical motor evoked potential monitoring during spinal deformity surgery: a review and discussion of the literature. Neurophysiol 2007;37(6):431–439 141. Sutter M, Deletis V, Dvorak J, et al. Current opinions and recommendations on multimodal intraoperative monitoring during spine surgeries. Eur Spine J 2007;16(2, Suppl 2):S232–S237 142. Gugino LD, Aglio LS, Edmonds HL, Gonzalez AA. Magnetic cortical stimulation techniques. In: Nuwer MR, ed. Intraoperative Monitoring of Neural Function. Handbook of Clinical Neurophysiology. Vol. 8. Amsterdam: Elsevier; 2008:282–318 143. Hallett M. Transcranial magnetic stimulation and the human brain. Nature 2000;406(6792):147–150 144. Tarapore PE, Picht T, Bulubas L, et al. Safety and tolerability of navigated TMS for preoperative mapping in neurosurgical patients. Clin Neurophysiol 2016;127(3):1895–1900 145. Kobayashi M, Pascual-Leone A. Transcranial magnetic stimulation in neurology. Lancet Neurol 2003;2(3):145–156 146. Wagner T, Valero-Cabre A, Pascual-Leone A. Noninvasive human brain stimulation. Annu Rev Biomed Eng 2007;9:527–565 147. Krieg SM, Shiban E, Buchmann N, et al. Utility of presurgical navigated transcranial magnetic brain stimulation for the

149. Picht T. Current and potential utility of transcranial magnetic stimulation in the diagnostics before brain tumor surgery. CNS Oncol 2014;3(4):299–310 150. Picht T, Schmidt S, Brandt S, et al. Preoperative functional mapping for rolandic brain tumor surgery: comparison of navigated transcranial magnetic stimulation to direct cortical stimulation. Neurosurgery 2011;69(3):581–588, discussion 588 151. Tarapore PE, Berger MS. Outlook on the potential of nTMS in neurosurgery. Navigated Transcranial Magnetic Stimulation in Neurosurgery. Cham, Switzerland: Springer; 2017:287–299 152. Krieg S. Navigated Transcranial Magnetic Stimulation in Neurosurgery. Cham, Switzerland: Springer; 2017:299 153. Di Iorio R, Rossini PM. Safety considerations of the use of TMS. Navigated Transcranial Magnetic Stimulation in Neurosurgery. Cham, Switzerland: Springer; 2017:67–83 154. Classen J, Witte OW, Schlaug G, Seitz RJ, Holthausen H, Benecke R. Epileptic seizures triggered directly by focal transcranial magnetic stimulation. Electroencephalogr Clin Neurophysiol 1995;94(1):19–25 155. Pascual-Leone A, Valls-Solé J, Wassermann EM, Hallett M. Responses to rapid-rate transcranial magnetic stimulation of the human motor cortex. Brain 1994;117(Pt 4):847–858 156. Chokroverty S, Hening W, Wright D, et al. Magnetic brain stimulation: safety studies. Electromyogr Clin Neurophysiol 1995;97(1):36–42 157. Goldring S, Harding GW, Gregorie EM. Distinctive electrophysiological characteristics of functionally discrete brain areas: a tenable approach to functional localization. J Neurosurg 1994;80(4):701–709

29

  Cortical Stimulation and Mapping Doris D. Wang, John D. Rolston, and Mitchel S. Berger

Summary In pediatric patients with medically refractory epilepsy, the goal of resective surgery is to remove the epileptogenic zone without causing postoperative deficits. Cortical and subcortical stimulation mapping is the gold standard for identification of functionally eloquent areas of the brain. In this chapter, we describe the indication and techniques of cortical and subcortical mapping for identification of language and motor areas during pediatric epilepsy surgery. We also discuss ways to avoid complications when using these techniques. Our goal is to provide a framework for neurosurgeons to safely utilize these tools while performing resective epilepsy surgery in eloquent areas. Keywords:  motor mapping, language mapping, intraoperative mapping, functional mapping, mapping technique, c­ ortical stimulation, subcortical stimulation, pediatric epilepsy, ­anesthesia

„„ Introduction In pediatric patients with medically refractory epilepsy, the goal of resective surgery is to remove the epileptogenic zone without causing postoperative deficits. Mapping of language and sensorimotor functions is the gold standard to achieve maximal safe resection for lesions in eloquent areas. Since the development of language and motor mapping techniques by Penfield in the context of epilepsy surgeries,1 modern advances have greatly facilitated effective resective surgeries in the eloquent cortex with reduced morbidity.2 Using methods of direct cortical and subcortical stimulation, Berger and colleagues were the first to report the use of intraoperative mapping techniques to localize language, sensorimotor pathways, and seizure foci in children with supratentorial brain tumor.3 This seminal work demonstrated that physiological mapping is safe, reliable, and valuable in identification of eloquent brain regions. Mapping allows for maximal tumor resection and eradication of epileptogenic zones while minimizing morbidity in the pediatric population (Fig. 29.1). The safety and efficacy of functional mapping in children has since been replicated in many reports describing its use in tumor resection as well as epilepsy surgery.4,​5,​6,​7 Physiological mapping is important to delineate the exact locations of essential cortical areas given the high intersubject

variability based on traditional maps.8,​9,​10 While functional imaging has made remarkable advances, functional magnetic resonance imaging (fMRI) is based on identifying cortical regions activated during specific tasks, and therefore cannot differentiate between essential and nonessential regions of the brain.11,​12,​13 Similarly, noninvasive mapping techniques using magnetoencephalography (MEG) and transcranial magnetic stimulation (TMS) may help in preoperative surgical evaluation. However, the sensitivity and specificity of these techniques remain inadequate to substitute intraoperative cortical stimulation mapping, and cannot offer real-time intraoperative information.14,​15 Diffusion tensor imaging (DTI) has been developed to determine subcortical anatomy, but it does not provide any functional detail.16,​17 Consequently, the use of intraoperative cortical and subcortical stimulation to accurately delineate functional regions and pathways is critical for safely resecting epileptic foci or lesions located near eloquent brain areas. In this chapter, we describe the indication and techniques of cortical and subcortical mapping for identification of language and motor areas during pediatric epilepsy surgery. We also discuss ways to avoid complications when using these techniques. Our goal is to provide a framework for neurosurgeons to safely utilize these tools while performing resective epilepsy surgery in eloquent areas.

„„ Language Mapping Indications for Language Mapping Preservation of language function must be balanced with the goal of maximal resection of epileptic focus or lesion, and therefore some argue that intraoperative language mapping for resection in the dominant hemisphere should be the rule, rather than the exception.18 The epileptic foci often reside within and are continuous with functional tissue. Given the significant variability in the anatomical and functional organization of the region, patients with epileptic focus or lesions located within and in proximity to any of the language pathways should undergo awake intraoperative mapping. These language areas typically comprise of cortical regions of the superior and middle temporal gyrus, pars opercularis, pars triangularis, posterior inferior frontal gyrus, as well as subcortical pathways involving

278

III  Surgical Anatomy and Mapping Techniques

Fig. 29.1  Preoperative axial fluid-attenuated inversion recovery (FLAIR) MRI of a 17-year-old adolescent boy who presented with seizures showing an expansile nonenhancing mass centered in the left parietal operculum (a) with extension into the insula (b). Postoperative axial FLAIR MRI (c, d) from the same patient showing near-total resection of the mass after successful awake intraoperative motor and language mapping.

the superior longitudinal fasciculus, the arcuate fasciculus, the uncinated fasciculus, and the inferior fronto-occipital fasciculus.9,​10,​19,​20,​21,​22 Intraoperative language mapping for language requires an awake, cooperative patient with relatively preserved baseline language functions, and can only be performed in the adolescent patient group. It is more difficult to identify language cortex in younger children (< 10 years of age), and they usually require placement of subdural electrodes in order to perform extraoperative functional mapping.23,​24 With adequate preoperative preparation, children aging from 11 to 17 can undergo successful awake craniotomy for resection of tumors or epileptic focus with language mapping with good psychological experience and surgical outcome.6,​7,​25 Some other relative contraindications for awake language mapping include large tumors causing cerebral edema, obesity, and sleep apnea causing airway concerns.26

Anesthesia for Language Mapping Specialized neuroanesthesia is critical for the success of awake craniotomy surgery. In our experience, patients are premedicated with midazolam and an arterial line, temperature probe and Foley’s catheter are placed prior to positioning. The surgeon injects local anesthetic consisting of a 1:1 mixture of 0.5% bupivacaine and 1% lidocaine with 1:100,000 epinephrine

to the scalp prior to Mayfield’s pinning.27 Sedation is achieved with propofol (up to 100 µg/kg/min) and remifentanil (0.05–2.0 µg/kg/min).26 Propofol and remifentanil boluses are used during Foley’s placement and Mayfield’s application. Anesthesia is maintained until craniotomy is complete, at which point all sedatives are discontinued and the patient is asked to hyperventilate prior to dura opening. The temporalis muscle and dura are then infiltrated with lidocaine around the middle meningeal artery to further decrease discomfort. No sedatives are administered during mapping. Propofol and topical ice-cold Ringer’s solutions are available for seizure suppression in case of stimulation-induced seizures. Once mapping is complete, sedation is achieved with dexmedetomidine (up to 1 µg/kg/min) and remifentanil (0.05 µg/kg/min and higher).26,​27 However, if subcortical mapping is to be done later, anesthesia will need to be periodically stopped or withheld during ­resection. Laryngeal mask airways (LMAs) may be held in reserve in the case of apnea or difficulty maintaining a patent airway. Nasal trumpets or oral airways can be used during the craniotomy portion or during tumor resection to maintain an open airway and to assist with ventilation as necessary.26

Language Mapping Technique Baseline language testing is usually performed 24 to 48 hours prior to surgery and included object naming, action naming,

29  Cortical Stimulation and Mapping

Fig. 29.2  Intraoperative photograph demonstrating placement of sterile numbered markers over the exposed cortical surface. Cortical stimulation up to 5 mA was applied at each cortical site during language tasks (picture naming and sentence production) using the Ojemann stimulator. Motor and sensory mapping were also performed at these sites. In this patient, cortical mapping revealed mouth sensory functions at sites 7 and 8. No positive motor or language sites were identified. Subcortical language and motor mapping were carried out during resection of the deeper portions of the tumor. SF, sylvian fissure.

reading, counting, and comprehension testing. The full naming battery is then modified to remove objects, words, and tasks that the patient has difficulty performing. Intraoperative testing only uses pictures and words that the patient is able to reliably answer correctly based on perioperative evaluation. We use a bipolar electrode (1-mm bipolar electrodes with 5-mm tip separation) and a constant-current stimulator to deliver biphasic square wave pulses (1.25-ms pulse wide) in 4-second trains at 60 Hz.26 We typically start cortical mapping at a low stimulus (1.5 mA) and increase to a maximum of 6 mA. Sterile numbered marks (~ 10–20 per subject) are placed over the exposed cortical surface (Fig. 29.2). Cortical stimulation is applied for 1 to 2 seconds at sequential cortical sites during a language task (object or action naming from slide presentation of line drawing, reading words, or counting). All tested language sites should be repeatedly stimulated for at least three times. A positive essential site is defined as a production of language errors in 66% or greater of the testing per site.27 Language deficits induced by electrical stimulation typically include speech arrest, anomia, and paraphasia.28 Speech arrest, or anarthria, occurs when the patient could not retrieve the coding of motor programs of speech while retaining the ability to move targeted muscles with visible articulatory effort.20 Anomia is the failure to generate the correct name for the object (picture naming) or verb (action naming) in the absence of damage to perceptual or articulatory mechanisms (i.e., still able to produce a sentence “this is a…”). Paraphasic error can be phonemic (“sucky” instead of “sunny”) or semantic (“chair” instead of “table”).28,​29 Continuous electrocorticography is performed using a crown to monitor afterdischarge potentials, which distinguish positive language sites from behavior changes caused by subclinical seizure activity. Some groups argue for lower ­stimulation (up

to 3 mA) to reduce risks of afterdischarges and seizures when performing mapping without electrocorticography.30 We recommend performing electrophysiological recordings to detect all afterdischarges. The cortex is deemed nonfunctional only if no behavior response is seen in the absence of afterdischarges after stimulation intensity of 6 mA is reached.31 In the cortex, a 1-cm margin of tissue should be preserved around each positive language site to protect functional tissue from the resection,27 though some suggest that it is possible to leave no margin with a higher incidence of transient postoperative deficit.32 With subcortical mapping, a 5-mm margin of tissue can be resected if no nearby fibers are detected with mapping.17

„„ Motor Mapping Indications and Contraindications for Motor Mapping Cortical and subcortical stimulation mapping allows for precise delimitation of the cortical and subcortical motor system. Knowing the boundaries of the motor system allows the surgeon to resect the maximal amount of pathological tissue while preserving eloquent motor cortex.9 While helpful, preoperative imaging modalities such as DTI, fMRI, MEG, and TMS are unable to localize the motor system with the precision of stimulation ­mapping.15 Thus, intraoperative stimulation mapping remains the gold standard for mapping and preserving eloquent cortex. The primary contraindication to motor mapping is severe hemiparesis or hemiplegia, in which case motor responses cannot be reliably evoked and monitored. This is true whether the mapping will be done “asleep,” under general anesthesia (GA),

279

280

III  Surgical Anatomy and Mapping Techniques or “awake,” with monitored anesthesia care (MAC). A relative contraindication is young age (~ 5 years or younger), when the brain is less excitable.33 In pediatric patients, the threshold for evoking motor responses decreases linearly with age, starting from a high intensity of approximately 20 mA at 5 to 7 years of age, but decreasing to adult levels during the teenage years.24,​33 At very young ages, there is difficulty evoking motor responses without triggering epileptiform afterdischarges, which has led most practitioners to avoid testing these patients.4,​24,​33

Awake versus Asleep Motor Mapping Unlike language mapping, motor mapping can be conducted with the patient either asleep (GA) or awake (minimum alveolar concentration, MAC). Awake mapping is generally considered more sensitive than asleep mapping, and additionally permits the testing of voluntary movements.34,​35 Regarding voluntary movements, at least one group has shown changes in evoked motor responses without commensurate changes in voluntary movement and vice versa (degradation of voluntary movements without changes in evoked motor potentials).35 Such case studies attest to the potential benefits of awake mapping. For patients who would benefit from the added sensitivity of awake mapping, but cannot tolerate a craniotomy under MAC, an alternative is the implantation of subdural electrodes with extraoperative motor mapping at a later time. For instance, for young children unable to participate in an awake case, an electrode grid could be implanted under GA, followed by mapping at the patient’s bedside one or more days later.

Anesthesia for Awake Motor Mapping Neuroanesthesia for awake motor mapping is similar to awake language mapping. The patient is sedated with a short-acting opioid (like remifentanil) combined with propofol or dexmedetomidine (and sometimes a combination of both).36 Inhaled anesthetics are generally avoided but have been reported by some institutions.37 A field block around the incision is created with a 1:1 mixture of 0.5% bupivacaine and 1% lidocaine with 1:100,000 epinephrine.26 Deep anesthesia is maintained until the craniotomy is complete, at which point sedation is stopped to permit mapping. Anesthesia can be restarted once the mapping is complete to permit a more comfortable resection of the lesion. Anesthesia can be periodically stopped or withheld if additional subcortical mapping is to be done during the case. As in language mapping cases, nasal trumpets, oral airways, and LMAs can be used during the craniotomy portion or during tumor resection to maintain an open airway and to assist with ventilation as necessary.

Anesthesia for Asleep Motor Mapping The anesthesia regimen for asleep mapping is similar to awake mapping. Short-acting opioids such as remifentanil, supplemented with propofol or dexmedetomidine are very common.38,​39 Almost all single agents at high doses suppress motor responses, which is why combinations are typical. During asleep mapping, inhalational anesthetics should be avoided except at low doses.40 While motor responses can still

be generated at up to 0.5 MAC, the suppressive effects can be difficult to overcome even at 0.25 MAC.40 Importantly, nitrous oxide seems to be the least suppressive of the inhalational agents.41 Propofol is often used in mapping cases, but suppresses motor responses by approximately 30 to 60% at levels of 1 to 2 μg/mL. Increasing concentrations can completely suppress responses. If propofol is used, a target level of less than 1 μg/mL appears safest.40 Other agents, such as ketamine, opioids, benzodiazepines, and etomidate, have minimal effect on motor responses when used at nonbolus levels. Therefore, the most common regimen tends to be short-acting opioids with propofol or dexmedetomidine. Muscle relaxants may be used during induction and pinning, but must be stopped or markedly reduced prior to mapping. If their use is continued during mapping, the levels must be kept constant, and a twitch monitor used to ensure a fixed level.38

Importance of Temperature and Other Physiological Parameters Particularly during asleep craniotomies, maintaining normal body temperature is important. Hypothermia, starting at less than 35°C, can increase motor response latency and decrease amplitude, or suppress the response completely. The same is true for hyperthermia, beginning at greater than 38°C.40 Other physiological parameters, such as blood pressure, oxygen saturation, and CO2 levels, can also affect motor responses. Hypoxia and hypotension should be avoided. Hypercapnia at levels of greater than 70 mm Hg may cause problems with both mapping and cerebral edema.42 Hypocapnia, which is typical for cranial surgery, appears to have no effect on motor responses at levels ranging from 13 to 30 mm Hg.40

Motor Mapping Technique Stimulation is carried out with a bipolar electrode (5-mm tip separation) or a monopolar ball probe (2–3 mm diameter tip) and a constant-current stimulator (Ojemann Cortical Stimulator; Integra LifeSciences Corporation, Saint Priest, France). The bipolar electrode configuration produces more constrained electric fields than the monopolar tip, and produces its highest currents between the two electrodes. Monopolar stimulation, which uses a remote ground electrode, produces a more homogeneous electric field, which decays approximately uniformly in a radial direction (actual currents are determined by tissue impedance, which varies in complex ways throughout the brain). Monopolar stimulation is therefore best suited to determine proximity to subcortical pathways during resection, while bipolar stimulation is best suited to mapping surface function. Stimulation trains consist of biphasic square pulses, typically delivered at 60 Hz. Pulses are typically 1 ms in duration (0.5 ms per phase). With the Ojemann stimulator, trains are initiated by contact with the tissue. Typical lengths are approximately 1 to 5 seconds. For awake patients, mapping usually starts at 2 mA and increases by 0.5 to 1.0 mA until a response is elicited or long trains of afterdischarges are observed (typical range is 2–5 mA). Asleep cases typically require higher-current intensities, 4 to 16 mA. Alternative stimulation train parameters

29  Cortical Stimulation and Mapping (e.g., frequency, duration, etc.) have been reported, with varying advantages and disadvantages.43 Each cortical site is tested three times. Positive sites are those that evoke motor responses in two out of three trials. During stimulation, the patient’s contralateral limbs and face should be exposed and visible to operating room (OR) staff, so that movements can be readily observed. In asleep cases, neurophysiologists can monitor electromyography (EMG) with inserted needle electrodes to detect evoked responses more reliably than by direct vision.44 For awake patients, sticker electrodes can be used in lieu of penetrating needle electrodes. In one report, EMG showed positive responses when no visual response was seen in 30% of surgeries.44 Positive sites are marked by placing small sterile tickets (numbered or lettered) as they are found. Alternatively, the numbers can be placed prior to mapping in a grid formation, with positive and negative sites recorded as the case proceeds (Fig. 29.2). How close can one resect tissue to a positive motor site? Multiple studies have examined positive mapping sites in relation to DTI-defined motor tracts, and the consensus appears that stimulation in the range of approximately 5 mA can activate tracts approximately 5 mm distant.16,​17,​45,​46 Positive cortical stimulation sites should therefore be considered to expand by at least this radius (5 mm) from the stimulation probe (noting that the bipolar probe itself is 5-mm wide, leading to a diameter of 1.5 cm around the stimulation site, which is consistent with language mapping results8). For subcortical mapping, the process is somewhat different. The probe is used to map the floor or wall of a resection cavity to test for nearby fibers. If nothing is detected, then the above results suggest that another 5 mm can be resected safely.17

„„ Complications Avoidance Seizures Seizures are provoked by stimulation mapping in approximately 1.5% of patients, regardless of whether they have preexisting epilepsy or not.47 Younger children require higher stimulation threshold (up to 20 mA) to elicit a functional response, likely due to insufficient myelination, and higher stimulation may provoke seizures.23,​24,​33 If symptomatic seizures are triggered, the brain should be irrigated with ice-cold irrigation and propofol or another anesthetic should be administered. Before ­mapping begins, the surgeon should verify that ice-cold irrigation is available and that the anesthesiologist has propofol or another agent in line and ready for delivery.

Registration Especially during awake craniotomies or following generalized seizures, there is a chance that the patient will move while in pins and perturb the navigation’s registration. For this reason,

it is advisable to capture additional skull-based fiducial marks after the scalp is opened but before the mapping begins. These marks can be a series of screws or small divots created by a high-speed burr.

Patient Intolerance One limitation of awake mapping is the difficulty of children to cooperate. Preoperative preparation is crucial to set patient expectations and reduce anxiety. We also have had success with having a child’s parents into the operating room during the mapping portion of the case. In addition, time is a major challenge of intraoperative mapping as patient may get fatigued after 2 hours of testing. Therefore, efficient language and motor tests should be employed to maximize information obtained to help with determining the resection.

Anesthesia During awake craniotomies, particularly in obese patients, there is a chance that anesthesia may make the patient apneic. In these cases, an LMA can be used to emergently intubate the patient, securing his or her airway and ventilating while the anesthesia is weaned off.

„„ Conclusion Cortical and subcortical stimulation mapping is the gold standard for identification of functionally eloquent areas of the brain. Neurosurgeons should use cortical and subcortical mapping in all circumstances when operating near the peri-Rolandic, perisylvian regions or near subcortical systems known to be involved in language or motor processing. Functional mapping has been shown to both significantly reduce late-stage postoperative neurological deficits and increase the proportion of gross total resections.48 While functional neuroimaging using fMRI, MEG, TMS, and DTI has evolved to provide useful supplementary information regarding the localization of eloquent areas and pathways, their sensitivity and specificity are not adequate to replace intraoperative direct cortical and subcortical mapping.14,​15 Intraoperative mapping is essential in providing functional information that ultimately inform the boundaries of resection. Using intraoperative mapping techniques, lesions located in the most highly eloquent areas that may not have been considered resectable, but are functionally silent from mapping, can be resected with an acceptable rate postoperative morbidity.49 With a multidisciplinary team approach among neurosurgeons, anesthesia, and neurophysiologists, cortical and subcortical mapping can be performed reliably and safely to maximizing resection of epileptic lesion in pediatric patients.

References 1. Penfield W. Epilepsy and the cerebral lesions of birth and infancy. Can Med Assoc J 1939;41(6):527–534 2. De Witt Hamer PC, Robles SG, Zwinderman AH, Duffau H, Berger MS. Impact of intraoperative stimulation brain mapping on glioma surgery outcome: a meta-analysis. J Clin Oncol 2012;30(20):2559–2565

3. Berger MS, Kincaid J, Ojemann GA, Lettich E. Brain mapping techniques to maximize resection, safety, and seizure control in children with brain tumors. Neurosurgery 1989;25(5):786–792 4. Jayakar P, Alvarez LA, Duchowny MS, Resnick TJ. A safe and effective paradigm to functionally map the cortex in childhood. J Clin Neurophysiol 1992;9(2):288–293

281

282

III  Surgical Anatomy and Mapping Techniques 5. Sala F, Krzan MJ, Deletis V. Intraoperative neurophysiological monitoring in pediatric neurosurgery: why, when, how? Childs Nerv Syst 2002;18(6–7):264–287 6. Balogun JA, Khan OH, Taylor M, et al. Pediatric awake craniotomy and intra-operative stimulation mapping. J Clin Neurosci 2014;21(11):1891–1894 7. Delion M, Terminassian A, Lehousse T, et al. Specificities of awake craniotomy and brain mapping in children for resection of supratentorial tumors in the language area. World Neurosurg 2015;84(6):1645–1652 8. Ojemann G, Ojemann J, Lettich E, Berger M. Cortical language localization in left, dominant hemisphere. An electrical stimulation mapping investigation in 117 patients. J Neurosurg 1989;71(3):316–326 9. Sanai N, Mirzadeh Z, Berger MS. Functional outcome after langu­age mapping for glioma resection. N Engl J Med 2008; 358(1):18–27 10. Chang EF, Breshears JD, Raygor KP, Lau D, Molinaro AM, Berger MS. Stereotactic probability and variability of speech arrest and anomia sites during stimulation mapping of the language dominant hemisphere. J Neurosurg 2017;126(1):114–121 11. Roux FE, Boulanouar K, Lotterie JA, Mejdoubi M, LeSage JP, Berry I. Language functional magnetic resonance imaging in preoperative assessment of language areas: correlation with direct cortical stimulation. Neurosurgery 2003;52(6):1335– 1345, discussion 1345–1347 12. Petrovich N, Holodny AI, Tabar V, et al. Discordance between functional magnetic resonance imaging during silent speech tasks and intraoperative speech arrest. J Neurosurg 2005;103(2):267–274 13. Spena G, Nava A, Cassini F, et al. Preoperative and intraoperative brain mapping for the resection of eloquent-area tumors. A prospective analysis of methodology, correlation, and usefulness based on clinical outcomes. Acta Neurochir (Wien) 2010;152(11):1835–1846 14. Ille S, Sollmann N, Hauck T, et al. Combined noninvasive language mapping by navigated transcranial magnetic stimulation and functional MRI and its comparison with direct cortical stimulation. J Neurosurg 2015;123(1):212–225 15. Ottenhausen M, Krieg SM, Meyer B, Ringel F. Functional preoperative and intraoperative mapping and monitoring: increasing safety and efficacy in glioma surgery. Neurosurg Focus 2015;38(1):E3 16. Bello L, Gambini A, Castellano A, et al. Motor and language DTI fiber tracking combined with intraoperative subcortical mapping for surgical removal of gliomas. Neuroimage 2008;39(1):369–382 17. Leclercq D, Duffau H, Delmaire C, et al. Comparison of diffusion tensor imaging tractography of language tracts and intraoperative subcortical stimulations. J Neurosurg 2010;112(3):503–511 18. Taylor MD, Bernstein M. Awake craniotomy with brain mapping as the routine surgical approach to treating patients with supratentorial intraaxial tumors: a prospective trial of 200 cases. J Neurosurg 1999;90(1):35–41 19. Chang EF, Raygor KP, Berger MS. Contemporary model of language organization: an overview for neurosurgeons. J Neurosurg 2015;122(2):250–261 20. Mandonnet E, Sarubbo S, Duffau H. Proposal of an optimized strategy for intraoperative testing of speech and language during awake mapping. Neurosurg Rev 2017;40(1):29–35 21. Tate MC, Herbet G, Moritz-Gasser S, Tate JE, Duffau H. Probabilistic map of critical functional regions of the human cerebral cortex: Broca’s area revisited. Brain 2014;137(Pt 10):2773–2782 22. Kilbride RD. Intraoperative functional cortical mapping of language. J Clin Neurophysiol 2013;30(6):591–596 23. Schevon CA, Carlson C, Zaroff CM, et al. Pediatric language mapping: sensitivity of neurostimulation and Wada testing in epilepsy surgery. Epilepsia 2007;48(3):539–545 24. Gallentine WB, Mikati MA. Intraoperative electrocorticography and cortical stimulation in children. J Clin Neurophysiol 2009;26(2):95–108

25. Soriano SG, Eldredge EA, Wang FK, et al. The effect of propofol on intraoperative electrocorticography and cortical stimulation during awake craniotomies in children. Paediatr Anaesth 2000;10(1):29–34 26. Hervey-Jumper SL, Berger MS. Technical nuances of awake brain tumor surgery and the role of maximum safe resection. J Neurosurg Sci 2015;59(4):351–360 27. Sanai N, Berger MS. Operative techniques for gliomas and the value of extent of resection. Neurotherapeutics 2009;6(3):478–486 28. Rofes A, Miceli G. Language mapping with verbs and sentences in awake surgery: a review. Neuropsychol Rev 2014;24(2):185–199 29. Ojemann G, Mateer C. Human language cortex: localization of memory, syntax, and sequential motor-phoneme identification systems. Science 1979;205(4413):1401–1403 30. Boetto J, Bertram L, Moulinié G, Herbet G, Moritz-Gasser S, Duffau H. Electrocorticography is not necessary during awake brain surgery for gliomas. World Neurosurg 2016;91:656–657 31. Sanai N, Berger MS. Mapping the horizon: techniques to optimize tumor resection before and during surgery. Clin Neurosurg 2008;55:14–19 32. Gil-Robles S, Duffau H. Surgical management of World Health Organization Grade II gliomas in eloquent areas: the necessity of preserving a margin around functional structures. Neurosurg Focus 2010;28(2):E8 33. Chitoku S, Otsubo H, Harada Y, et al. Extraoperative cortical stimulation of motor function in children. Pediatr Neurol 2001;24(5):344–350 34. Eseonu CI, Rincon-Torroella J, ReFaey K, et al. Awake craniotomy vs craniotomy under general anesthesia for perirolandic gliomas: evaluating perioperative complications and extent of resection. Neurosurgery 2017;81(3):481–489 35. Suzuki K, Mikami T, Sugino T, et al. Discrepancy between voluntary movement and motor-evoked potentials in evaluation of motor function during clipping of anterior circulation aneurysms. World Neurosurg 2014;82(6):e739–e745 36. Meng L, Berger MS, Gelb AW. The potential benefits of awake craniotomy for brain tumor resection: an anesthesiologist’s perspective. J Neurosurg Anesthesiol 2015;27(4):310–317 37. Peruzzi P, Bergese SD, Viloria A, Puente EG, Abdel-Rasoul M, Chiocca EA. A retrospective cohort-matched comparison of conscious sedation versus general anesthesia for supratentorial glioma resection. Clinical article. J Neurosurg 2011;114(3):633–639 38. Guo L, Gelb AW. The use of motor evoked potential monitoring during cerebral aneurysm surgery to predict pure motor deficits due to subcortical ischemia. Clin Neurophysiol 2011;122(4):648–655 39. Pechstein U, Nadstawek J, Zentner J, Schramm J. Isoflurane plus nitrous oxide versus propofol for recording of motor evoked potentials after high frequency repetitive electrical stimulation. Electroencephalogr Clin Neurophysiol 1998;108(2):175–181 40. Lotto ML, Banoub M, Schubert A. Effects of anesthetic agents and physiologic changes on intraoperative motor evoked potentials. J Neurosurg Anesthesiol 2004;16(1):32–42 41. Ubags LH, Kalkman CJ, Been HD, Drummond JC. Differential effects of nitrous oxide and propofol on myogenic transcranial motor evoked responses during sufentanil anaesthesia. Br J Anaesth 1997;79(5):590–594 42. Short LH, Peterson RE, Mongan PD. Physiologic and anesthetic alterations on spinal-sciatic evoked responses in swine. Anesth Analg 1993;76(2):259–265 43. Tate MC, Guo L, McEvoy J, Chang EF. Safety and efficacy of motor mapping utilizing short pulse train direct cortical stimulation. Stereotact Funct Neurosurg 2013;91(6):379–385 44. Yingling CD, Ojemann S, Dodson B, Harrington MJ, Berger MS. Identification of motor pathways during tumor surgery facilitated by multichannel electromyographic recording. J Neurosurg 1999;91(6):922–927

29  Cortical Stimulation and Mapping 45. Kamada K, Todo T, Ota T, et al. The motor-evoked potential threshold evaluated by tractography and electrical stimulation. J Neurosurg 2009;111(4):785–795 46. Ozawa N, Muragaki Y, Nakamura R, Iseki H. Identification of the pyramidal tract by neuronavigation based on intraoperative diffusion-weighted imaging combined with subcortical stimulation. Stereotact Funct Neurosurg 2009;87(1):18–24 47. Szelényi A, Joksimovic B, Seifert V. Intraoperative risk of seizures associated with transient direct cortical stimulation

in patients with symptomatic epilepsy. J Clin Neurophysiol 2007;24(1):39–43 48. Duffau H, Moritz-Gasser S, Mandonnet E. A re-examination of neural basis of language processing: proposal of a dynamic hodotopical model from data provided by brain stimulation mapping during picture naming. Brain Lang 2014;131:1–10 49. Krieg SM, Schnurbus L, Shiban E, et al. Surgery of highly eloquent gliomas primarily assessed as non-resectable: risks and benefits in a cohort study. BMC Cancer 2013;13:51

283

30

  Subcortical Mapping During Intracranial Surgery in Children Francesco Sala and Davide Giampiccolo

Summary While resection or disconnection of an epileptogenic lesion may lead to the cessation of seizures, damage to essential networks produces permanent deficits. As current perspectives in neurosurgery have moved from strict cortical localizationism to hodology, intraoperative recognition and preservation of essential cortical–subcortical structures have become a crucial goal for surgical resection. Although the resurgence of awake surgery and the advent of tractography have allowed for unprecedented progression in knowledge and treatment of lesions involving subcortical white matter structures, other strategies may have to be applied in conditions of scarce compliance, such as surgery in pediatric patients. In this scenario, intraoperative neurophysiological monitoring (IOMN) may be the strategy of choice, allowing for the identification of functionally defined but anatomically ambiguous eloquent areas and respective white matter connectivity. Here, we review current advances and future perspectives in mapping subcortical networks in children. Keywords:  epilepsy surgery, neurophysiology, pediatric patients, brain mapping, motor evoked potentials

„„ Introduction IONM is a well-established discipline, aimed not merely to predict but also to prevent neurological injury during neurosurgical procedures. It is based on the use of mapping and monitoring techniques. Mapping techniques identify anatomically ambiguous neural structures, while monitoring techniques assess the functional integrity of sensory, motor, visual, and auditory pathways. Some of these techniques are particularly valuable during supratentorial surgery to localize eloquent cortex and to identify and preserve subcortical connectivity. Whether the goal is a lesionectomy or, in the case of nonlesional epilepsy surgery, to remove the epileptogenic zone or to disrupt the spreading of seizure activity, it is imperative to minimize morbidity. To maintain the integrity of neurological functions is a prerequisite to warrant preservation of quality of life, and this is becoming more and more priority in both tumor and epilepsy surgery, considering the increase in survival rates. While motor cortex can be identified also during asleep surgery, language and other cognitive functions can be assessed

intraoperatively only during awake craniotomy. However, this could be challenging in young and/or scarcely collaborating children and most of the experience in awake craniotomy is limited to adolescents and children older than 12 to 14 years of age.1,​2,​3 Whenever awake craniotomy is not an option, the traditional approach (especially in epilepsy surgery) is a two-stage operation, with grid implantation and extraoperative neurophysiological evaluation before the resection procedure. This latter strategy has the drawbacks of requiring two operations and the risks connected with implanted grid electrodes,4 but also the advantages of extended time for mapping cortical function, as compared to the intraoperative setting.5 However, cortical mapping, regardless the intra- or extraoperative setting, does not assess subcortical pathways and connectivity. In the past, most of the interest and research was focused on cortical—rather than subcortical—mapping, due to the classical concept of eloquent cortex, in which location of function was reputed to be mostly located in specific cortical regions. Broca’s area, Wernicke’s area, and the precentral gyrus (Brodmann’s area 4) were considered “no entry zones,” to the extent that P ­enfield wrote “in the therapeutic approach, it should be pointed out that only very rarely has the Rolandic area been included in any excision and never has this region of the brain been touched unless a lesion was present that could be demonstrated grossly by operative inspection. This digression is made in the hope of discouraging surgical removal of normal brain from the Rolandic area, or elsewhere, whatever may have been the pattern of epileptic seizure.”6 However, the large interindividual variability of brain functional organization7 as well as a lack of consistent clinical presentations following damage to the same cortical region challenged this rigid cortical functional organization concept. Although theories of higher brain function as associative connections of cortical areas were initially proposed by Karl Wernicke, Norman Geschwind’s disconnection syndromes in animals and man8 provided a model by which brain function would be constituted by spatially distributed dynamic networks involving different subcortical pathways connecting distant cortices; this theory was indeed supported by classical description of disconnection syndromes.9 In a brain composed of localized but connected specialized areas, disconnection leads to dysfunction,10 and loss of function subsequent to white matter damage may be irreversible since damage to tracts that connect to an eloquent area may cause as much deficit as damage to the area itself.11 This approach has refreshed the concept of hodotopy, where the human brain is seen as an “integrated, wide, plastic network

30  Subcortical Mapping During Intracranial Surgery in Children made up of cortical functional epicenters,” “topic organization,” connected by both short-local and large-scale white matter fibers,12 moving the attention from the cortex to the subcortical anatomy. This is of particular importance, since gray matter and white matter do not show the same degree of plasticity. Herbet et al11 demonstrated that the cortex possesses high capability for reorganization after damage, while conversely, white matter tract plasticity is low. Evidence has been obtained that when maintaining subcortical integrity of white matter tracts, it is possible to perform surgical resection of eloquent areas such as Broca’s and Wernicke’s area without incurring irreversible functional deficits.13,​14 As a result, current approaches in supratentorial surgery use cortical stimulation to recognize surgical entry points, while subcortical stimulation can be used for maintaining function through preservation of the integrity of subcortical white matter tracts. The resurgence of awake surgery and the advent of tractography, a postacquisition derivation of MR diffusion imaging in which the main direction of water molecules is used to infer white matter structure, have offered new insights on subcortical connectivity. Thanks to the unique opportunity to identify white matter tracts in presurgical functional planning using diffusion imaging tractography15 and to test these tracts intraoperatively using subcortical stimulation.16 From this perspective, intraoperative neurophysiological subcortical mapping enables the opportunity for investigating the functional role of specific subcortical networks. As much as IONM techniques in pediatric neurosurgery are largely based on the same techniques used in adult patients, the immaturity of the motor system in young children requires some adjustments of these techniques. Studies of human corticospinal tract (CST) development have shown that CST axons reach the medulla by 8-week postconceptional age (PCA), and the lower cervical spinal cord by 24-week PCA.17,​18 Corticospinal connections reach the sacral levels between 18 and 28 weeks’ PCA and are completed at birth.19,​20 There is, however, a discrepancy between the anatomical and neurophysiological development of the motor pathways, as the neurophysiological maturation of the CST progresses throughout childhood and adolescence.21 Moreover, different white matter tracts may show different patterns of maturation during childhood and adolescence. In Paus et al,22 computational analysis of structural magnetic resonance images obtained in 111 children and adolescents revealed age-related increases in white matter density in fiber tracts for putative corticospinal and frontotemporal pathways. While the maturation of the corticospinal tract was bilateral, the frontotemporal pathway was found predominantly in the left hemisphere, providing evidence of different maturation of presumably motor and language pathways. This is supported also by Lebel et al,23 in which the arcuate fasciculus would develop more slowly than the other connections and could correlate with Schevon et al,24 in which language cortex failed to be recognized with functional mapping in patients younger than 10 years old. In a developing brain and in accordance with the different stages of white matter maturation, a tailored approach is needed whenever IONM techniques are applied to pediatric patients, especially infants. For example, in adults, the motor cortex is usually located around 45 to 50 mm behind the coronal suture in the midline. However, in 2004 a study by Rivet

et al,25 documented that M1 in young children is displaced more ventrally, and in infants under the age of 3 years it can be as close as just 20 mm behind the coronal suture. This should guide the positioning of cortical stimulating electrodes in babies. Moreover, pediatric neurosurgeons should be aware that a craniotomy exposing the first 2 to 3 cm behind the coronal suture may well include the motor strip in young children, with all the related implications in terms of surgical approach. In this chapter, we will briefly address the basics of ­tractography—the neuroimaging technique playing a major role in subcortical mapping—and we will then review some of the principles of cortical mapping, while focusing more specifically on subcortical mapping and monitoring of motor evoked ­potentials during brain surgery in children. While the vast majority of the authors’ experience is in the field of brain tumor s­urgery,26,​27,​28 the same principles can be applied to epilepsy surgery.

„„ Tractography In 1994, Basser et al29 first described MR diffusion tensor imaging measuring the anisotropic diffusion of water molecules in a pork loin, although diffusion MR were already integrated in common clinical practice for the detection of ischemic stroke. In this seminal paper, they showed that if diffusion is measured along at least six different directions, it is possible to obtain a mathematical description of a diffusion tensor, an ellipsoid that accounts for the overall displacement of water molecules. In diffusion MR, the signal is usually sensitized to the displacement of water molecules along a selected direction. If the composition of the tissue is isotropic (i.e., its physical properties are identical in all directions), the diffusion of water molecules is reduced equally along all orientations assuming the shape of a sphere. If the diffusion of water is anisotropic, water diffuses along a particular direction, as in the case of the muscle fibers in the pork loin or equally in white matter of the brain and the spinal cord, since myelination in the axon will force water to move along a major axis, which is the orientation of the white matter bundle. In diffusion tensor imaging, it is assumed that fibers in a voxel follow a single direction, hence diffusion tensor imaging cannot solve fiber crossing. This is an important drawback of the technique, since it has been reported that around 90% of voxels in the brain contain fiber crossing.30 As a consequence, it is not possible to visualize motor tracts originating from the lateral motor cortex, and this problem ­limits the application of tractography only to patients where the most dorsal component of the CST connecting to the leg and hand area is a ­ ffected.31 To overcome this limitation, advanced diffusion imaging techniques that solve fiber crossing within a voxel, such as high angular resolution diffusion imaging (HARDI), have been developed, but still do not meet FDA criteria for approval in clinical practice.32,​33,​34 Popularized by Catani in 2002,35 this technique has allowed for dissection of in vivo white matter tracts and has been used for research and clinical purposes. Since awake surgery may not be an option because of limited collaboration or age, tractography could be the best approach available to evaluate subcortical anatomy in children. Nevertheless, since brain shift phenomena could occur during the surgery, caution must

285

286

III  Surgical Anatomy and Mapping Techniques be paid in the interpretation of tractography results,36 and the surgeon must not rely solely on the tractography when performing the operation.16 Another use of tractography in surgical practice is to evaluate surgical risk of specific postoperative deficits. In particular, Powell et al37 compared preoperative tractography of the optic radiation and the postoperative structural MRI showing that postoperative superior homonymous quadrantanopia occurred following resection of the Meyer’s loop. Follow-up studies36,​38 reported that the location of the tip of the Meyer loop and the extent of resection were significant predictors of postoperative visual ­deficits.

„„ Direct Cortical Stimulation: Traditional Penfield’s Technique Direct cortical stimulation (DCS) is an old technique, popularized in the first half of the last century by W. Penfield6 and, in children, it has been used mainly in epilepsy surgery.39,​40,​41 Until recently, in pediatric neurosurgery, DCS has traditionally been performed using Penfield’s technique, which is characterized by a continuous cortical stimulation over a few seconds with a frequency of 50 to 60 Hz and a biphasic stimulus of 0.5-ms duration, and current intensity up to a maximum of 18 to 20 mA.42 If no responses are recorded with intensity up to 20 mA, that part of the cortex might be considered not functional. However, before labeling a cortical area as nonfunctional, it is essential to repeat the stimulation for consistency, and to exclude any technical problem because a stimulation study that is entirely negative never provides adequate security to plan a resection, unless there is a severe or even complete preoperative deficit. Penfield’s technique is still considered a standard method nowadays to perform cognitive mapping, but it has some disadvantages for motor mapping. First, an incidence of intraoperative seizures reported as high as 20%.43 A second limitation of the Penfield technique is the inability to provide continuous monitoring of the motor pathways because of the stimulation parameters, so that their functional integrity cannot be assessed continuously during surgery. The third, and most relevant limitation in the pediatric population, as consistently reported by several groups with robust experience in epilepsy surgery, is the very low success rate in eliciting a motor response when DCS is performed in children under the age of 5 to 6 years.5,​39,​40,​44 Penfield’s technique, therefore, remains valuable mainly for language and other cognitive mapping, but, since this requires an awake and cooperative patient, it is of limited use in the pediatric population.

„„ Short Train Technique and Monitoring of Motor Evoked Potentials As an alternative to Penfield’s technique, the so-called “short train of stimuli technique” became available in the mid-1990s and, since then, it has progressively replaced

Penfield’s technique with regard to motor mapping and ­ ­monitoring.45,​46 This technique was originally introduced to allow continuous monitoring of muscle motor-evoked potentials (mMEPs) during transcranial electrical stimulation (TES) of the brain. A short train of five to seven pulses (each of 0.5ms duration and with interstimulus interval around 4.1 ms) is applied to the skull, using either corkscrew or needle electrodes, placed on the scalp according to the 10–20 international electroencephalography system. To avoid penetrating injury, these electrodes should be carefully placed in infants under 12 to 18 months of age with an open fontanel; when a shunt system is present, care should be taken to avoid injuring the valve or the catheter.26 mMEPs are then recorded by placing needle electrodes in the contralateral limb muscles. The selection of muscles that have richer corticospinal innervation is fundamental to obtain robust mMEPs. The abductor pollicis brevis (APB) and the long forearm flexor or extensor have been shown to be good options for the upper limb, while the abductor hallucis brevis (AHB) is the best muscle for the lower extremities. For the orofacial muscles, the orbicularis oris and orbicularis oculi muscles are generally used, as well as the genioglossus and other muscles involved in the articulation of speech. Using different montages of stimulating electrodes provides flexibility to optimize elicitation of mMEPs without muscle twitching, which can interfere with surgery. In most cases, C1–C2 is a better electrode montage for eliciting mMEPs in all contralateral limbs. Occasionally, the montage Cz–1 cm vs. Cz+6 cm can better elicit mMEPs from lower extremities, offering also the advantage of less intense muscle twitching than other montages. The more lateral electrode montages (C3–C4, C3–Cz, C4–Cz) can induce vigorous muscle twitching and, if high stimulation intensities are used, these montages also incur a higher chance of activating the deeper portion of the corticospinal tract;47 this latter may result in false-negative results if the surgical trauma occurs more superficially, that is, proximal to the point of activation of the corticospinal tract. So, although TES is a safe technique,48 whenever feasible, mMEPs elicited through DCS are preferable, as the required stimulation current is much lower (usually < 20 mA, as compared to up to 200 mA in TES) and the activation of the motor pathways more superficial, therefore with a lower risk of distal activation of the corticospinal tract (Fig. 30.1). In the DCS technique, a multicontact strip electrode is placed under the dura overlapping the precentral gyrus; this latter can be indirectly identified using the phase reversal technique to localize the central sulcus.49 An electrode at the Fpz serves as cathode. The electrode with the lowest stimulation threshold to elicit a contralateral muscle response is chosen for continuous MEP monitoring. MEPs are recorded from upper and lower extremity muscles in the same way as following TES. In brain surgery, warning criteria for mMEPs are not well defined. Preservation of mMEPs at the end of the procedure, with amplitudes similar to those at opening baselines, is indicative of good motor outcome with no or minimal and transient postoperative paresis.50 The appearance of a significant drop in mMEPs amplitude (range: 50–80%) may be indicative of injury to the motor pathways.51,​52,​53 A persistent decrease

30  Subcortical Mapping During Intracranial Surgery in Children

Fig. 30.1  Transcranial versus cortical stimulation for muscle motor-evoked potential (mMEP) monitoring. (a, b) A strip electrode positioned across the central sulcus to elicit cortical mMEP and corkscrew scalp electrodes to elicit transcranial mMEPs. Stimulation intensity is usually up to 20 mA for cortical stimulation and up to 200 mA for transcranial stimulation (assuming trains of five stimuli, 0.5-ms duration each, interstimulus interval of about 4 ms). In the same patient, a spinal motor evoked potential (D wave) recorded from an epidural electrode positioned on the thoracic spinal cord presents a much shorter latency following transcranial (4.4 ms) than cortical (6.3 ms) stimulation. (c) With direct cortical stimulation the spreading of the current is limited and the activation of the corticospinal tract is very superficial. Vice versa, transcranial electrical stimulation, particularly at high intensities and with a lateral electrode montage such as C3–C4, may activate the corticospinal tract distally, resulting in a shorter latency of the D wave and exposing to the risk of falsely negative mMEPs.

in amplitude or a reversible loss of the mMEPs may correlate with a transient motor deficit or, rarely, a permanent deficit. A complete disappearance of the mMEPs, on the other hand, strongly correlates with a permanent postoperative motor deficit, although false-positive results could occur.50 Interpretation of mMEP changes is not trivial, and it is complicated by the large variability of mMEP amplitude, even under physiological conditions. Anesthesia, body temperature, blood pressure, and parameters of stimulation may all affect the reproducibility of mMEPs; accordingly, adequate expertise is required to interpret these signals. One of the main advantages of mMEP monitoring is that it allows to monitor the functional integrity of the motor pathways from the cortex to the muscles. MEP changes can occur not only due to a mechanical injury to the CSTs secondary to coagulation, traction, or tissue damage, but also because of an ischemic injury in the case of vessel occlusion or vasospasm. These vascular derangements can be identified, and

possibly prevented, only using mMEP monitoring, not cortical or subcortical mapping techniques. With regard to the pediatric population, it should be considered that in children younger than 4 to 5 years of age, higher stimulating thresholds may be needed due to the immaturity of the motor cortex and the subcortical motor pathways.26,​54 On the other hand, while young children have slower conduction velocities due to the immaturity of their motor pathways and some degree of latency shift may be expected, these patients also have shorter limbs than adults and therefore, in the end, these two factors partially compensate each other. Remarkably, in 2009, a review on cortical stimulation in children by Gallentine and Mikati41 included only studies where Penfield’s technique was used, despite the fact that MEP monitoring and cortical mapping using the so-called “short train technique” was available since the mid-1990s.45,​46 This

287

288

III  Surgical Anatomy and Mapping Techniques bias toward Penfield’s technique clearly reflects the scarce diffusion of the latter MEP technique in pediatric neurosurgery centers, especially in North America were IONM devices for transcranial MEP monitoring were not FDA approved until 2002.55 In fact, the short train technique offers several advantages as it allows both continuous monitoring of mMEPs, through either TES or DCS, and mapping of the motor cortex through DCS. Anecdotal reports suggest that this technique has a significantly higher success rate for DCS than the traditional Penfield’s technique.26

„„ Subcortical Stimulation At the cortical level, mapping is aimed to help the surgeon deciding “where to enter” to gain access to subcortical lesions, while identifying eloquent sites which should be spared. At subcortical level, vice versa, the main goal is to decide “when to stop” the removal of an epileptogenic lesion or the disruption of epileptogenic circuits, in order to avoid injuring either cortical afferent and efferent pathways, as well as white matter bundles connecting different cortical sites. The CST can be localized at a subcortical level using the same techniques used for DCS. Our own preference is toward the use of the short train MEP technique—using a monopolar ­probe—

rather than Penfield’s technique, for both cortical and subcortical mapping. However, it is necessary to clarify the issue of monopolar versus bipolar stimulation, as this is often a matter of confusion. “Monopolar” and “bipolar” strictly refer to the type of probe used to deliver the current, not to the parameter of stimulation. Traditionally, Penfield’s technique is performed using a bipolar handheld stimulator with two ball tips about 1 cm apart, while the short train technique is usually performed through a monopolar probe, with a reference electrode that can be inserted in the temporalis muscle. With a monopolar stimulation, the current field is more diffuse, and the volume of brain tissue stimulated increases with the intensity of the stimulation, with the possibility of activating motor pathways at some distance (20–25 mm) from the point of stimulation. Conversely, with a bipolar stimulator, the electrical field is more circumscribed, and there is a lower risk of distal activation of the motor tracts, but the stimulation will produce no response unless the probe is almost directly on the tract (Fig. 30.2).56 Our own preference is for monopolar stimulation because the surgeon can determine whether or not he/she is approaching the tract of interest judging by the current threshold to elicit a motor response. Nevertheless, we would like to emphasize that, regardless of the technique, the most important variable for a successful mapping is the experience of the team (neurosurgeon and neurophysiologist) with that specific technique.

Fig. 30.2  (a) Bipolar (B) and monopolar (M) probes for direct cortical stimulation. (b, c) Electric field distribution using bipolar (b) and monopolar (c) probes. A, anode; C, cathode. (Reproduced with permission from Sala F et al 2010.26)

30  Subcortical Mapping During Intracranial Surgery in Children In the recent past, there has been a great interest to understand the relationship between the threshold current necessary to elicit a subcortical motor response (subcortical threshold) and the distance between the stimulation site and the CST itself. Current evidence suggests that a subcortical threshold current of 1 mA correlates with a 1-mm distance between stimulation site and the CST.53,​57,​58 However, it should be considered that this “rule of thumb” of 1 mA equal 1 mm depends also on the parameters of stimulation. As recently pointed out in a paper by Shiban et al,59 this correlation is more proximal to the 1:1 when cathodal rather than anodal stimulation is used and pulse duration is 0.5 to 0.7 ms. So, with these parameters, one could expect that a subcortical motor threshold of 10, 5, and 1 mA to elicit a motor response from contralateral muscles translates in approximately 10, 5, and 1 mm of distance from the CST, respectively. However, whether or not this same rule applies to children and especially to very young children remains undetermined. Schucht et al60 published a series of 8 patients on the use of low-threshold monopolar motor mapping for the resection of lesions in motor eloquent areas in children and adolescents. They showed a very high success rate for both MEP monitoring and cortical–subcortical mapping in children. However, only one child in this study was younger than 6 years of age, and therefore results cannot be extrapolated to the younger children. With regard to monopolar versus bipolar stimulation, although applied to an adult population, the study by Szelényi et al61 confirmed the fact that for subcortical mapping, the combination of short train technique and monopolar rather than bipolar stimulation offers the best chances for a successful mapping, as compared to Penfield’s technique and bipolar stimulation. In fact, there is very little data on subcortical motor mapping in children, and virtually no data exist in the pediatric population with regard to the subcortical mapping of other, nonmotor, pathways. A matter of debate is the correlation between subcortical mapping thresholds and the risk of postoperative deficits. Again, most of the data available in the literature are related to motor function. It is intuitive that the lower the threshold capable of eliciting a response, the higher is the risk of postoperative deficit because of the proximity to the motor pathways. So, for example, if the threshold is lower than 3 to 4 mA, there is a significant risk of injury because the CST is only 3 to 4 mm away from the dissection. In general, thresholds above 5 mA are usually considered safe while there is some consistency in reporting a significantly higher risk of postoperative, at least transient, paresis for subcortical threshold of 3 mA or lower.53,​62 While all these conclusions are culled from studies in adult patients, whether or not the same criteria apply to children remains uncertain. Recently, technological innovation has helped to push the limits of subcortical mapping to the edge. Nowadays, both suction devices and ultrasonic aspirators combined with a stimulating probe are available.63,​64 These tools allow to perform continuous or subcontinuous subcortical mapping without the need of periodically alternating tissue removal with stimulation. It is likely that, apart from vascular injuries, irreversible mMEP changes during tissue removal at apparently safe subcortical mapping threshold (above 3–5 mA) may

be due to inappropriate temporal and spatial coupling of the surgical field. It could happen that, between two consecutive subcortical mapping procedures, the surgeon removes too much tissue ending up in very close proximity to the white matter tracts. With the technical adjustments offered by these new tools, some authors started to suggest that subcortical threshold as low as 1 to 2 mA can be tolerated, resulting in only transient deficit. Again, however, we would like to stress the fact that such outstanding results are based on a robust and well-established experience, which combine expertise in both intraoperative neurophysiology and brain tumor surgery. As much as the limits of subcortical mapping could and will be pushed forward in the years to come, we would consider as critical any threshold below 3 to 5 mA, and we would recommend to stop tissue removal at this stage to avoid any permanent injury to the subcortical pathways. Illustrative cases of subcortical mapping are presented in Fig. 30.3, Fig. 30.4, Fig. 30.5, and Fig. 30.6.

„„ Cortico-Cortical Evoked Potentials In corticocortical evoked potentials (CCEPs), time-locked stimulation and registration of two distant cortical areas are used to reveal their subcortical connection. First described by Matsumoto et al,65 using CCEPs in neurosurgery has been a useful in vivo method to track white matter connections in real time in epilepsy patients. This technique could be considered complimentary to tractography to the extent that while tractography allows for structural visualization of white matter anatomy, CCEPs provide a unique chance to approach the subcortical white matter anatomy by measuring the relationship between the electrical activity of two or more specialized cortical areas. Having implanted subdural electrodes for presurgical evaluation, 20 to 100 square electrical pulses with a duration of 0.3 ms are applied over the eloquent cortices at alternating polarity with a frequency of 1 Hz in the awake patient, and then averaged in relation to the stimulus onset. The intensity is set at 10 to 12 mA if no clinical signs or afterdischarges are present at 15 mA, and the stimulation is repeated in at least two different trials to confirm the reproducibility.65 The cortical area evaluated covers the epileptic hemisphere and particular attention, with relation to language network, is paid to the connections between anterior (fronto-opercular) and posterior (posterotemporal) language regions. The resulting electrocorticogram between these areas shows two peaks, a first peak latency (N1) ranged from 22 to 36 ms (mean: 27.9 ms) while a second (N2) from 113 to 164 ms (mean: 144.6 ms). Moreover, the distribution of N2 (3–21 electrodes) is larger than that of N1 (1–20 electrodes). In one study, the distribution of N2 potentials has suggested the presence of a wider posterior language region in the temporo-parieto-occipital junction, while the peak latency of N1 is suggested to represent the direct connection between anterior (Broca’s) and posterior (Wernicke’s) area. These results were then supported by Catani and Mesulam,66 who described the arcuate fasciculus as a perisylvian language network composed of a direct segment, the long segment (LS) of the arcuate fasciculus, connecting Wernicke’s and Broca’s area, and two indirect segments, an anterior segment (AS) and a posterior segment, connecting Wernicke’s and Bro-

289

290

III  Surgical Anatomy and Mapping Techniques

Fig. 30.3  Illustrative case of subcortical mapping. A 3-year-old boy presented with recurrent focal seizures in the form of absences. A threedimensional reconstruction of the MRI shows a left medial frontal lesion occupying the supplementary motor area (SMA), M1, and cingulate gyrus (a). HARDI tractography revealed that the corticospinal tract (in yellow) was displaced laterally and was in contact with the lesion (b). During surgery, cortical and subcortical mapping was performed with the train-of-five technique (c). Subcortical mapping showed responses in the right tibialis anterior (TA) and abductor hallucis brevis (AHB) using a stimulation intensity of 10 mA at site A. At site B, there was no response at 10 mA, but increasing the stimulation intensity to 20 mA produced responses in both the abductor pollicis brevis (APB) and TA. At site C, these responses were maintained using a stimulation intensity of 25 mA. The decrease in amplitude of TA and AHB was over 50%, although it was still possible to evoke responses. Postoperatively, the patient suffered from mild motor deficits in the right leg and, pathology revealed the lesion to be cortical dysplasia. No seizures occurred after the surgery, and motor function was completely recovered at 6-month follow-up.

ca’s in the so-called Geschwind’s area in the inferior parietal cortex.67 Another study by Matsumoto et al has also confirmed the potential of CCEPs in evaluating the motor networks with the same technique.68 Interestingly, as the incidence of intraoperative seizure has shown to be low when using CCEPs and the technique does not require patient cooperation, CCEPs may be a valid alternative to DCS in pediatric patients that cannot show sufficient compliance to undergo awake surgery. CCEPs were applied in ­oncological supratentorial surgery by Yamao with both awake and asleep paradigms.69 In this study, tractography and fMRI

were used to predict the location of anterior and posterior language areas. The predicted regions were then tested during the awake surgery and CCEPs were used during the operation to assess functional integrity. Even with decrease in amplitude, N1 was preserved in all subjects, and no patient showed long-term sequelae from the operation and the presence of N1 was proposed as a marker to assess white matter integrity. Nevertheless, a pattern between CCEPs modification and postoperative recovery was not shown. To sum up, studies using CCEPs are promising but few, and a wider clinical validation is needed.

30  Subcortical Mapping During Intracranial Surgery in Children

Fig. 30.4  Intraoperative neuronavigation snapshots during surgery for a high-grade glioma of the Rolandic area in a 17-year-old adolescent boy who presented with right hand paresthesia. The preoperative axial (a), coronal (b), and sagittal (c) contrast-enhanced T1-weighted MR images show the anteromesial displacement of the corticospinal tract (light blue), revealing that the tumor is posterolateral to this tract. The red point indicates the site of intraoperative direct subcortical stimulation toward the end of tumor removal. (d, e) Contralateral muscle responses elicited by subcortical mapping. A consistent response from the right abductor pollicis brevis, extensor digitorum communis, and hypoglossal nerve was recorded at the end of tumor removal with stimulation intensities ranging from 10 to 15 mA (short train-of-five stimuli, 0.5-ms duration, interstimulus interval 0.4 ms, repetition rate 1 Hz). These relatively high subcortical mapping thresholds and the lack of any significant preoperative motor deficit suggest that the corticospinal tract was rather displaced by the bleeding tumor and the surrounding edema. (f) Continuous muscle motor evoked potential (mMEP) monitoring in cascade mode shows the presence of contralateral mMEPs throughout the procedure. The right abductor hallucis response is not showed in this particular screen, because it required a different stimulation montage. Postoperative coronal (g), sagittal (h), and axial (i) contrast-enhanced T1-weighted MR images show complete tumor resection. Pathology revealed a glioblastoma. The patient presented no additional neurological deficits postoperatively and went for adjuvant therapies with Karnofsky’s score of 100. R VII-FACC, right orbicularis oris; R EXT, right extensor digitorum communis; R ABP, right abductor pollicis brevis; R TIB ANT, right tibialis anterior; R AB ALL, right abductor hallucis; R XII-HYPO, right genioglossus.

291

292

III  Surgical Anatomy and Mapping Techniques

Fig. 30.5  Subcortical mapping. Axial (a), sagittal (b), and coronal (c) contrast-enhanced, T1-weighted images of a lesion occupying the right thalamus and displacing anteriorly the internal capsule in an 8-year-old boy who presented with a left hemiparesis. (d, e) Tractography revealed the corticospinal tract (in blue) displaced anteriorly and laterally to the tumor. The red point indicates the site of intraoperative direct subcortical stimulation toward the end of tumor removal.

„„ Conclusion Electrophysiological stimulation of the human brain is a fascinating research field with a remarkable impact on the clinical practice. The last two decades have seen the resurgence of awake craniotomy and a shift in the concept of eloquent brain areas, which are nowadays no more limited to the cortex but more and more related to the subcortical connectivity. Subcortical mapping plays, therefore, a key role, and new paradigms of stimulation as well as warning criteria have been established. However, the surgeon should always keep in mind that a number of variables can affect the results of mapping, such as pulse duration, current intensity, number of pulses, train duration, frequency of stimulation, and stimulating probe (bipolar vs. monopolar). Subcortical mapping is very useful in localizing function at the subcortical level but cannot assess the functional i­ ntegrity

of subcortical pathways. To monitor these tracts either an awake craniotomy is needed for language and other cognitive function, or mMEPs can be monitored to assess corticospinal pathways in the asleep patient. In 2003 we strongly recommended the combination of subcortical mapping with mMEP monitoring in order to not only avoid a mechanical injury to motor pathways, but also to timely detect an impending vascular injury and apply corrective measures to counteract a possible ischemia.70 Since then, a number of larger studies have been published confirming the critical role of MEP monitoring and subcortical mapping in preventing injury to the descending tracts in brain surgery. Warning criteria have been set for both MEP amplitude changes and subcortical mapping thresholds. Whether or not these same criteria apply to children remains largely undetermined and it should be investigated in both pediatric brain tumor surgery and pediatric epilepsy surgery.

30  Subcortical Mapping During Intracranial Surgery in Children

Fig. 30.6  (a) HARDI tractography reconstruction showing the corticospinal tract (in yellow) adjacent to the tumor. (b) At surgery, TES evoked MEP from the left abductor pollicis brevis (APB) and the left tibialis anterior (TA) were elicited at 120 and 200 mA, respectively, while no MEPs were obtained following DCS up to 35 mA (not shown). (c) Response from the left APB, extensor digitorum communis (EXT), and TA were recorded following subcortical stimulation at 10 mA as showed in the stimulation point A. (d) Approaching the core of the lesion, response from the left APB, EXT, and TA was recorded at stimulation intensity of 3 mA in point B. (e) At the end of tumor resection, subcortical mapping at stimulation point C showed responses from the left APB, EXT, and TA at 2 mA, when stimulating the anterior superolateral margin of the tumor. Noteworthy, from a single spot multiple muscle response are elicited due to the convergence of corticospinal tract fibers in the proximity of the internal capsule. Pathology revealed a pilomyxoid astrocytoma (WHO II). L OOM, left orbicularis oris; L EXT, left extensor digitorum communis; L APB, left abductor pollicis brevis; L TIB ANT, left tibialis anterior.

References 1. Soriano SG, Eldredge EA, Wang FK, et al. The effect of propofol on intraoperative electrocorticography and cortical stimulation during awake craniotomies in children. Paediatr Anaesth 2000;10(1):29–34 2. Delion M, Terminassian A, Lehousse T, et al. Specificities of awake craniotomy and brain mapping in children for resection of supratentorial tumors in the language area. World Neurosurg 2015;84(6):1645–1652 3. Balogun JA, Khan OH, Taylor M, et al. Pediatric awake craniotomy and intra-operative stimulation mapping. J Clin Neurosci 2014;21(11):1891–1894 4. Önal C, Otsubo H, Araki T, et al. Complications of invasive subdural grid monitoring in children with epilepsy. J Neurosurg 2003;98(5):1017–1026 5. Chitoku S, Otsubo H, Harada Y, et al. Extraoperative cortical stimulation of motor function in children. Pediatr Neurol 2001;24(5):344–350 6. Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation. Brain 1937:389–443

7. Tzourio-Mazoyer N, Josse G, Crivello F, Mazoyer B. Interindividual variability in the hemispheric organization for speech. Neuroimage 2004;21(1):422–435 8. Geschwind N. Disconnexion syndromes in animals and man. II. Brain 1965;88(3):585–644 9. Thiebaut de Schotten M, Dell’Acqua F, Ratiu P, et al. From phineas gage and monsieur leborgne to H.M.: revisiting disconnection syndromes. Cereb Cortex 2015;25(12):4812–4827 10. Catani M. From hodology to function. Brain 2007;130 (Pt 3):602–605 11. Herbet G, Maheu M, Costi E, Lafargue G, Duffau H. Mapping neuroplastic potential in brain-damaged patients. Brain 2016;139(Pt 3):829–844 12. De Benedictis A, Duffau H. Brain hodotopy: from esoteric concept to practical surgical applications. Neurosurgery 2011;68(6):1709–1723, discussion 1723 13. Duffau H. The error of Broca: From the traditional localizationist concept to a connectomal anatomy of human brain. J Chem Neuroanat 2018;89:73–81

293

294

III  Surgical Anatomy and Mapping Techniques 14. Sarubbo S, Latini F, Sette E, et al. Is the resection of gliomas in ­ Wernicke’s area reliable?: Wernicke’s area resection. Acta ­Neurochir (Wien) 2012;154(9):1653–1662

35. Catani M, Howard RJ, Pajevic S, Jones DK. Virtual in vivo interactive dissection of white matter fasciculi in the human brain. Neuroimage 2002;17(1):77–94

15. Voets NL, Bartsch A, Plaha P. Brain white matter fibre tracts: a review of functional neuro-oncological relevance. J Neurol Neurosurg Psychiatry 2017;88(12):1017–1025

36. Chen X, Weigel D, Ganslandt O, Buchfelder M, Nimsky C. Prediction of visual field deficits by diffusion tensor imaging in temporal lobe epilepsy surgery. Neuroimage 2009;45(2):286–297

16. Duffau H. Stimulation mapping of white matter tracts to study brain functional connectivity. Nat Rev Neurol 2015;11(5):255–265

37. Powell HWR, Parker GJM, Alexander DC, et al. MR tractography predicts visual field defects following temporal lobe resection. Neurology 2005;65(4):596–599

17. O’Rahilly R, Muller F. Human Embryonic Brain. An Atlas of Developmental Stages. New York, NY: Wiley-Liss; 1994. Available at: https://scholar.google.it/scholar?hl=en&as_sdt=0%252C5&q= Human+embryonic+brain.+An+atlas+of+developmental+stages&btnG=. Accessed January 20, 2018

38. Yogarajah M, Focke NK, Bonelli S, et al. Defining Meyer’s loop-­ temporal lobe resections, visual field deficits and diffusion tensor tractography. Brain 2009;132(Pt 6):1656–1668

18. Humphrey T. The development of the pyramidal tracts in human fetuses correlated with cortical differentiation. In: Tower DB, Schade JP, eds. Structure and Function of the Cortex: Proceedings of the Second International Meeting of Neuro-Biologists. Vol Amsterdam: Elsevier; 1960:93–103

40. Riviello JJ, Kull L, Troup C, Holmes GL. Cortical stimulation in children: techniques and precautions. Tech Neurosurg 2001; 7(1):12–18. Available at: https://insights.ovid.com/techniques -neurosurger y/techn/2001/03/000/cortical-stimulation-­ children-techniques/4/00127927. Accessed January 20, 2018

19. Eyre JA, Miller S, Clowry GJ, Conway EA, Watts C. Functional corticospinal projections are established prenatally in the human foetus permitting involvement in the development of spinal motor centres. Brain 2000;123(Pt 1):51–64 20. Eyre JA. Corticospinal tract development and its plasticity after perinatal injury. Neurosci Biobehav Rev 2007;31(8):1136–1149 21. Müller K, Hömberg V, Lenard HG. Magnetic stimulation of motor cortex and nerve roots in children. Maturation of cortico-­ motoneuronal projections. Electroencephalogr Clin Neurophysiol 1991;81(1):63–70 22. Paus T, Zijdenbos A, Worsley K, et al. Structural maturation of neural pathways in children and adolescents: in vivo study. Science 1999;283(5409):1908–1911

39. Duchowny M, Jayakar P. Functional cortical mapping in children. Adv Neurol 1993;63:149–154

41. Gallentine WB, Mikati MA. Intraoperative electrocorticography and cortical stimulation in children. J Clin Neurophysiol 2009;26(2):95–108 42. Sala F, Kržan MJ, Deletis V. Intraoperative neurophysiological monitoring in pediatric neurosurgery: why, when, how? Childs Nerv Syst 2002;18(6–7):264–287 43. Szelényi A, Joksimovič B, Seifert V. Intraoperative risk of seizures associated with transient direct cortical stimulation in patients with symptomatic epilepsy. J Clin Neurophysiol 2007;24(1):39–43

23. Lebel C, Walker L, Leemans A, Phillips L, Beaulieu C. Microstructural maturation of the human brain from childhood to adulthood. Neuroimage 2008;40(3):1044–1055

44. Resnick TJ, Alvarez LA, Duchowny MS. Cortical stimulation thresholds in children being evaluated for resective surgery. Annual Meeting of the American Epilepsy Society. Vol San Francisco, C ­ alifornia: Raven Press; 1988:651–652. Available at: https://ci.nii.ac.jp/naid/10026281618/. Accessed January 20, 2018

24. Schevon CA, Carlson C, Zaroff CM, et al. Pediatric language mapping: sensitivity of neurostimulation and Wada testing in epilepsy surgery. Epilepsia 2007;48(3):539–545

45. Taniguchi M, Cedzich C, Schramm J. Modification of cortical stimulation for motor evoked potentials under general anesthesia: technical description. Neurosurgery 1993;32(2):219–226

25. Rivet DJ, O’Brien DF, Park TS, Ojemann JG. Distance of the motor cortex from the coronal suture as a function of age. Pediatr Neurosurg 2004;40(5):215–219

46. Pechstein U, Cedzich C, Nadstawek J, Schramm J. Transcranial high-frequency repetitive electrical stimulation for recording myogenic motor evoked potentials with the patient under general anesthesia. Neurosurgery 1996;39(2):335–343, discussion 343–344

26. Sala F, Manganotti P, Grossauer S, Tramontanto V, Mazza C, Gerosa M. Intraoperative neurophysiology of the motor system in children: a tailored approach. Childs Nerv Syst 2010;26(4):473–490 27. Sala F, Coppola A, Tramontano V, Babini M, Pinna G. Intraoperative neurophysiological monitoring for the resection of brain tumors in pediatric patients. J Neurosurg Sci 2015;59(4):373–382 28. Coppola A, Tramontano V, Basaldella F, Arcaro C, Squintani G, Sala F. Intra-operative neurophysiological mapping and monitoring during brain tumour surgery in children: an update. Childs Nerv Syst 2016;32(10):1849–1859 29. Basser PJ, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and imaging. Biophys J 1994;66(1):259–267 30. Jeurissen B, Leemans A, Tournier JD, Jones DK, Sijbers J. Investigating the prevalence of complex fiber configurations in white matter tissue with diffusion magnetic resonance imaging. Hum Brain Mapp 2013;34(11):2747–2766 31. Catani M, Dell’Acqua F. Mapping white matter pathways with diffusion imaging tractography: focus on neurosurgical applications. Brain Mapp. 2011:61–75 32. Jbabdi S, Johansen-Berg H. Tractography: where do we go from here? Brain Connect 2011;1(3):169–183 33. Dell’acqua F, Scifo P, Rizzo G, et al. A modified damped Richardson-Lucy algorithm to reduce isotropic background effects in spherical deconvolution. Neuroimage 2010;49(2):1446–1458 34. Thiebaut de Schotten M, Dell’Acqua F, Forkel SJ, et al. A lateralized brain network for visuospatial attention. Nat Neurosci 2011;14(10):1245–1246

47. Rothwell J, Burke D, Hicks R, Stephen J, Woodforth I, Crawford M. Transcranial electrical stimulation of the motor cortex in man: further evidence for the site of activation. J Physiol 1994;481 (Pt 1):243–250 48. MacDonald DB. Safety of intraoperative transcranial electrical stimulation motor evoked potential monitoring. J Clin Neurophysiol 2002;19(5):416–429–.– Available at: –Accessed January 20, 2018 49. Cedzich C, Taniguchi M, Schäfer S, Schramm J. Somatosensory evoked potential phase reversal and direct motor cortex stimulation during surgery in and around the central region. Neurosurgery 1996;38(5):962–970 50. Neuloh G, Pechstein U, Cedzich C, Schramm J. Motor evoked potential monitoring with supratentorial surgery. Neurosurgery 2004;54(5):1061–1070, discussion 1070–1072 51. Krieg SM, Schäffner M, Shiban E, et al. Reliability of intraoperative neurophysiological monitoring using motor evoked potentials during resection of metastases in motor-eloquent brain regions: clinical article. J Neurosurg 2013;118(6):1269–1278 52. Szelényi A, Hattingen E, Weidauer S, Seifert V, Ziemann U. Intraoperative motor evoked potential alteration in intracranial tumor surgery and its relation to signal alteration in postoperative magnetic resonance imaging. Neurosurgery 2010;67(2):302–313 53. Nossek E, Korn A, Shahar T, et al. Intraoperative mapping and monitoring of the corticospinal tracts with neurophysiological assessment and 3-dimensional ultrasonography-based navigation. Clinical article. J Neurosurg 2011;114(3):738–746

30  Subcortical Mapping During Intracranial Surgery in Children 54. Journee H, Hoving E, Mooij J. P27.4 Stimulation threshold–age relationship and improvement of muscle potentials by preconditioning transcranial stimulation in young children. Clin Neurophysiol 2006;117:115 55. Macdonald DB, Skinner S, Shils J, Yingling C; American Society of Neurophysiological Monitoring. Intraoperative motor evoked potential monitoring—a position statement by the American Society of Neurophysiological Monitoring. Clin Neurophysiol 2013;124(12):2291–2316 56. Szelényi A, Bello L, Duffau H, et al; Workgroup for Intraoperative Management in Low-Grade Glioma Surgery within the European Low-Grade Glioma Network. Intraoperative electrical stimulation in awake craniotomy: methodological aspects of current practice. Neurosurg Focus 2010;28(2):E7 57. Kamada K, Todo T, Ota T, et al. The motor-evoked potential threshold evaluated by tractography and electrical stimulation. J Neurosurg 2009;111(4):785–795 58. Ohue S, Kohno S, Inoue A, et al. Accuracy of diffusion tensor magnetic resonance imaging-based tractography for surgery of gliomas near the pyramidal tract: a significant correlation between subcortical electrical stimulation and postoperative tractography. Neurosurgery 2012;70(2):283–293, discussion 294 59. Shiban E, Krieg SM, Haller B, et al. Intraoperative subcortical motor evoked potential stimulation: how close is the corticospinal tract? J Neurosurg 2015;123(3):711–720 60. Schucht P, Seidel K, Murek M, et al. Low-threshold monopolar motor mapping for resection of lesions in motor eloquent areas in children and adolescents. J Neurosurg Pediatr 2014;13(5):572–578 61. Szelényi A, Senft C, Jardan M, et al. Intra-operative subcortical electrical stimulation: a comparison of two methods. Clin ­Neurophysiol 2011;122(7):1470–1475

62. Seidel K, Beck J, Stieglitz L, Schucht P, Raabe A. The ­warning-sign hierarchy between quantitative subcortical motor mapping and continuous motor evoked potential monitoring during resection of supratentorial brain tumors. J Neurosurg 2013;118(2):287–296 63. Shiban E, Krieg SM, Obermueller T, Wostrack M, Meyer B, Ringel F. Continuous subcortical motor evoked potential stimulation using the tip of an ultrasonic aspirator for the resection of motor eloquent lesions. J Neurosurg 2015;123(2):301–306 64. Raabe A, Beck J, Schucht P, Seidel K. Continuous dynamic mapping of the corticospinal tract during surgery of motor eloquent brain tumors: evaluation of a new method. J Neurosurg 2014;120(5):1015–1024 65. Matsumoto R, Nair DR, LaPresto E, et al. Functional connectivity in the human language system: a cortico-cortical evoked potential study. Brain 2004;127(Pt 10):2316–2330 66. Catani M, Mesulam M. The arcuate fasciculus and the disconnection theme in language and aphasia: history and current state. Cortex 2008;44(8):953–961 67. Catani M, Jones DK, ffytche DH. Perisylvian language networks of the human brain. Ann Neurol 2005;57(1):8–16 68. Matsumoto R, Nair DR, LaPresto E, Bingaman W, Shibasaki H, Lüders HO. Functional connectivity in human cortical motor ­system: a cortico-cortical evoked potential study. Brain 2007;130(Pt 1):181–197 69. Yamao Y, Matsumoto R, Kunieda T, et al. Intraoperative dorsal language network mapping by using single-pulse electrical stimulation. Hum Brain Mapp 2014;35(9):4345–4361 70. Sala F, Lanteri P. Brain surgery in motor areas: the invaluable assistance of intraoperative neurophysiological monitoring. J ­Neurosurg Sci 2003;47(2):79–88

295

IV Part IV

Surgical Treatment of E ­ pilepsy

IVa.    Anesthesia

  31 Anesthetic Considerations and Postoperative Intensive Care Unit Care in Pediatric Epilepsy Surgery  301

  32  Pediatric Awake Craniotomy

307

IVb.    Intracranial Electrode ­Placement for Invasive ­Monitoring   33 Implantation of Strip, Grid, and Depth Electrodes for Invasive Electrophysiological Monitoring

315

  34 Stereoelectroencephalography in Children: Methodology and Surgical Technique

323

IVc.    Temporal Lobe Epilepsy and Surgical Approaches

  35 Mesial Temporal Sclerosis in Pediatric Epilepsy 333

  36 Anteromesial Temporal Lobectomy 341

  37 Selective Amygdalohippocampectomy

351

  38 Surgical Management of Lesional Temporal Lobe Epilepsy

364

  39 Surgical Management of MRI-Negative Temporal Lobe Epilepsy

376

297

IV IVd.    E  xtratemporal Lobe ­Epilepsy and Surgical Approaches

298

  40 Surgical Management of InsularOpercular Epilepsy in Children

386

  41 Focal Cortical Dysplasia: Histopathology, Neuroimaging, and Electroclinical ­Presentation

404

  42 Surgical Approaches in Cortical Dysplasia

414

  43 Tuberous Sclerosis Complex

424

  44 Resective Epilepsy Surgery for Tuberous Sclerosis Complex

432

  45 Extratemporal Resection and Staged Epilepsy Surgery in Children

437

  46 Supplementary Sensorimotor Area Surgery

447

  47 Rolandic Cortex Surgery

452

  48 Anterior Peri-insular Quadrantotomy

459

  49 Posterior Peri-insular Quadrantotomy

465

  50 Tailored Extratemporal Resection in Children with Epilepsy

472

IV   51 Surgical Management of MRI-Negative Extratemporal Lobe Epilepsy

481

  52 Surgical Management of Hypothalamic Hamartomas

489

IVe.    Hemispheric Surgery Techniques   53    Hemispherectomy and Hemispherotomy Techniques in Pediatric Epilepsy Surgery

497

  54   Multifocal Epilepsy and Multilobar Resections

506

  55   Anatomical Hemispherectomy

510

  56   Hemidecortication for Intractable Epilepsy

519

  57   Functional Hemispherectomy at the University of California, Los Angeles

530

  58   Transsylvian Hemispheric Deafferentation 

542

  59   Vertical Parasagittal Hemispherotomy

553

  60   Peri-insular Hemispherotomy

562

  61   Endoscope-Assisted Hemispherotomy

572

299

IV IVf.     Other Disconnective ­Procedures   62 Corpus Callosotomy

581

  63 Endoscope-Assisted Corpus Callosotomy with Anterior, Hippocampal, and Posterior Commissurotomy

589

  64 Endoscopic Disconnection of Hypothalamic Hamartomas

597

  65 Multiple Subpial Transections in Children with Refractory Epilepsy

603

  66 Hippocampal Subpial Transection

608

IVg.    Neuromodulation ­Procedures   67  Vagal Nerve Stimulation

613

  68  Cortical and Deep Brain Stimulation 619

IVh.  Radiosurgery and Ablative ­Procedures   69 Radiosurgical Treatment for Epilepsy

625

  70 Stereotactic Laser Ablation for Hypothalamic Hamartomas

640

  71 MRI-Guided Laser Thermal Therapy in Pediatric Epilepsy Surgery

645

31

  Anesthetic Considerations and Postoperative Intensive Care Unit Care in Pediatric Epilepsy Surgery Sulpicio G. Soriano and Michael L. McManus

Summary The surgical management of intractable seizures is advancing rapidly, and surgical intervention is increasingly common among children with epilepsy. Here, we review the anesthetic considerations in caring for children and their general application in preoperative, intraoperative, and postoperative care. Attention is given to specific procedures including grids and strips placement, stereotactic ablation, seizure focus excision, awake craniotomy, corpus callosotomy, hemispherectomy, and placement of vagal nerve stimulators. Keywords:  pediatric, epilepsy, seizures, grids and strips, ­stereotactic, corpus callosotomy, hemispherectomy, vagal nerve stimulator

„„ Introduction The surgical management of intractable epilepsy has evolved due to rapid advances in intraoperative neuroimaging, electroencephalography (EEG), and surgical approaches. These advances have dramatically improved the outcome in infants and children. The aim of this section is to highlight the agedependent aspects of the perioperative management of the pediatric neurosurgical patient in an evolving field.

„„ Physiological Differences in Pediatrics Age-dependent differences in cerebrovascular physiology have a significant impact on the perioperative management of neurosurgical patients. Cerebral blood flow (CBF) is coupled tightly to metabolic demand and is approximately 40 mL/100 g/min in healthy neonates.1 CBF peaked between 2 and 4 years and settled at 7 to 8 years.2 These changes mirror changes in neuroanatomical development. A simplified view of the cerebral autoregulatory range in a normal newborn lies between blood pressures of 20 and 60 mm Hg, which reflect the relatively low cerebral metabolic requirements and blood pressure of the perinatal period. However, recent reports demonstrate a

wide range of cerebral autoregulation in pediatric patietns.3,​4 Although children younger than 2 years of age have lower baseline mean arterial pressures, they have lower autoregulatory reserve and can theoretically be at greater risk of cerebral ischemia.5,​6 These factors place the infant at risk for significant hemodynamic instability during neurosurgical procedures as compared to adults.

„„ Preoperative Evaluation and Preparation A thorough preoperative organ system-based evaluation of the pediatric patient is essential to minimize perioperative morbidity since infants are at higher risk for perioperative morbidity and mortality than any other age group.7 Neurosurgical lesions and associated surgery in this vulnerable age group have been linked to increased mortality and decreased academic achievement scores at adolescence.8 Furthermore, outcome studies reveal increased perioperative morbidity and mortality in this patient population.9,​10 Respiratory and cardiac-related events account for a majority of these complications and necessitate a thorough history and physical examination. A thorough review of the patient’s history can reveal conditions that may require further evaluation and be optimized before surgery. A complete airway examination is essential, since some craniofacial anomalies may require specialized techniques to secure the airway.11 Optimal cardiac function is crucial intraoperatively since massive blood loss, swings in blood pressure, electrolyte shifts, and aggressive fluid administration may lead to depression of myocardial contractility and acute myocardial failure. Therefore, a pediatric cardiologist should be consulted to evaluate all patients with suspected problems in order to identify any lesion and assess cardiac function prior to surgery. A variety of medical conditions often accompany pediatric patients with epilepsy, and these need to be addressed when formulating the anesthetic plan. Tuberous sclerosis (TS) is a hamartomatous disease that usually presents with cutaneous and intracranial lesions, the later leading to medically intractable epilepsy.12 Hamartomatous lesions frequently infiltrate and disturb the cardiac, renal, and pulmonary systems as

302

IVa Anesthesia well. Cardiac rhabdomyomas can be detected in a majority of these patients and can lead to dysrhythmias, obstruction of intracardiac blood flow, and abnormal conduction through the bundle of His. Therefore, all patients with TS should have a preoperative echocardiogram and electrocardiograph in order to detect any functional defects. Renal lesions often result in hypertension and azotemia, both of which may complicate the conduct of anesthesia. Sturge–Weber syndrome, or encephalotrigeminal angiomatosis, is another of the phakomatoses and is characterized by port-wine facial stains and ipsilateral leptomeningeal angiomas. Intracranial angiomata with calcification (“railroad sign”) produce cerebral atrophy, mental retardation, and seizures that are often refractory to medical management. Extracranial angiomata, including lesions involving the airway, have been reported and congenital glaucoma is present in a third of patients. Thus, airway management, intraocular pressure, and intraoperative hemorrhage are important considerations. Preoperative laboratory tests should be tailored to the proposed neurosurgical procedure. Open craniotomies for lobectomies and hemispherectomies are associated with significant blood loss and mobidity.13 Hypercoagulation develops early after resection of brain tissue in pediatric neurosurgical patients, as assessed by thromboelastography.14 Given the risk of significant blood loss associated with craniotomies, a hematocrit, prothrombin time, and partial thromboplastin time should be obtained in order to uncover any insidious hematological or coagulation disorders. Type and cross-matched blood should be available prior to all craniotomies. All patients presenting for epilepsy surgery have a history of pharmacological therapy of their seizures. Each drug class has side effects that can affect the conduct of anesthesia. The traditional anticonvulsant drugs, phenobarbital, phenytoin, and carbamazepine, respectively, are potent inducers of hepatic microsomal P-450 enzymes. Fosphenytoin and Carbatrol/ oxcarbazepine are recent reformulations of the latter two drugs. The hepatic P-450 enzymes mediate biotransformation and enhanced elimination of many drugs. Long-term administration of these specific anticonvulsant drugs results in drug resistance and increase requirements for both nondepolarizing muscle relaxants and opioids administered during general anesthesia.15 Patients on chronic anticonvulsant drug therapy with phenobarbital, phenytoin, and carbamazepine ­ need to be closely monitored for drug effect, and dosage should be increased accordingly.16 In general, the newer classes of anticonvulsant drugs do not appear to alter the metabolism of anesthetic drugs. However, other side effects have been reported with chronic administration of these new drugs. Topiramate has been shown to cause an asymptomatic anion gap metabolic acidosis due to inhibition of carbonic anhydrase.17 This can amplify the metabolic acidosis that often occurs as a result of hypoperfusion due to massive blood loss. Sodium valproate is associated with platelet abnormalities and can cause bleeding disorders. Sodium valproate and felbamate can induce liver failure, and patient receiving these drugs should have the appropriate laboratory tests to determine the baseline platelet and liver function prior to surgery. The ketogenic diet is a high-fat, low-carbohydrate, lowprotein regime that promotes a chronic metabolic state of ketosis and acidosis. For reasons that remain unclear, this has proven to be a very useful adjunct in the treatment of many

children with intractable epilepsy. Although adequate calories are ­provided with fat, carbohydrate intake is limited to 5 to 15 g/d, and hypoalbuminemia is common. Since the metabolic state can be disrupted by administration of carbohydrate containing intravenous solutions or by ingestion of the sweetened syrups contained in some premedications, these should be avoided by the anesthesiologist.18 While both normal saline and lactated Ringer’s are acceptable fluid choices, there is limited margin for additional acidosis and the acid-base status must be monitored closely during surgery. Bicarbonate, plasma glucose, and serum ketone levels should be measured preoperatively and then sampled regularly to avoid hypoglycemia or excessive ketosis (serum or urine ketones > 160 mg/dL).

„„ Anesthetic Management Premedication Separation from parents and perioperative anxiety play a significant role in the care of the pediatric patient and are related to the cognitive development and age of the child. Preoperative sedatives given prior to the induction of anesthesia can ease the transition from the preoperative holding area to the operating room.19 Midazolam administered orally is particularly effective in relieving anxiety and producing amnesia. If an indwelling intravenous catheter is in place, midazolam can be slowly titrated to achieve sedation. If intraoperative electrocorticography (ECoG) is planned, the dosage of midazolam should be reduced in order to minimize the depressant effects of benzodiazepines on the ECoG.

Induction of Anesthesia The patient’s neurological status and coexisting medical conditions will dictate the appropriate technique and drugs for induction of anesthesia. In infants and young children, general anesthesia can be induced with inhalation of sevoflurane and nitrous oxide in oxygen. Sevoflurane has been shown to have epileptogenic potential.20 However, the mechanism of this phenomenon is unclear. Alternatively, if the patient already has intravenous (IV) catheter in place, anesthesia can be induced with the sedative hypnotic drug propofol (3–4 mg/kg). These drugs rapidly induce unconsciousness and can blunt the hemodynamic effects of tracheal intubation. A nondepolarizing muscle relaxant is then administered after induction of general anesthesia in order to facilitate intubation of the trachea. Patients with nausea or gastroesophageal reflux disorder are at risk for aspiration pneumonitis and should have a rapid-sequence induction of anesthesia performed with thiopental or propofol immediately followed by a rapid-acting muscle relaxant and cricoid pressure. Rocuronium can be used when succinylcholine is contraindicated, such as for patients with spinal cord injuries or paretic extremities. In these instances, succinylcholine can result in sudden, ­catastrophic hyperkalemia.

Airway Management Given the high incidence of respiratory morbidity and mortality in pediatric patients, a thorough examination of the

31  Anesthetic Considerations and Postoperative Intensive Care Unit Care in Pediatric Epilepsy Surgery airway and the use of appropriate equipment and techniques are mandatory. Since the trachea is relatively short, an endotracheal tube can easily migrate into a mainstem bronchus if an infant’s head is flexed or turned. Therefore, great care should be devoted to assuring proper position of the endotracheal tube during tracheal intubation. Patients undergoing awake craniotomies are always at risk for airway compromise due to sedation, seizure, or obstruction due to positioning. These factors mandate that the patient’s face should be accessible to the anesthesiologist for manipulation of the airway and ventilation of the lungs.

Positioning Patient positioning for surgery requires careful preoperative planning to allow adequate access to the patient for both the neurosurgeon and anesthesiologist. This issue is especially important in patients undergoing awake craniotomies. In this case, the patient has to be in a comfortable position throughout the surgical procedure. A clear channel should be created in front of the patient’s face in order to facilitate communication and facial observation during the neuropsychological assessment. If cortical stimulation or induction of the seizure is planned, the patient’s limbs must be easily visualized. Frequently, neurosurgical procedures are performed with the head slightly elevated to facilitate venous and cerebral spinal fluid drainage from the surgical site. However, superior sagittal sinus pressures decrease with increasing head elevation, and this increases the likelihood of venous air embolus (VAE).21 Extreme rotation of the head can impede venous return through the jugular veins and lead to impaired cerebral perfusion and increased intracranial pressure (ICP) and venous bleeding.

Vascular Access Because of limited access to the patient (especially small children) during neurosurgical procedures, optimal intravenous access is mandatory prior to the start of surgery. Typically, two large-bore venous cannulae are sufficient for most craniotomies. Should initial attempts fail, central venous cannulation may be necessary. Utilization of the femoral vein avoids the risk of pneumothorax associated with subclavian catheters, and does not interfere with cerebral venous return as may be the case with jugular catheters. Placement of peripherally inserted central catheter through either the cephalic, basilica, or brachial veins and advanced to the distal superior vena cava under ultrasound guidance have the advantage of being a noninvasive and long-term access to the central circulation.22 Since significant blood loss and hemodynamic instability can occur during craniotomies, cannulation of the radial artery would provide direct blood pressure monitoring and sampling for blood gas analysis. Other useful arterial sites in infants and children include the dorsalis pedis and posterior tibial artery.

Maintenance of Anesthesia There are several classes of drugs used to maintain general anesthesia. Potent, volatile anesthetic agents (sevoflurane, isoflurane, and desflurane) are administered by inhalation. These

drugs are potent cerebrovascular dilators and cerebral metabolic depressants, which can mediate dose-dependent uncoupling of cerebral metabolic supply and demand and increased cerebral blood volume and ICP. Moreover, the use of these agents can be associated with a significant decrease in cerebral perfusion pressure, primarily due to a dose-dependent reduction in arterial blood pressure.23 They depress the EEG and may interfere with intraoperative ECoG. Given these issues, volatile anesthetics are rarely used as the sole anesthetic for neurosurgery. Intravenous anesthetics are categorized as sedative or hypnotics and opioids.24,​25 These drugs are also potent cerebral metabolic depressants but do not cause cerebrovasodilation. The sedative or hypnotics, propofol, midazolam, and thiopental rapidly induce anesthesia and attenuate the EEG. Opioid drugs can depress the EEG, but not as severe as the sedative hypnotics. Fentanyl and other related synthetic opioids, including sufentanil, have their context-sensitive half-times increase with repeated dosing or prolonged infusions and require hepatic metabolism. As a result, the narcotic effects, such as respiratory depression and sedation, of these drugs may be prolonged. Deep neuromuscular blockade with a nondepolarizing muscle relaxant is maintained to avoid patient movement and minimize the number of anesthetic agents needed. Muscle relaxants should be withheld or permitted to wear off when assessment of motor function during neurosurgery is planned.

Intraoperative Fluid and Electrolyte Management Given the unexpected nature of sudden blood loss, normovolemia should be maintained throughout the procedure. Estimation of the patient’s blood volume is essential in determining the amount of allowable blood loss and when to transfuse blood. Blood volume depends on the age and size of the patient. Normal saline is commonly used as the maintenance fluid during neurosurgery, because it is mildly hyperosmolar (308 mOsm/kg). However, rapid infusion of large quantities of normal saline (> 60 mL/kg) can be associated with hyperchloremic acidosis.26 Given the relatively large blood volume of the neonate and infant, the maintenance rate of fluid administration depends on the weight of the patient. The maximum allowable blood loss should be determined in advance in order to determine when blood should be transfused to the patient. Initially, blood losses should be replaced with 3 mL of normal saline for 1 mL of blood loss or a colloid solution such as 5% albumin equal to the blood loss. Hematocrits of 21 to 25% should provide some impetus for blood transfusion. Massive transfusion of packed red blood cells can lead to dilutional thrombocytopenia and hypocalcemia. Hypotension should be aggressively treated with a combination of crystalloid, blood products, and vasopressors (dopamine, norepinephrine, and epinephrine). Brain swelling can be initially managed by hyperventilation and elevating the head above the heart. Should these maneuvers fail, mannitol can be given at a dose of 0.25 to 1.0 g/kg intravenously. However, repeated dosing can lead to extreme hyperosmolality, renal failure, and further brain edema.27 Furosemide is a useful adjunct to mannitol in decreasing acute cerebral edema and has been shown in vitro to prevent rebound swelling due to mannitol.

303

304

IVa Anesthesia

„„ Anesthetic Considerations for Specific Procedures Anesthesia for Placement of Grids and Strips Open placement of the intracranial grids and strips through a craniotomy carries the same operative risks noted previously. The level of anesthesia to be decreased in order to minimize its depressant effects on the EEG. Since the anticonvulsant drugs are usually held in order to detect and characterize the seizure, postoperative monitoring of the patient should focus on uncontrolled status epilepticus. The patient typically returns within 1 week for a repeat craniotomy for removal of the grids and strips and resection of the seizure focus or foci. It is important to avoid administration of nitrous oxide until the dura is opened, since intracranial air can persist up to 3 weeks following a craniotomy, and nitrous oxide in these situations can cause rapid expansion of air cavities and result in tension pneumocephalus.28 Recent advances in minimally invasive stereotactic technique for placement of intracerebral electrodes have been favorable in pediatric epilepsy patients.29 Given the lack of an open craniotomy, patients undergoing this minimally invasive technique are hemodynamically stable and usually do not require invasive hemodynamic monitoring (arterial line) and blood transfusion. These patients should be closely monitored during the postoperative period for the possibly of cerebral edema and intracranial hemorrhage.

Stereotactic-Guided Ablations Ablations of seizure foci have been in vogue in adult neurosurgery and gaining traction in the pediatric arena.30 Recently, applications of this technique include radiofrequency thermocoagulation of complex cortical foci and laser ablation of hypothalamic hamartomas in children.31,​32 As noted above, the intraoperative course of these procedures is typically uneventful, but vigilant postoperative monitoring is essential.33

Resection of Seizure Focus The depressant effects of many anesthetic agents limit the utility of intraoperative neurophysiological monitoring to be used during the surgical procedure. In general, ECoG and EEG can be utilized during low levels of volatile anesthetics. Cortical stimulation of the motor cortex necessitates observation of motor movement of the specific area of the homunculus. Therefore, muscle relaxation should be avoided or permitted to dissipate during the monitoring period. Since some epileptogenic foci are in close proximity to cortical areas controlling speech, memory, motor or sensory function, monitoring of patient and electrophysiological responses is frequently utilized to minimize iatrogenic injury to these areas.34

Awake Craniotomy Neurological function is best assessed in an awake and cooperative patient. Positioning of the patient is critical for the success of this technique. The patient should be in a s­ emilateral position to allow both patient comfort as well as surgical and

airway access to the patient. A variety of techniques have been advocated to facilitate intraoperative assessment of motor–­ sensory function and speech. These range from no sedation with local anesthesia to “asleep-awake-asleep” techniques, where the general anesthesia is induced prior to and after functional testing.35,​36 Propofol does not interfere with the ECoG, when discontinued 20 minutes before monitoring in children undergoing an awake craniotomy.37 Other drug regiments include remifentanil and dexmedetomidine. However, it is imperative that candidates for an awake craniotomy be mature and psychologically prepared to participate in this procedure.

Corpus Callosotomy Since intraoperative EEG is not required during this procedure, any anesthetic regiment can be utilized. However, the risk for uncontrolled hemorrhage and VAE still exists since the surgical approach encroaches upon the sagittal sinus. Lethargy and somnolence occur after complete division of the corpus c­ allosum and place these patients at risk for aspiration pneumonitis and airway obstruction during the immediate postoperative period. Maintenance of postoperative tracheal intubation and mechanical ventilation should be contemplated until the patient fully regains consciousness.

Hemispherectomy Hemispherectomies have the highest morbidity of all surgical procedures for intractable epilepsy. These procedures result in the loss of more than one blood volume and are associated with coagulopathy, hypokalemia, and hypothermia.38 Sustained bleeding and subsequent transfusions can progress into a low cardiac output state characterized by hypotension and limited response to administration of fluid and inotropic drugs. Given the continuous blood loss, administration of tranexamic acid has been shown to significantly reduced blood loss in children undergoing cardiac and spinal surgery.39,​40 Since the prothrombotic effect of tranexamic acid is unknown, a reduce dose of 50 mg/kg load followed by an infusion of 5 μg/kg/h may be effective. Low cardiac output states can be temporized by instituting a dopamine infusion (5–10 μg/kg/min). If this is ineffective, an epinephrine infusion (0.05–1.0 μg/kg/min) along with aggressive fluid resuscitation may be necessary. Given the large fluid shifts and somnolence that ensues, it is prudent to leave and patient’s trachea intubated and institute mechanical ventilation for the first postoperative day.

Vagal Nerve Stimulator Vagal nerve stimulation is another therapeutic option for intractable seizures. Placement of these devices is ­commonly performed under general anesthesia as same day surgical patients. Intraoperative bradycardia and transient asystole have been reported during initial stimulation and positioning of the electrodes, but these events were short lived and no morbidity evolved. Possible mechanisms for the bradycardia or asystole include stimulation of cervical cardiac branches of the vagus nerve either by collateral current spread or directly by inadvertent placement of the electrodes on one of these branches. Close observation of the electrocardiogram is prudent during testing of the vagus nerve.

31  Anesthetic Considerations and Postoperative Intensive Care Unit Care in Pediatric Epilepsy Surgery

„„ Postoperative Management Close observation in an intensive care unit with serial neurological examinations and invasive hemodynamic monitoring is helpful for the prevention and early detection of postoperative problems. It is important to recognize that postoperative seizures can still occur in these patients and lead to significantly increased morbidity. When seizures do occur, the response must be promp31: first with basic life support algorithms addressing airway, breathing and circulation, then with administration of sufficient anticonvulsant drug to stop the seizure. A common approach involves lorazepam 0.1 mg/ kg IV (repeated after 10 minutes if necessary) for immediate control, followed by fosphenytoin 20 mg/kg, phenobarbital 20 mg/kg, or levetiracetam 10 mg/kg on a regular schedule for more lasting coverage. Postoperative nausea and vomiting can result from surgery, anesthesia, or postoperative medications. Since retching will cause sudden increases in ICP, nausea should be treated with a nonsedating antiemetic. Although the efficacy of intraoperative administration of antiemetic drugs as a prophylactic measure is controversial,32 ondansetron (50 μg/kg), dexamethasone (0.25 mg/kg), and/or metoclopramide (150 μg/kg) can effectively treat nausea and vomiting in the postoperative period. Fluid and electrolyte derangements are common in the postoperative neurosurgical patient. Hyponatremia can develop due to excessive free water administration (hypotonic fluids given in the setting of high antidiuretic hormone [ADH] levels) or due to abnormal losses of sodium (cerebral salt wasting). The syndrome of inappropriate ADH secretion is marked by low sodium and hypervolemia while cerebral salt wasting is marked by low sodium and hypovolemia. Sudden falls in serum sodium concentration can precipitate seizures that resist conventional drug therapy but respond to modest volumes of hypertonic saline (3% saline, 4 mL/kg). Hypernatremia occurs less frequently, but can result from increased insensible losses (especially in infants) or the presence of diabetes insipidus. In the latter, urine

volumes in excess of 4 mL/kg are usually observed, and infusion of arginine vasopressin is necessary for correction.33 Intraoperative events that most commonly necessitate postoperative tracheal intubation and mechanical ventilation include neurological dysfunction and massive blood loss followed by volume resuscitation with fluid shifts. In these patients, continuous infusions of a narcotic (fentanyl, morphine) and a benzodiazepine (midazolam) provide effective sedation. Although propofol infusion can provide reliable sedation with quick wake-ups, in pediatric patients it may lead to a syndrome of metabolic acidosis and progressive multiorgan failure (the “propofol infusion syndrome”).34 When propofol must be used, it is prudent to minimize infusion rates to less than 3 mg/kg/h and durations to less than 12 hours. Dexmedetomidine has analgesic properties and a useful agent for reversible sedation.41 It does not cause apnea, so spontaneous ­breathing is easily maintained. Transient bradycardia, hypotension, and hypertension are associated side effects. The treatment of postoperative pain is a major component of the perioperative management of the neurosurgical patient. Since most craniotomies are observed in a critical care unit, intravenous opioids (morphine 0.1 mg/kg IV every 2–4 hours as ­needed) can be carefully titrated to blunt pain but not to the point of oversedation. Patients recovering in an unmonitored setting may receive acetaminophen (10–15 mg/kg) with minimal side effects.

„„ Conclusion The perioperative management of pediatric patients for epilepsy surgery should focus on the specific problems unique to the disease state, age of the child, and operative conditions. Thorough preoperative evaluation and open communication between members of the epilepsy team are important. A basic understanding of age-dependent variables and the interaction of anesthetic and surgical procedures are essential in minimizing perioperative morbidity.

References 1. Cross KW, Dear PR, Hathorn MK, et al. An estimation of intracranial blood flow in the new-born infant. J Physiol 1979;289: 329–345 2. Wintermark M, Lepori D, Cotting J, et al. Brain perfusion in ­children: evolution with age assessed by quantitative perfusion computed tomography. Pediatrics 2004;113(6):1642–1652 3. Lee JK. Cerebral perfusion pressure: how low can we go? Paediatr Anaesth 2014;24(7):647–648 4. Brady KM, Mytar JO, Lee JK, et al. Monitoring cerebral blood flow pressure autoregulation in pediatric patients during cardiac surgery. Stroke 2010;41(9):1957–1962 5. McCann ME, Schouten AN, Dobija N, et al. Infantile postoperative encephalopathy: perioperative factors as a cause for concern. Pediatrics 2014;133(3):e751–e757 6. McCann ME, Schouten AN. Beyond survival; influences of blood pressure, cerebral perfusion and anesthesia on neurodevelopment. Paediatr Anaesth 2014;24(1):68–73 7. Habre W, Disma N, Virag K, et al; APRICOT Group of the ­European Society of Anaesthesiology Clinical Trial Network. Incidence of severe critical events in paediatric anaesthesia (APRICOT): a

prospective multicentre observational study in 261 hospitals in Europe. Lancet Respir Med 2017;5(5):412–425 8. Hansen TG, Pedersen JK, Henneberg SW, Morton NS, Christensen K. Neurosurgical conditions and procedures in infancy are associated with mortality and academic performances in adolescence: a nationwide cohort study. Paediatr Anaesth 2015;25(2):186–192 9. Campbell E, Beez T, Todd L. Prospective review of 30-day morbidity and mortality in a paediatric neurosurgical unit. Childs Nerv Syst 2017;33(3):483–489 10. Kuo BJ, Vissoci JR, Egger JR, et al. Perioperative outcomes for pediatric neurosurgical procedures: analysis of the National Surgical Quality Improvement Program-Pediatrics. J Neurosurg Pediatr 2017;19(3):361–371 11. Nargozian C. The airway in patients with craniofacial abnormalities. Paediatr Anaesth 2004;14(1):53–59 12. Shenkman Z, Rockoff MA, Eldredge EA, Korf BR, Black PM, Soriano SG. Anaesthetic management of children with tuberous sclerosis. Paediatr Anaesth 2002;12(8):700–704

305

306

IVa Anesthesia 13. Goobie SM, DiNardo JA, Faraoni D. Relationship between transfusion volume and outcomes in children undergoing noncardiac surgery. Transfusion 2016;56(10):2487–2494 14. Goobie SM, Soriano SG, Zurakowski D, McGowan FX, Rockoff MA. Hemostatic changes in pediatric neurosurgical patients as evaluated by thrombelastograph. Anesth Analg 2001;93(4):887–892 15. Soriano SG, Martyn JAJ. Antiepileptic-induced resistance to neuromuscular blockers: mechanisms and clinical significance. Clin Pharmacokinet 2004;43(2):71–81 16. Soriano SG, Sullivan LJ, Venkatakrishnan K, Greenblatt DJ, Martyn JA. Pharmacokinetics and pharmacodynamics of vecuronium in children receiving phenytoin or carbamazepine for chronic anticonvulsant therapy. Br J Anaesth 2001;86(2):223–229 17. Groeper K, McCann ME. Topiramate and metabolic acidosis: a case series and review of the literature. Paediatr Anaesth 2005;15(2):167–170 18. Valencia I, Pfeifer H, Thiele EA. General anesthesia and the ketogenic diet: clinical experience in nine patients. Epilepsia 2002;43(5):525–529 19. McCann ME, Kain ZN. The management of preoperative anxiety in children: an update. Anesth Analg 2001;93(1):98–105 20. Gibert S, Sabourdin N, Louvet N, et al. Epileptogenic effect of ­sevoflurane: determination of the minimal alveolar concentration of sevoflurane associated with major epileptoid signs in children. Anesthesiology 2012;117(6):1253–1261 21. Grady MS, Bedford RF, Park TS. Changes in superior sagittal sinus pressure in children with head elevation, jugular venous compression, and PEEP. J Neurosurg 1986;65(2):199–202 22. Westergaard B, Classen V, Walther-Larsen S. Peripherally inserted central catheters in infants and children—indications, techniques, complications and clinical recommendations. Acta Anaesthesiol Scand 2013;57(3):278–287 23. Sponheim S, Skraastad Ø, Helseth E, Due-Tønnesen B, Aamodt G, Breivik H. Effects of 0.5 and 1.0 MAC isoflurane, sevoflurane and desflurane on intracranial and cerebral perfusion pressures in children. Acta Anaesthesiol Scand 2003;47(8):932–938 24. Modica PA, Tempelhoff R, White PF. Pro- and anticonvulsant effects of anesthetics (Part I) Anesth Analg 1990;70(3):303–315 25. Modica PA, Tempelhoff R, White PF. Pro- and anticonvulsant effects of anesthetics (Part II) Anesth Analg 1990;70(4):433–444 26. Scheingraber S, Rehm M, Sehmisch C, Finsterer U. Rapid saline infusion produces hyperchloremic acidosis in patients undergoing gynecologic surgery. Anesthesiology 1999;90(5):1265–1270 27. McManus ML, Soriano SG. Rebound swelling of a ­stroglial cells exposed to hypertonic mannitol. Anesthesiology 1998; 88(6):1586–1591

28. Reasoner DK, Todd MM, Scamman FL, Warner DS. The incidence of pneumocephalus after supratentorial craniotomy. Observations on the disappearance of intracranial air. Anesthesiology 1994;80(5):1008–1012 29. Cossu M, Schiariti M, Francione S, et al. Stereoelectroencephalography in the presurgical evaluation of focal epilepsy in infancy and early childhood. J Neurosurg Pediatr 2012;9(3):290–300 30. LaRiviere MJ, Gross RE. Stereotactic laser ablation for medically intractable epilepsy: the next generation of minimally invasive epilepsy surgery. Front Surg 2016;3:64 31. Wilfong AA, Curry DJ. Hypothalamic hamartomas: o ­ptimal approach to clinical evaluation and diagnosis. Epilepsia 2013;54(Suppl 9):109–114 32. Cossu M, Fuschillo D, Casaceli G, et al. Stereoelectroencephalography-guided radiofrequency thermocoagulation in the epileptogenic zone: a retrospective study on 89 cases. J Neurosurg 2015;123(6):1358–1367 33. Medvid R, Ruiz A, Komotar RJ, et al. Current applications of MRI-guided laser interstitial thermal therapy in the treatment of brain neoplasms and epilepsy: a radiologic and neurosurgical overview. AJNR Am J Neuroradiol 2015;36(11):1998–2006 34. Adelson PD, Black PM, Madsen JR, et al. Use of subdural grids and strip electrodes to identify a seizure focus in children. Pediatr Neurosurg 1995;22(4):174–180 35. Sarang A, Dinsmore J. Anaesthesia for awake craniotomy—­ evolution of a technique that facilitates awake neurological testing. Br J Anaesth 2003;90(2):161–165 36. Stevanovic A, Rossaint R, Veldeman M, Bilotta F, Coburn M. Anaesthesia management for awake craniotomy: systematic review and meta-analysis. PLoS One 2016;11(5):e0156448 37. Soriano SG, Eldredge EA, Wang FK, et al. The effect of propofol on intraoperative electrocorticography and cortical stimulation during awake craniotomies in children. Paediatr Anaesth 2000;10(1):29–34 38. Brian JE Jr, Deshpande JK, McPherson RW. Management of cerebral hemispherectomy in children. J Clin Anesth 1990;2(2):91–95 39. Goobie SM, Meier PM, Pereira LM, et al. Efficacy of tranexamic acid in pediatric craniosynostosis surgery: a double-blind, ­placebo-controlled trial. Anesthesiology 2011;114(4):862–871 40. Faraoni D, Goobie SM. The efficacy of antifibrinolytic drugs in children undergoing noncardiac surgery: a systematic review of the literature. Anesth Analg 2014;118(3):628–636 41. Mason KP, Lerman J. Review article: dexmedetomidine in children: current knowledge and future applications. Anesth ­ Analg 2011;113(5):1129–1142

32

  Pediatric Awake Craniotomy Gaston Echaniz, Michael Tan, Ibrahim Jalloh, Samuel Strantzas, and Tara Der

Summary Operating on lesions within or near eloquent areas may result in permanent speech and motor deficits. Pediatric awake craniotomy with direct intraoperative mapping allows for maximal and increased functional resection with considerable reduction in postoperative neurological deficits. The patient must have a high level of endurance, motivation, and psychological preparedness to participate in the procedure. Many urgent intraoperative anesthetic complications may occur. The success of the technique depends on proper preoperative and anesthetic planning that includes careful patient selection, neuropsychological assessment, psychiatric evaluation, and psychological preparation. Critical steps for anesthetic management of awake craniotomy are discussed in this chapter, based on current evidence and institutional experience. Keywords:  awake craniotomy, pediatric anesthesia, epilepsy surgery, tumor resection, eloquent areas

„„ Introduction Operating on lesions within or adjacent to eloquent areas carries the risk of permanent neurological deficits. Awake craniotomy with direct intraoperative brain mapping is the gold standard for identifying eloquent areas.1 Mapping different cortical pathways allows for maximal and a more functional resection with considerable reduction in postoperative speech and motor deficits.2 Sacko et al prospectively compared patients who underwent lesionectomy in eloquent areas under awake surgery with intraoperative mapping and the patients in whom the surgical resection was performed under general anesthesia. They reported that neurological outcome, extent of resection, and survival at 80 months were significantly better in patients who underwent awake craniotomy. These patients had shorter hospital stay, no significant anesthetic complications, and no need for conversion to general anesthesia.3 Awake craniotomies are well described in the adult literature and are routinely used in adults undergoing tumor, epilepsy, and functional neurosurgery.2,​4,​5,​6 However, there are limited data in children undergoing awake craniotomy. This may be related to the epidemiology of brain tumors in children as most are not in eloquent areas. Few cases and limited series have been

published and many aspects still need to be elucidated.7,​8,​9 The majority of cases are performed using the asleep-awake-asleep approach whereby the child is awoken for mapping and excision of the lesion and is asleep for the craniotomy and surgical closure. The ethical barriers of this procedure include negative psychological experience and possible emotional distress.10 Anesthesia for pediatric awake craniotomy presents many challenges to the anesthesiologist who must provide an anesthetic that ensures optimal analgesia, patient cooperation with minimal sedation, minimal psychological trauma, and allows for satisfactory intraoperative monitoring.11 The anesthesiologist must create a safe and calm environment for the child to obtain maximal cooperation and be prepared to deal with different urgent or emergent intraoperative complications.12 Some of the complications reported in adults are airway obstruction, respiratory depression, hemodynamic instability, nausea and vomiting, disinhibited behavior, seizures, and pain.6,​13 Therefore, the success of the technique, in children, is largely focused on proper preoperative planning and anesthetic preparation, including careful patient selection, neuropsychological assessment, psychiatric evaluation, and psychological preparation.

„„ Patient Selection and Evaluation After a child is identified by neurosurgery as having a lesion amenable to awake craniotomy, a series of evaluations must be followed. These include multidisciplinary consultations with neurology, epilepsy team, neuropsychology, neuropsychiatry, and anesthesia. One of the most important conditions for deciding the appropriateness for awake craniotomy is the neuropsychological or cognitive development of the patient. It is imperative that candidates for awake craniotomy show a high level of endurance and motivation and a psychological capacity to participate.14,​15,​16 An acceptable age minimum of 11 or 12 years has been suggested for pediatric awake craniotomies;12 however, the appropriateness should be based on a child’s level of neuropsychological development including maturity and capacity. Children as young as 9 years of age enduring awake craniotomy for resection of lesions such as glioblastoma have been described.17 In our institution, the youngest patient for awake craniotomy was a 7-year-old who successfully underwent a left frontal tumor resection. Therefore, no prohibition

308

IVa Anesthesia to awake craniotomy below a minimum age should be imposed. Cases should be considered individually taking into account the patient’s maturity and special requirements of the resection. A dedicated multidisciplinary team is the cornerstone in identifying and optimizing the child for this challenging procedure.

„„ Anesthetic Evaluation At our center, a detailed consultation with the anesthesiologist performing the procedure is needed to establish patient rapport. The assessment should be focused on the medical history, details regarding seizure characteristics, and neurological examination findings. Current medications should be evaluated for ongoing perioperative seizure control. One relative contraindication to awake surgery is difficult airway. Patient positioning, often with a neck flexion while the head in surgical head pins, makes intraoperative airway management difficult. Therefore, a detailed airway assessment is vital. Building a close relationship with patients and their families is essential to build trust. Detailed information should be provided about the various stages of the surgery, making the patient aware of what they can expect to experience, and the role they will play. Potential complications such as seizures and airway difficulties should be discussed. Pictures and videos of the procedure are shown to familiarize the child with the operating room setup and tasks involved.

„„ Neuropsychological Evaluation Neuropsychological assessment by a neuropsychologist before, during, and after surgery identifies cognitive level and speech or memory deficits. A full neuropsychological assessment measuring intellectual abilities (IQ), expressive or receptive or ­higher-order language skills, memory skills, visual–motor and visual–spatial skills, problem solving, manual dexterity, executive function, as well as academics (reading, spelling, writing, and math) are performed. Since the most concerning complication is injury to the eloquent cortex, a focus should be the patient’s language skill testing. Preexisting deficits must be known to compare any intraoperative changes. The neuropsychological assessment is also a great opportunity to assess their conversational skills, their ability to answer questions, and to identify topics the child can fluently discuss in the operating room. Finally, this is an important opportunity to build rapport and trust with the child.

„„ Psychological Preparation Several reports of awake craniotomies in pediatric patients have been published though most are small series in teenaged patients.9,​10,​17,​18,​19,​20,​21,​22 While older teenagers may have a similar physiology to an adult, they may be psychologically immature and ill prepared for this type of procedure compared to an adult. Adults presenting for awake craniotomy can easily understand its benefit. Thus, they are often highly ­motivated and

c­ ooperative to successfully minimize perioperative ­deficits. The greatest challenge for pediatric awake craniotomy is a ­ dopting techniques for younger children given their wide spectrum of neuropsychological and cognitive development. In addition to the same complications described in adults, issues such as ­agitation, restlessness, fear, anxiety, and lack of cooperation might be more frequent in this population. Therefore, psychological screening for suitability and extensive preparation for the procedure is paramount. Various strategies have been described in order to improve the psychological preparation, such as hypnosis conditioning, meeting a child who has undergone an awake craniotomy, showing pictures or videos describing the atmosphere of the operating room, and meeting the surgical and anesthetic team well before surgery.10 Psychological preparation has been shown to reduce the stress in children undergoing minor p ­ rocedures.23 Klimek 17 et al described an intensive psychological care for a child in their report, and they were able to keep a 9-year-old patient lightly sedated and cooperative during an awake craniotomy. Hypnosis has also been described to be effective and easily performed technique to improve the psychological experience and cooperation of children, especially in those with anxious personality or depression.10 Parental presence throughout the operative procedure might also be considered to reduce anxiety and improve patient’s cooperation.17 In this case, parents must receive a detailed explanation of the procedure and how to react in case of any intraoperative complications. Should there be concerns about parental ability to cooperate or facilitate the procedure, parental presence should be discouraged.

„„ Intraoperative Technique Operating Room Setup Operating room ergonomics and equipment movement must be optimized prior to patient arrival. The majority of brain lesions in awake surgery cases are left sided, but some are right sided. This information should be detailed prior to room preparation as they require mirror image setups. A typical setup is demonstrated in Fig. 32.1 and Video 32.1. The overall aim is to allow sufficient space for sterile surgical equipment and access, room for the scrub nurse, direct airway, facial, and line access for the anesthesiologist to manage the airway and hemodynamics, space for patient comfort and communication, and space for the neuropsychologist. Whether solely sedated or asleep-awake-asleep techniques are used, the anesthesiologist must be prepared to convert to a full general endotracheal anesthesia if needed. Airway management might be particularly challenging due to the patient’s

Video 32.1 Awake craniotomy and cortical ­stimulation. (This video is provided courtesy of Tara Der.) https://www.thieme.de/de/q.htm?p=opn/tp/ 255910102/9781626238176_c032_v001&t=video

32  Pediatric Awake Craniotomy

Fig. 32.1  Scalp innervation.

flexed and rotated neck position and limited range of motion while the patient’s head is pinned. Airway tools including supraglottic devices such as an laryngeal mask airway (LMA), oral and nasal airways, masks, a fiberoptic bronchoscope, as well as induction drugs should be prepared.24 After placement of any lines and airway devices in supine position, the patient is repositioned in a semilateral or lateral position. Ideally, lines should be placed on the nonoperative side for easier intraoperative access. Although patient ­comfort is essential, unobstructed access to the airway is mandatory. Minimizing neck flexion and rotation with proper body positioning helps to decrease airway obstruction and allows for better c­ erebral jugular venous return. Careful positioning and padding of limbs minimize patient discomfort during the awake portion of surgery. Blanket and fluid warming devices should be provided to avoid shivering during surgical resection. Intravenous poles and Mayo stand supports are placed to allow tenting of surgical drapes over the patient’s upper body. This prevents claustrophobia and allows for communication and direct access to the face and airway.

Anesthetic Technique Adult awake craniotomies are often performed with the patient awake for the entire procedure with varying degrees of sedation (awake-awake-awake). Pediatric awake craniotomies are usually performed with an asleep-awake-asleep technique and patients remain awake for lesion mapping and excision only. Premedication with sedatives is quite controversial, and the decision should be based on the patient’s anxiety, comorbidities, and proposed anesthetic technique. In our experience, anxiolytic premedication is usually not required for the asleep-awake-asleep technique as patients are highly motivated and well prepared for the procedure. We also offer parental presence for induction, as there is literature suggesting decreased perioperative anxiety for some patients.25,​26,​27 Furthermore, the patient is usually under general anesthesia for stimulating portions of surgery, and the persistence of benzodiazepines or alpha-2 agonists used for p ­ remedication

can lead to delayed wake-up, excessive sedation, and poor cooperation during the awake phase. Acetaminophen 15 mg/kg orally preoperatively should be part of a multimodal ­analgesic approach.28,​29

Induction and Initial Asleep Phase For the asleep-awake-asleep technique, anesthetic induction should follow the same principles as other craniotomies to ensure minimizing rises in intracranial pressures and maintenance of cerebral perfusion. Again, due to patient selection, intravenous catheter placement is usually well tolerated with topical anesthetic creams. Inhalational induction is seldom required. Muscle relaxants are usually avoided, since this may interfere with motor evoked potentials and cause residual muscle weakness during the awake phase. Intravenous induction can be achieved with 1 to 2 μg/kg fentanyl followed by 2 to 3 mg/kg propofol. Antibiotic prophylaxis and nonsedating antiemetic combinations such as dexamethasone 0.1 mg/kg and ondansetron 0.1 mg/kg are also given during induction. If inhalational induction is performed, volatile agents must be washed out quickly to minimize interference with neuromonitoring. An additional large-bore intravenous catheter is placed in case of unexpected bleeding and an arterial line for blood pressure monitoring and blood sampling is usually required. A urinary catheter is inserted if surgical time is prolonged but may be avoided as this adds to patient discomfort during the awake phase. LMA is the preferred choice of airway management and has been reliably used in both the adult and pediatric awake ­craniotomy literature.20,​30,​31,​32,​33,​34 It is easily removed during the awake phase and more readily reinserted for the final asleep phase of the procedure (compared to endotracheal intubation). Some centers have attempted endotracheal intubation; ­however, this is more likely to cause undesirable coughing and injury in a patient with an exposed craniotomy site and skull fixed in pins.35,​36 Compared to sedation and use of nasal prongs, airway stabilization with an LMA allows for controlled ventilation particularly if pCO2 control is needed.

309

310

IVa Anesthesia

Regional Anesthesia for Awake Craniotomy Absence of pain is a cornerstone to facilitate good cooperation to better psychological experience. Therefore, integral to the s­ uccess of the awake phase is the adequacy of regional ­anesthesia.37 Six nerves supply sensation to the scalp: supratrochlear, supraorbital, zygomaticotemporal, auriculotemporal, and the lesser and greater occipital nerves. Furthermore, in rare situations, the greater auricular nerve and third occipital nerve can supply sensation to areas of the scalp in the surgical field (Fig. 32.2).38 The scalp block is performed once the child is asleep prior to insertion of pins. A 25-gauge needle is used to infiltrate 0.25% bupivacaine hydrochloride with epinephrine 1:200,000, which provides up to 8 to 12 hours of analgesic action. Infiltration begins anteriorly, just above the eyebrow in the region of the supraorbital nerve, using the supraorbital notch as a landmark. The needle is directed medially to also cover the supratrochlear nerve. Next, the anterior temporal region is infiltrated lateral to the orbit to target the zygomaticotemporal branches of the maxillary division of the trigeminal nerve. Further infiltration targets the auriculotemporal nerve at approximately

1 cm anterior to the tragus at the level of the zygomatic arch, being cautious to not infiltrate below the zygoma, which could partially block the facial nerve. The great auricular nerves are s­ ubsequently infiltrated just posterior to the pinna at the ­level of the zygomatic arch. Finally, the greater, lesser, and third occipital nerves are collectively injected by inserting the needle at the midpoint between the inion and mastoid, along the superior nuchal line, infiltrating toward the pinna. Supplementation of the scalp block with pin site infiltration should be considered. Local anesthetic toxicity is not an issue with this regimen if weight-based dosing is followed. During the awake phase, the anesthesiologist and surgeon must be prepared to supplement local anesthesia below and above the drapes.

Maintenance of Anesthesia Maintenance of anesthesia can be achieved with total intravenous anesthesia (TIVA) consisting of propofol (100– ­ 150 μg/kg/min) and remifentanil (0.1–0.4 μg/kg/min) with doses closely titrated and mirrored to electroencephalogram (EEG) depth. These drugs are ideal as they are easy to titrate and result in a cooperative patient with rapid wakening in 5 to

Fig. 32.2  Typical setup of an operating room for a left-sided brain lesion. A mirror image setup is required for right-sided lesions.

32  Pediatric Awake Craniotomy 20 minutes, with minimal residual sedation during the awake phase of the procedure. Propofol also has antiemetic and anticonvulsant properties and reduces intracranial pressure and cerebral oxygen consumption.39 Furthermore, propofol infusion at the dose range used for awake craniotomy does not appear to interfere with electrocorticography (ECoG) if it is stopped 20 minutes before recording for pediatric patients.21,​40 Remifentanil has a very short context-sensitive half-life even after ­prolonged infusion, which is minimally altered by age or kidney or liver dysfunction. Similar to propofol, remifentanil does not appear to interfere with ECoG monitoring.41,​42 Other centers have also used benzodiazepines and neuroleptic analgesia in the past17,​35,​36 and alpha-2 agonists have recently gained popularity.8,​19,​33,​43,​44,​45,​46 However, neuroleptic analgesia (droperidol + fentanyl) has been associated with prolonged sedation, seizures, QT prolongation, decreasing patient cooperation, and predisposition to cardiac dysrhythmias. Dexmedetomidine is a highly selective alpha-2-adrenergic receptor agonist with dose-dependent anxiolytic, sedative, and analgesic effects. Thus, it allows for a unique “cooperative sedation” that can be easily reversed with verbal stimulation and without respiratory depression.46,​47,​48 Additionally, dexmedetomidine may provide neuroprotective and sympatholytic effects that allow for hemodynamic stability at critical moments of neurosurgical stimulation.49 Hence, it may be a useful sedative agent in patients who undergo complex neurocognitive testing during awake craniotomy.50 No differences have been found in terms of quality of intraoperative brain mapping and efficacy of sedation with dexmedetomidine, though there are fewer respiratory adverse events when compared to remifentanil–propofol in adults undergoing awake craniotomy for supratentorial tumor resection.51 Dexmedetomidine has shown to provide effective and safe conscious sedation during awake craniotomy with a shorter arousal time than propofol infusion.52 However, in our experience, dexmedetomidine was associated with a prolonged wakening time. Of 23 cases in our center, dexmedetomidine was added to our standard TIVA in two cases, and both had a prolonged wake-up time in excess of 50 minutes at doses of 0.2 to 1.8 μg/kg/h. Wakening

times with only remifentanil–­propofol on average was 21 minutes (unpublished data, submission in progress). Depth of anesthesia monitoring is advised to allow more accurate titration of TIVA and enables more accurate timing of patient wakening. Monitors such as bispectral index are used in some centers, but we find that raw EEG monitoring is sufficient for this purpose.35

Electroencephalography for Depth of Anesthesia and Detection of Afterdischarges Emergence from anesthesia for the awake portion is monitored by utilizing raw EEG, which provides a gauge for depth of anesthesia. Monitoring the depth of anesthesia using commercial devices, such as the bispectral index monitor, can be used; however, we have found the use of raw EEG to be more accurate, particularly in the pediatric population. There are many advantages to using raw EEG; raw EEG allows for immediate identification of depth without acquisition delay. It also enables the interpreter to quickly identify artifacts such as electromyography, electrocautery, and 60 Hz interference, which could ­otherwise contaminate readings obtained from commercial devices. In addition, no proprietary electrodes or special equipment is required and, if recording somatosensory evoked potentials, the same electrodes can be used to monitor EEG. Finally, the ability to monitor depth of anesthesia from each hemisphere of the brain is also possible using raw EEG. EEG is derived from the collective activity of the cerebral cortex and measures various electrical frequencies. Typically, with deepening anesthesia, there is a predominance of high-­ amplitude delta (1–3 Hz) and theta (4–7 Hz) slow wave activity, and a loss of low-amplitude, irregular alpha (8–12 Hz) and beta (> 12 Hz) fast activity. This pattern is reversed with the return of consciousness, and as faster frequencies predominate the EEG, the patient begins to emerge from anesthesia in preparation for speech testing (Fig. 32.3).

Fig. 32.3  Variations in raw EEG, as the patient moves from a deeper state of anesthetic depth to an awake state. Three channels of longitudinal bipolar EEG are recorded in each 5-second snapshot. The upper graph shows a higher concentration of slow delta and theta waves, which diminish (middle graph) and eventually disappear (lower graph) as the patient emerges from anesthesia. The awake state contains predominantly low-amplitude, high-frequency irregular beta activity.

311

312

IVa Anesthesia Although a well-trained eye is needed to ensure EEG r­ ecordings are meaningful and not contaminated with artifact, ­interpreting EEG for depth of anesthesia is a skill that requires a handful of cases to become comfortable with. Barnard et al53 showed that after a brief didactic session, anesthetists were able to categorize EEGs as anesthetized, sedated, or awake, with precise accuracy.

Awake Phase Once the surgeon has performed the craniotomy, TIVA is reduced in close accordance to EEG depth. After the dura is opened and secured, preparations are made for the awake phase of the procedure. Propofol and remifentanil infusions are paused and nonsedating antiemetics, such as ondansetron, are given or re-dosed. Once the child begins spontaneous respirations, a “no-touch” technique is used to reduce patient coughing and bucking. The LMA is removed once the child is awake and supplemental oxygen by mask or nasal prongs is provided. During the awake phase of the procedure, the eloquent and/or motor cortex areas of the brain are tested constantly. The requirement for a fully alert and cooperative patient means sedating drugs are avoided, if possible, during testing. Preferably, a neurologist or neuropsychologist (or often the anesthesiologist), who is familiar with the patient baseline deficits, is present to assist with testing. Intraoperatively, voluntary motor movement and motor commands are assessed, as well as cognitive and language testing. Depending on the patient’s age and language skills, we use different tasks such as pictorial cue cards, counting, and complex arithmetic skills (e.g., count down by 3 from 100), memory games (previously identified during preoperative visits), and constant speech and motor tasks. Pain scores are frequently obtained to ensure patient’s comfort during this stage. The patient is encouraged to speak loudly, and ambient noise in the room is reduced (e.g., lower monitor alarm volumes), so the surgical team can follow the testing in progress. Any observed changes in function must be communicated immediately to the surgical team. At the conclusion of the awake phase after lesionectomy, anesthesia is induced once again by a bolus dose of propofol, and TIVA is restarted and titrated to EEG depth. A new nonsoiled LMA is replaced and the anesthesiologist resumes control of ventilation. Alternatively, in a cooperative and adequately blocked child, the propofol infusion can be resumed and closure done under sedation with nasal pongs (asleep-awake-awake).

Pain commonly arises from pin sites. If regional scalp nerve block and pin site infiltration by surgeon or anesthesiologist do not relieve the pain, intravenous opioid is administered. Restarting the remifentanil infusion at this stage is ideal because oversedation is quickly reversed by reducing or stopping the infusion. Intraoperative seizures during awake craniotomies are difficult to predict, and the incidence ranges from 5 to 20% in the adult population.56 This risk seems to be higher in younger patients with low-grade glioma in the frontal lobe and previous history of seizures.54 Considering the potentially devastating consequences of a patient having generalized tonic–clonic movements while fixed in head pins, the anesthesiologists should be prepared to treat them promptly, and the surgical team must be extremely alert to the initiation of seizure activity. Antiepileptic drugs should be optimized before the procedure and reloaded intraoperatively if necessary. Cold saline for brain irrigation should be prepared and placed close to the surgical field. If a seizure occurs, flooding of the surgical field with cold irrigation can abort the seizure.36 If necessary, benzodiazepines such as 0.1 mg/kg midazolam can be administered and loading with phenytoin or fosphenytoin consid-

Table 32.1  Intraoperative complications and management

Complication

Management

Seizures

• Irrigation of surgical field with cold saline irrigation • Medications

Pain

• Local anesthetic to pin sites • Intravenous opioid, e.g., restarting remifentanil infusion • Conversion to general anesthesia

Nausea/vomiting

• Antiemetics as prophylaxis and/or treatment, e.g., dexamethasone and ondansetron

Airway compromise

• Prevent with careful positioning • Minimize sedation • Judicious drug treatment of seizures and anxiety • Airway cart prepared

Raised ICP

• Decrease sedation if hypoventilation leads to raised pCO2 • Ask patient to hyperventilate • Airway management and assisted ventilation • Osmolar therapy, e.g., mannitol

„„ Intraoperative Complications

Patient anxiety/poor cooperation

The attending anesthesiologist must be prepared for a wide range of complications during all phases of surgery but particularly, the awake phase. Complications are summarized in Table 32.1. Some of the reported intraoperative complications in adults include pain, seizures, respiratory depression, airway obstruction, hemodynamic instability, vomiting, and disinhibited behaviors.6,​13,​14,​24,​54,​55 Despite lack of such data for children, the anesthesiologist should be prepared to address and treat these situations.

• Consider decreasing or increasing sedation depending on the clinical situation • Address any physical discomfort • Provide psychological/emotional support • Treat symptoms of pain, nausea, or vomiting • Consider abandoning awake procedure and converting to general anesthesia

Abbreviation: ICP, intracranial pressure.

32  Pediatric Awake Craniotomy ered. Mapping should be paused until speech or motor abilities recover. The ­neurosurgeon should identify the area that triggered the event to avoid restimulation. Treatment with small doses of propofol has been reported, but the risk of oversedation leading to airway obstruction and hypoxia is high. If seizure activity continues, conversion to general anesthesia and airway instrumentation should ensue. The team may then elect to abort the procedure as planned. Airway complications include obstruction and subsequent hypoxia following seizure or oversedation. A nasopharyngeal airway may be sufficient to relieve obstruction, but inducing general anesthesia with replacement of an LMA may be necessary. If pulmonary aspiration occurs following vomiting, airway protection with an endotracheal tube is difficult in a patient with an exposed craniotomy site and head secured in pins. The anesthesiologist may need to perform either oral or nasal endotracheal intubation through the LMA or fiberoptic-assisted intubation.24 As a last alternative, surgeons should be prepared to cover the field and come out of pins for direct laryngoscopy and intubation. Hypercapnia due to sedation and subsequent hypoventilation can lead to brain edema in rare situations. Controlling

the pCO2 in a conscious patient with an unsecured airway is difficult. A cooperative patient may be able to assist by hyperventilating upon request; however, it is also possible that this situation has arisen due to an oversedated patient who is ­ unable to follow commands. If this occurs, pharmacological management with mannitol or hypertonic saline is often more practical. Hemodynamic instability, including hypertension, hypotension, and tachycardia, is more common using the asleepawake-asleep technique compared with the awake-awake-awake technique, but rarely causes harm to patients when occurring in isolation. Improved hemodynamic stability occurs with maintenance of low-dose sedation, but runs the increased risk of respiratory complications.

„„ Conclusion Awake craniotomy in children is vitally necessary in certain neurosurgical procedures, including some epilepsy cases. Anesthetic management of these cases, although challenging, is attainable with careful planning and execution.

References 1. Ojemann G, Ojemann J, Lettich E, Berger M. Cortical language localization in left, dominant hemisphere. An electrical stimulation mapping investigation in 117 patients. J Neurosurg 1989;71(3):316–326

13. Danks RA, Rogers M, Aglio LS, Gugino LD, Black PM. Patient tolerance of craniotomy performed with the patient under local anesthesia and monitored conscious sedation. Neurosurgery 1998;42(1):28–34, discussion 34–36

2. Ebel H, Ebel M, Schillinger G, Klimek M, Sobesky J, Klug N. Surgery of intrinsic cerebral neoplasms in eloquent areas under local anesthesia. Minim Invasive Neurosurg 2000;43(4):192–196

14. Nossek E, Matot I, Shahar T, et al. Failed awake craniotomy: a retrospective analysis in 424 patients undergoing craniotomy for brain tumor. J Neurosurg 2013;118(2):243–249

3. Sacko O, Lauwers-Cances V, Brauge D, Sesay M, Brenner A, Roux FE. Awake craniotomy vs surgery under general anesthesia for resection of supratentorial lesions. Neurosurgery 2011;68(5):1192– 1198, discussion 1198–1199

15. Khu KJ, Doglietto F, Radovanovic I, et al. Patients’ perceptions of awake and outpatient craniotomy for brain tumor: a qualitative study. J Neurosurg 2010;112(5):1056–1060

4. Dziedzic T, Bernstein M. Awake craniotomy for brain tumor: indications, technique and benefits. Expert Rev Neurother 2014;14(12):1405–1415 5. Sahjpaul RL. Awake craniotomy: controversies, indications and techniques in the surgical treatment of temporal lobe epilepsy. Can J Neurol Sci 2000;27(Suppl 1):S55–S63, discussion S92–S96 6. Archer DP, McKenna JM, Morin L, Ravussin P. Conscious-sedation analgesia during craniotomy for intractable epilepsy: a review of 354 consecutive cases. Can J Anaesth 1988;35(4):338–344 7. Akay A, Rükşen M, Çetin HY, Seval HO, İşlekel S. Pediatric awake craniotomy for brain lesions. Pediatr Neurosurg 2016;51(2): 103–108 8. Ard J, Doyle W, Bekker A. Awake craniotomy with dexmedetomidine in pediatric patients. J Neurosurg Anesthesiol 2003;15(3):263–266 9. Balogun JA, Khan OH, Taylor M, et al. Pediatric awake craniotomy and intra-operative stimulation mapping. J Clin Neurosci 2014;21(11):1891–1894 10. Delion M, Terminassian A, Lehousse T, et al. Specificities of awake craniotomy and brain mapping in children for resection of supratentorial tumors in the language area. World Neurosurg 2015;84(6):1645–1652

16. Leal RT, da Fonseca CO, Landeiro JA. Patients’ perspective on awake craniotomy for brain tumors-single center experience in Brazil. Acta Neurochir (Wien) 2017;159(4):725–731 17. Klimek M, Verbrugge SJ, Roubos S, van der Most E, Vincent AJ, Klein J. Awake craniotomy for glioblastoma in a 9-year-old child. Anaesthesia 2004;59(6):607–609 18. Berger MS, Kincaid J, Ojemann GA, Lettich E. Brain mapping techniques to maximize resection, safety, and seizure control in children with brain tumors. Neurosurgery 1989;25(5):786–792 19. Everett LL, van Rooyen IF, Warner MH, Shurtleff HA, Saneto RP, Ojemann JG. Use of dexmedetomidine in awake craniotomy in adolescents: report of two cases. Paediatr Anaesth 2006;16(3):338–342 20. Hagberg CA, Gollas A, Berry JM. The laryngeal mask airway for awake craniotomy in the pediatric patient: report of three cases. J Clin Anesth 2004;16(1):43–47 21. Soriano SG, Eldredge EA, Wang FK, et al. The effect of propofol on intraoperative electrocorticography and cortical stimulation during awake craniotomies in children. Paediatr Anaesth 2000;10(1):29–34 22. Tobias JD, Jimenez DF. Anaesthetic management during awake craniotomy in a 12-year-old boy. Paediatr Anaesth 1997;7(4):341–344

11. Frost EA, Booij LH. Anesthesia in the patient for awake craniotomy. Curr Opin Anaesthesiol 2007;20(4):331–335

23. Kolk AM, van Hoof R, Fiedeldij Dop MJ. Preparing children for venepuncture. The effect of an integrated intervention on distress before and during venepuncture. Child Care Health Dev 2000;26(3):251–260

12. McClain CD, Landrigan-Ossar M. Challenges in pediatric neuroanesthesia: awake craniotomy, intraoperative magnetic resonance imaging, and interventional neuroradiology. Anesthesiol Clin 2014;32(1):83–100

24. Matsuda A, Mizota T, Tanaka T, Segawa H, Fukuda K. [Difficult ventilation requiring emergency endotracheal intubation during awake craniotomy managed by laryngeal mask airway] Masui 2016;65(4):380–383

313

314

IVa Anesthesia 25. Kain ZN, Mayes LC, Caramico LA, et al. Parental presence during induction of anesthesia. A randomized controlled trial. Anesthesiology 1996;84(5):1060–1067 26. Sadeghi A, Khaleghnejad Tabari A, Mahdavi A, Salarian S, Razavi SS. Impact of parental presence during induction of anesthesia on a ­ nxiety level among pediatric patients and their parents: a randomized clinical trial. Neuropsychiatr Dis Treat 2017;12:3237–3241 27. Kita T, Yamamoto M. [Parental presence is a useful method for smooth induction of anesthesia in children: a postoperative questionnaire survey] Masui 2009;58(6):719–723 28. Wick EC, Grant MC, Wu CL. Postoperative multimodal analgesia pain management with nonopioid analgesics and techniques: a review. JAMA Surg 2017;152(7):691–697 29. Lönnqvist PA, Morton NS. Postoperative analgesia in infants and children. Br J Anaesth 2005;95(1):59–68 30. Tongier WK, Joshi GP, Landers DF, Mickey B. Use of the laryngeal mask airway during awake craniotomy for tumor resection. J Clin Anesth 2000;12(8):592–594 31. Murata H, Nagaishi C, Tsuda A, Sumikawa K. Laryngeal mask airway Supreme for asleep-awake-asleep craniotomy. Br J Anaesth 2010;104(3):389–390 32. Shinokuma T, Shono S, Iwakiri S, Shigematsu K, Higa K. [Awake craniotomy with propofol sedation and a laryngeal mask airway: a case report] Masui 2002;51(5):529–531 33. Chung YH, Park S, Kim WH, Chung IS, Lee JJ. Anesthetic management of awake craniotomy with laryngeal mask airway and dexmedetomidine in risky patients. Korean J Anesthesiol 2012;63(6):573–575 34. Gadhinglajkar S, Sreedhar R, Abraham M. Anesthesia management of awake craniotomy performed under asleep-awake-asleep technique using laryngeal mask airway: report of two cases. Neurol India 2008;56(1):65–67 35. Sarang A, Dinsmore J. Anaesthesia for awake craniotomy— evolution of a technique that facilitates awake neurological testing. Br J Anaesth 2003;90(2):161–165 36. Meng L, McDonagh DL, Berger MS, Gelb AW. Anesthesia for awake craniotomy: a how-to guide for the occasional practitioner. Can J Anaesth 2017;64(5):517–529 37. Guilfoyle MR, Helmy A, Duane D, Hutchinson PJ. Regional scalp block for postcraniotomy analgesia: a systematic review and meta-analysis. Anesth Analg 2013;116(5):1093–1102 38. Kemp WJ III, Tubbs RS, Cohen-Gadol AA. The innervation of the scalp: a comprehensive review including anatomy, pathology, and neurosurgical correlates. Surg Neurol Int 2011;2:178 39. Marik PE. Propofol: therapeutic indications and side-effects. Curr Pharm Des 2004;10(29):3639–3649 40. Herrick IA, Craen RA, Gelb AW, et al. Propofol sedation during awake craniotomy for seizures: patient-controlled administration versus neurolept analgesia. Anesth Analg 1997;84(6):1285–1291

41. Beers R, Camporesi E. Remifentanil update: clinical science and utility. CNS Drugs 2004;18(15):1085–1104 42. Herrick IA, Craen RA, Blume WT, Novick T, Gelb AW. Sedative doses of remifentanil have minimal effect on ECoG spike activity during awake epilepsy surgery. J Neurosurg Anesthesiol 2002;14(1):55–58 43. Gignac E, Manninen PH, Gelb AW. Comparison of fentanyl, sufentanil and alfentanil during awake craniotomy for epilepsy. Can J Anaesth 1993;40(5, Pt 1):421–424 44. Bulsara KR, Johnson J, Villavicencio AT. Improvements in brain tumor surgery: the modern history of awake craniotomies. Neurosurg Focus 2005;18(4):e5 45. Almeida AN, Tavares C, Tibano A, Sasaki S, Murata KN, Marino R Jr. Dexmedetomidine for awake craniotomy without laryngeal mask. Arq Neuropsiquiatr 2005;63(3B):748–750 46. Bekker AY, Kaufman B, Samir H, Doyle W. The use of dexmedetomidine infusion for awake craniotomy. Anesth Analg 2001;92(5):1251–1253 47. Bekker A, Sturaitis MK. Dexmedetomidine for neurological surgery. Neurosurgery 2005;57(1, Suppl):1–10, discussion 1–10 48. Hall JE, Uhrich TD, Barney JA, Arain SR, Ebert TJ. Sedative, amnestic, and analgesic properties of small-dose dexmedetomidine infusions. Anesth Analg 2000;90(3):699–705 49. Kondavagilu SR, Pujari VS, Chadalawada MV, Bevinguddaiah Y. Low dose dexmedetomidine attenuates hemodynamic response to skull pin holder application. Anesth Essays Res 2017;11(1):57–61 50. Mack PF, Perrine K, Kobylarz E, Schwartz TH, Lien CA. Dexmedetomidine and neurocognitive testing in awake craniotomy. J Neurosurg Anesthesiol 2004;16(1):20–25 51. Goettel N, Bharadwaj S, Venkatraghavan L, Mehta J, Bernstein M, Manninen PH. Dexmedetomidine vs propofol-remifentanil conscious sedation for awake craniotomy: a prospective randomized controlled trial. Br J Anaesth 2016;116(6):811–821 52. Shen SL, Zheng JY, Zhang J, et al. Comparison of dexmedetomidine and propofol for conscious sedation in awake craniotomy: a prospective, double-blind, randomized, and controlled clinical trial. Ann Pharmacother 2013;47(11):1391–1399 53. Barnard JP, Bennett C, Voss LJ, Sleigh JW. Can anaesthetists be taught to interpret the effects of general anaesthesia on the electroencephalogram? Comparison of performance with the BIS and spectral entropy. Br J Anaesth 2007;99(4):532–537 54. Nossek E, Matot I, Shahar T, et al. Intraoperative seizures during awake craniotomy: incidence and consequences: analysis of 477 patients. Neurosurgery 2013;73(1):135–140, discussion 140 55. Skucas AP, Artru AA. Anesthetic complications of awake craniotomies for epilepsy surgery. Anesth Analg 2006;102(3):882–887 56. Sartorius CJ, Wright G. Intraoperative brain mapping in a community setting—technical considerations. Surg Neurol 1997;47(4):380–388

33

  Implantation of Strip, Grid, and Depth Electrodes for Invasive Electrophysiological Monitoring Oğuz Çataltepe and Julie G. Pilitsis

Summary Intracranial monitoring techniques are frequently used in epilepsy surgery procedures at the pediatric age group. Intracranial monitoring is especially helpful when the preoperative data obtained with noninvasive monitoring techniques are inconclusive or provide insufficient information for precise localization of epileptogenic zone (EZ). The main contribution and value of extraoperative monitoring with intracranial electrodes in epilepsy surgery is its role in the precise determination and mapping of the EZ and eloquent cortex. The most commonly used electrodes in intracranial monitoring are subdural strip, grid, and depth electrodes. Subdural grid electrodes are excellent choices for covering a large cortical area in its entirety to record both interictal and ictal epileptogenic activity and to perform extraoperative cortical stimulation and mapping. Depth electrodes are most valuable in assessing deep cortical structures such as amygdala, hippocampus, cingulum, insula, and orbitofrontal cortex. Invasive monitoring techniques may also include a combination of depth, strip, and grid electrodes, and the precise configuration is determined by individual patient ­anatomo-electro-clinical (AEC) findings. It is prudent to know the advantages and limitations of each invasive monitoring technique to choose the most appropriate modality for individual patients. The surgeon and epilepsy team must be aware of the advantages, limitations, and potential complications of each technique and must be versatile with them in order to provide the best possible electrophysiological data to define the best surgical approaches in these cases. Keywords:  intracranial monitoring, strip electrode, grid electrode, depth electrode

„„ Introduction The main goal of resective surgical intervention in the management of medically refractory epilepsy is the complete removal of the EZ while preserving functionally critical cortical areas. Therefore, the precise mapping of the AEC network responsible from seizures and determining its relation to eloquent cortex has critical significance for achieving a postoperative seizure-free outcome.1,​2,​3,​4,​5 Preoperative investigations aim to

define AEC characteristics of EZ, its relations with structural lesions and eloquent cortex, as well as to determine the feasibility of surgical resection of the EZ without any new neurological deficit. Naturally, precise localization and mapping the extent of the EZ as well as its functional status are crucial for good surgical outcome and also preserving neurological function.5,​6 Although scalp electroencephalography (EEG) is very helpful in determining the location of the epileptogenic zone, it does not delineate the surgical target satisfactorily in many MRI-­ negative cases. A report from the Cleveland stated that noninvasive studies achieved the goal of identifying the EZ in around 70% of the patients who are operated on in 2012. In the remaining 30%, either formulation of a clear AEC hypothesis was not possible, or the precise location of EZ relative to eloquent cortex was not defined satisfactorily for a safe surgical resection.5 These patients are frequently candidates for invasive monitoring. The main contribution and value of extraoperative monitoring with intracranial electrodes in epilepsy surgery is its role in the precise determination and mapping of the epileptogenic zone and eloquent cortex. Invasive monitoring is required frequently in pediatric age group because of the high frequency of extratemporal epilepsy and cortical dysplasia. In these cases, preoperative data obtained with noninvasive monitoring techniques may be inconclusive or provide insufficient information for precise localization of EZ. This group constitutes 25 to 40% of cases in some pediatric series.2,​3,​7,​8,​9 Although invasive monitoring in pediatric population has been used commonly, application of these techniques in infants is still limited, although increasing.4,​10 Invasive monitoring in infants is tending to increase due to difficulty in interpreting clinical seizure semiology and imaging findings in immature brains. Further, there is a high frequency of extratemporally localized EZ in infants.10

„„ Indications Invasive monitoring is indicated if preoperative assessment with noninvasive tests results in the following:5 • MRI-negative cases: ◦◦ Electroclinical information is suggestive for focal seizures but fails to identify the EZ with sufficient accuracy for surgical planning.

316

IVb  Intracranial Electrode Placement for Invasive Monitoring yy MRI shows a lesion but: ◦◦ Its location is not concordant with electroclinical hypothesis. ◦◦ Lesion borders are ill-defined. ◦◦ Electroclinical evidence implies a larger epileptogenic zone than lesional zone seen on MRI, and simple lesionectomy would not provide adequate seizure control. ◦◦ Structural lesion is adjacent to eloquent cortex. yy MRI shows large focal, multifocal, hemispheric, or bilateral abnormalities but: ◦◦ Electroclinical data imply more localized or lateralized ictal onset. ◦◦ Working hypothesis is limited focal resection of epileptogenic area might be sufficient for good seizure control. yy Multiple lesions or dual pathologies are present but: ◦◦ One or both of them are discordant with electroclinical hypothesis. ◦◦ It is not clear which one is epileptogenic. yy Noninvasive electrographical tests: ◦◦ Are inconclusive, such as: –– Presence of electrographic data suggestive for unilateral temporal onset along bilateral imaging abnormalities. –– Ill-defined electrographic abnormalities with a unilateral structural abnormality on imaging studies. ◦◦ Provide poorly defined lateralization or localization of epileptogenic zone: –– Semiological and electrographic difficulty in differentiating temporal lobe seizures from possible insular or frontal lobe seizures. ◦◦ Provide noncongruent data. ◦◦ Show multiple epileptiform areas.2,​3,​6,​7,​8,​9,​11,​12,​13 yy EZ involves to eloquent cortex.

„„ Invasive Monitoring Techniques Several invasive monitoring electrodes may be used in epilepsy patients. The most commonly used electrodes are ­subdural strip, grid, and depth electrodes (Fig. 33.1). Epidural

peg electrodes have fallen out of favor. Invasive monitoring electrodes can be chosen among commercially available electrodes or can be custom-made based on the specific needs of the case. These can be used separately or in combination to determine the EZ. The extent of coverage, appropriate electrode types and configuration are determined after reviewing the data for each patient. The initial step in invasive monitoring is determining the cortical area with the highest likelihood of epileptogenicity. Scalp EEG and video monitoring, MRI, ­magnetoencephalography (MEG), and functional imaging s­ tudies such as positron emission tomography (PET) and single-­photon emission computed tomography (SPECT) provide the most valuable data in defining the cortical area that needs to be covered.8,​9 Invasive electrodes may be placed bilaterally, but if lateralization of the seizures is clear but localization is questionable, coverage may be limited with cortical areas in one hemisphere.6

Subdural Strip and Grid Electrodes Both subdural strip and grid electrodes are thin, biologically inert, Silastic or Teflon sheets with embedded nickel–­ chromium or platinum, electrically isolated electrode contacts. Each electrode contact is 2 to 4 mm in diameter, and generally the interelectrode distance is 5 to 10 mm. Subdural strips are single row of contact electrodes (1 × 4, 1 × 6, 1 × 8, etc.). Subdural grids are larger plates of rectangular arrays with several parallel rows of up to 64 electrodes (2 × 4, 2 × 6, 4 × 8, 8 × 8, etc.). Both subdural strips and grid electrodes are very thin and flexible sheets and can be contoured to the underlying cortical surface. They are also transparent and allow visualization of the underlying anatomical structures. There are many commercially available strip and grids with various configurations and shapes, such as curvilinear, dualfaced grid electrodes for interhemispheric coverage. Custom-made options are also readily available. Strip electrodes are placed through burr holes or can be slid under craniotomy edges if they will be used in combination with grid electrodes (Fig. 33.2).

Fig. 33.1  Subdural strip and grid electrodes. (This image is provided courtesy of Integra Neurosciences, Plainsboro, New Jersey, USA.)

33  Implantation of Strip, Grid, and Depth Electrodes for Invasive Electrophysiological Monitoring Multiple subdural strip electrodes using different trajectories to sample large cortical areas can be placed through a s­ ingle burr hole. Strip electrodes are ideal to lateralize and grossly localize the most suspicious cortical areas for seizure onset. Grid electrode placement is performed with a large craniotomy (Fig. 33.3). Grid electrode recording is an ideal technique to locate the EZ and to stimulate and map the adjacent eloquent cortex by covering large cortical surfaces completely. Subdural grid electrodes are frequently used in pediatric epilepsy surgery to cover large extratemporal cortical areas.7,​8,​9,​11,​14 However, their application in infants, especially children younger than 2 years of age, is limited because of size of the grids and relative fullness of the infant brain.4

Depth Electrodes Depth electrodes are most valuable in assessing deep cortical structures such as amygdala, hippocampus, cingulum, insu-

Fig. 33.2  Interhemispheric grid and frontal multiple strip electrode placement through a frontal craniotomy.

la, and orbitofrontal cortex.15 Depth electrodes in pediatric ­ pilepsy surgery are not used as frequent as subdural grid elece trodes, and when they are used, it is frequently associated with subdural grid electrodes. However, their indications have been changed, and its application has been increasing in both adults and children parallel to popularization of the stereoelectroencephalography (SEEG) technique in North America. SEEG is discussed in detail in Chapters 17 and 34. Depth electrodes are multicontact electrode arrays, consisting of up to 16 nickel–chromium or platinum contacts embedded in a thin, tubular, biologically inert Silastic material. They can be placed through a drill hole using a stereotactic frame, frameless neuronavigation guidance, or robotic surgical assistant systems. Depth electrodes are typically used to record electrographical activities from deep structures. However, they do not provide data form cortical surface as good as subdural grids, and they have limited stimulation capabilities. Therefore, depth electrodes are used in combination with subdural strip and grid electrodes in some cases.7 If the preferred approach is not SEEG, then depth electrodes combined with subdural electrodes provide good coverage for both cortical surface and deep structures.7,​11,​15 Depth electrodes can be placed into targets such as mesial temporal structures through two different approaches: occipitotemporal and temporal (Fig. 33.4 and Fig. 33.5). Occipitotemporal approach provides a trajectory parallel to the long axis of hippocampus, whereas the temporal approach provides a trajectory orthogonal to hippocampus. Occipitotemporal approach has the advantage of placing multiple contacts throughout the amygdala, hippocampal head and body with a single depth ­electrode. Because this approach would not p ­ rovide any information regarding neocortical electrical activity, additional subdural strips or grids may need to be placed to cover temporal neocortex at the same time. Conversely, the temporal orthogonal approach has an advantage of providing data both from ­mesial as well as neocortical structures through depth ­electrode. However, this approach often necessitates multiple depth electrodes and results in only one or two c­ ontacts truly in the mesial structures, whereas the other contacts are in white matter and neocortex. This approach provides very limited neocortical ­electrographic data compared with subdural electrodes.

Fig. 33.3  (a) Intraoperative photograph of frontal multiple grid placement. Note the absence of any gap between the two grid electrode plates. (b) Postoperative skull X-ray showing large grid coverage on the right frontal convexity and interhemispheric space as well as strip sampling from the left frontal cortex.

317

318

IVb  Intracranial Electrode Placement for Invasive Monitoring

Fig. 33.4  Postoperative MRI images of bilateral hippocampal depth electrodes through occipitotemporal approach: axial (a), sagittal (b), and coronal (c) planes.

Fig. 33.5  Unilateral hippocampal depth electrode placement through lateral temporal approach. Postoperative MRI images: axial (a), sagittal (b), and coronal (c) planes.

Advantages and Limitations of Invasive Monitoring Techniques Invasive electrode recording has many advantages as well as limitations. One of the main advantages of invasive monitoring is the reliability of the electrophysiological recording, because it is directly obtained from a small number of neurons with stable impedances with higher amplitudes and also because signal attenuation caused by scalp and skull, as well as muscle artifacts are eliminated.9,​11,​16 Other advantages are being able to detect EEG activity from deep structures such as the amygdala, hippocampus, basal frontal or temporal cortex, and interhemispheric cortical surfaces. In addition to these, providing an opportunity to perform extraoperative stimulation and cortical mapping is also very beneficial, especially in the pediatric age group.9,​14,​16 It is prudent to know the advantages and limitations of each invasive monitoring technique to choose the most appropriate modality for individual patients. The main advantage of subdural strip electrodes is their versatility. They can be easily placed through a small burr hole without requiring any special equipment or expertise. Multiple strip electrodes can be placed in different trajectories through a single burr hole, and large cortical areas including basal temporal and frontal cortex can be sampled. Strip electrodes can even be used for mesial temporal lobe coverage and provide valuable information, although they record from the parahippocampus rather than hippocampus.

Conversely, the placement of strip electrodes is not as precise, and the location of the electrodes can easily be suboptimal because the placement is done blindly. Neuronavigation techniques may be used to increase the reliability of the placement, but it may still easily deviate from intended trajectory. Strip electrodes can provide limited information. They can only provide cortical sampling data because their coverage over the cortex is discontinuous and limited over a linear trajectory. Injury to cortex, cortical veins, and associated bleeding are main risks for strip electrodes. The risk of cerebrospinal fluid (CSF) leak and infection is also much higher with strip electrodes compare to depth electrodes.11,​14 Subdural grid electrodes are excellent choices for covering a large cortical area in its entirety to record both interictal and ictal epileptogenic activity and to perform extraoperative cortical stimulation and mapping. However, grids can be placed satisfactorily only after seizures are grossly localized with other modalities.14 Although extraoperative cortical stimulation and mapping ability is one of the greatest advantages of grid electrodes, limited spatial and cortical sampling, especially due to incomplete or inadequate coverage of intrasulcal, deep brain, and interhemispheric spaces, constitutes its weakness. It does not provide a three-dimensional large functional network sampling in contrast to SEEG.5 Also, subdural grid electrode placement in small children can be challenging because of bulky volume and associated risks. Additionally, value of stimulation and mapping data obtained through subdural grid electrodes

33  Implantation of Strip, Grid, and Depth Electrodes for Invasive Electrophysiological Monitoring in young children is controversial. There are number of challenges for cortical stimulation and mapping in young children. Absence of responses to cortical stimulation in children younger than 4 years old does not indicate nonfunctional cortex, and cortical stimulation results in this age group are inconsistent, at best. Motor response may be elicited up to 25% of children younger than 3 years of age, and identification of language cortex with extraoperative cortical stimulation in children younger than 10 years old is not always reliable.7,​17,​18 Subdural grid electrode placement is performed with a large craniotomy and thus exposes the patient to considerable postoperative risks, including cerebral edema, mass effect, cortical injury, bleeding, and related problems. All these risks are much more substantial in young children because of relative fullness of the intracranial contents and limited blood volume. The ideal candidate for subdural grid electrode placement is a patient with a well-defined superficial cortical lesion who also needs cortical stimulation and functional mapping because of a location adjacent to eloquent cortex. However, this technique should be used cautiously if the patient had a previous craniotomy for diagnostic or excisional purposes in the same area. If the targeted coverage aims interhemispheric cortical surfaces, basal frontal or temporal cortex, insula, deep sulcal areas, or deep structures such as hippocampus, then depth electrodes or SEEG might be more appropriate. Depth electrodes are excellent choices for recording from deep structures. They may be placed with a small stab incision and a drill hole. They can even be removed at bedside easily without going back to the operating room.7 The most significant advantage of depth electrodes is their ability to place the electrodes precisely to an intended deep-seated target and record directly from these structures, including amygdala, hippocampus, and insula. This is especially relevant in mesial temporal lobe epilepsy (TLE) and insular epilepsy, because seizure activities originating from these structures may spread very rapidly, and thus seizure onset may not be clearly defined with surface recording alone.14 Conversely, the placement of depth electrodes requires sophisticated equipment such as a stereotactic frame, neuronavigation, or robotic surgical assistant systems as well as surgical expertise and experience with these techniques, and higher associated cost. Invasive monitoring techniques have many advantages as described previously; however, they also have an inherent bias secondary to limited cortical sampling (tunnel vision) that is prone to false localization. Because of limited spatial coverage, invasive monitoring techniques may define the EZ incompletely if the targeted coverage defined by AEC hypothesis for EZ is less than perfect. Electrographic abnormal activity detected by intracranial electrodes can also easily be a propagated activity more than a true seizure onset.11 While a single scalp electrode can record electrical activity from a relatively large cortical area (~ 6 cm2), a single subdural or depth electrode contact can cover only a few square millimeters of cortical area.13 This limited sampling provides perfect recording at the contact area but can easily miss seizure onset from adjacent areas. It should also be emphasized that subdural grid and strip coverage is basically a surface sampling limited to the crown portion of the cortex. The majority of the cortex in covered areas stays buried in the sulci below the surface or under opercular surfaces such as insular cortex. Therefore, strip and grid electrodes do not have any direct contact with cortical tissues in these embedded cortical areas. Precise placement to cover the intended area is quite

difficult with subdural grid and strip electrodes if coverage of interhemispheric and basal cortical surfaces of the brain is targeted. In addition to this, placement of subdural grid and strips in these areas may also be challenging and carry risk because of limited visual exposure, presence of bridging veins and adhesions. On the other hand, depth electrode placement is the i­ deal method for precise placement and excellent recording from these deep structures. Another well-known potential technical problem of invasive monitoring is signal detection problems secondary to orientation of the angle of the dipole.1,​7,​8,​9 Other disadvantages and limitations are cost and patient discomfort secondary to the surgical procedure and immobility during the postoperative monitoring period. Each procedure also carries standard neurosurgical risks of morbidity, and even mortality, and two separate surgical interventions are generally required. It is also possible that invasive monitoring still may not provide any localizing information at the end of the monitoring period and further cortical resection to treat seizures may not be an option for the patient. This rate has been reported as 12 to 34% in some large series.1,​2,​3,​7,​8

„„ Surgical Technique Subdural Strip Electrode Placement Subdural strip electrodes are placed through burr holes. If strip electrodes are placed to cover temporal lobe, a burr hole is placed 2 cm above the zygoma, just anterior to pinna. Then, dura is opened carefully while protecting the underlying arachnoid. The temporal strip electrodes are gently slid into the s­ ubdural space using smooth forceps and continuous irrigation. Mannitol (0.25 g/kg) and Decadron might be given if the ­subdural space is not comfortable. A Penfield #3 dissector (Codman Inc., Raynham, MA) may be helpful during strip electrode placement to guide the electrodes to intended trajectory. We place six to eight contact electrodes to cover temporal pole and anterior parahippocampus. The electrode is placed on a trajectory toward the medial portion of the temporal pole just lateral to the sphenoid wing so that it curves down at the temporal pole and lies under the anterior parahippocampus. Then, a second electrode with six contacts is placed, perpendicular to the sylvian fissure and toward base of the middle fossa to cover the middle, inferior, and fusiform gyri with the most distal electrode being under the middle posterior portion of the parahippocampus. The third electrode (four to six contacts) is placed posteriorly to cover the posterior temporal region over the middle superior temporal gyri. If the strip electrode does not slide in smoothly, it should be pulled back and reoriented to a slightly different trajectory for a second attempt.3,​16 Neuronavigation guidance would be quite helpful in cases more precise topographic placement is needed. Strip electrode for extratemporal coverage can be used especially to sample frontopolar, lateral, and mesial frontal cortical surfaces. The burr hole location is determined in these cases individually on the basis of the targeted cortical areas and planned coverage sites. After placement of strip electrodes, the cables are tunneled 3- to 4 cm away from incision using an angiocath or specially designed tunnelers. The burr hole is plugged with ­Gelfoam (Pfizer Inc, New York, NY). The galea and scalp layers are closed separately.

319

320

IVb  Intracranial Electrode Placement for Invasive Monitoring

Subdural Grid Electrode Placement The patient’s head is either placed on a horseshoe head holder or secured with three-point fixation system in supine position. The extent of skin incision and scalp flap is determined by considering the electrode coverage area as well as the projected craniotomy for possible resective surgery in the future. The surgical procedure for grid placement is a standard craniotomy and is performed under general anesthesia. A generous bone flap is removed to expose and to cover a larger area than presumed epileptogenic region and surrounding cortex. Dura is opened largely, and tacking sutures are placed. At this stage, neuronavigation guidance can also be used again to determine the cortical coverage area more precisely. Then the grid plate is placed over the cortex with smooth forceps. Although it is possible to slide the grid plate slightly beyond the craniotomy edges, this type of blind advancement should be avoided as much as possible to decrease cortical injury and bleeding risk. Subdural grid electrodes can also be used to cover interhemispheric cortical surfaces. Dual-sided grid electrodes can be used for interhemispheric bilateral coverage. We place our craniotomy flap twothirds anterior and one-third posterior to the coronal suture in these cases if the plan is to place only interhemispheric grid and a few subdural strips (Fig. 33.6). Although grid electrodes can often be slid into the interhemispheric space easily, we place interhemispheric grid electrodes after fully dissecting and exploring the interhemispheric space because of the frequent presence of adhesions and bridging veins in this area. We also prefer to expose and dissect the interhemispheric space all the way down to corpus callosum to cover cingulum as well. Both cingulate gyri are almost always somewhat attached to each other, and if the subdural grid is placed blindly by sliding the plate without separating both gyri, the edge of the grid electrode may either stay above cingulum or injure them. After subdural grid placement is completed, the edges of the plate should be checked carefully to avoid any compression on large cortical or bridging veins. If the cortical area covered by the grid plate is a site of previous surgery or trauma, or if there is an underlying mass lesion, more caution is needed while placing grid electrodes and during the closure. It is also critical to be sure that all electrodes firmly contact with cortical surface because grid plates can be folded under the bone. Large grid electrode plates are prone to buckling, therefore dividing the grid along the electrode arrays may help to reshape the curve of the plate for a firm contact with the cortical surface.

In some cases, adjacent grid–grid or grid–strip combinations may be needed to gain more extensive coverage. In these cases, these adjacent electrodes should be placed, so that no space is present between subdural electrodes to prevent herniation of brain or cortical vasculature between the edges of the two subdural electrode plates (Fig. 33.3a). After placement of the subdural grid electrode, cables are tunneled greater than 5 cm away from the incision lines using an angiocath or specifically designed cable tunnelers. Then a digital photo is taken to document the grid position, and dural closure is performed. Duraplasty is almost always needed to have a comfortable ­closure using a watertight technique as much as possible. Then, EEG recording is performed by checking each electrode contact separately, and then bone flap is replaced with loose sutures if recording is satisfactory. At this stage, it is necessary to ensure that leads are not compressed between bone edges. We perform meticulous hemostasis before skin closure to avoid any need for subgaleal drain placement. Some surgeons prefer leaving a subdural drain for CSF drainage over several days as a preventive measure in case of cerebral edema. Then the galea and skin are closed as separate layers with appropriate sutures. All cable exit points are closed with purse-string sutures, and all cables are marked and numbered separately to identify the related electrodes. A bulky head dressing is applied.

Intraparenchymal Depth Electrode Placement Intraparenchymal depth electrode placement in epilepsy patients has traditionally been performed with s­tereotactic frames. However, frameless stereotactic techniques using ­neuronavigation system guidance and robotic surgical assistance have been becoming increasingly popular in recent years.8,​19 Although, application of a stereotactic frame in the pediatric age group is limited by age, some groups have been placing depth electrodes even in infants.10 We initially placed depth electrodes using stereotactic frame (Leksell Elekta, S ­ tockholm, Sweden) in our patients, but more recently, we have been using the ROSA (MedTech, Montpellier, France) robotic assistance ­system, especially for SEEG cases. Frameless s­ ystems have many advantages including ability to use in younger patients, more freedom to choose appropriate trajectories, c­ oncurrent use in guiding precise strip electrode trajectories and grid placement in the same setting, and eliminating the need to remove or work around

Fig. 33.6  Mesial frontal cortex coverage with interhemispheric dual-sided grid electrodes and bilateral frontal convexity coverage with multiple strip electrodes. (a) Head computed tomography scan shows grid electrode contacts in the interhemispheric space and multiple strip electrode contacts in different trajectories on the frontal convexities. (b) Skull X-ray shows interhemispheric grid plates and multiple bilateral strip electrodes covering frontal convexities.

33  Implantation of Strip, Grid, and Depth Electrodes for Invasive Electrophysiological Monitoring the stereotactic frame if craniotomy is needed.8,​19 Conversely, the stereotactic frame has a very reliable track record and ­precision for depth electrode placement. Placing depth electrodes with a traditional stereotactic frame also takes longer than SEEG with the ROSA robotic assistant system. In the procedures with stereotactic frame, brain MRI with gadolinium is obtained a few days before surgery date. Then the stereotactic frame is attached to the patient’s head in the operating room under general anesthesia, and the patient is transferred to the CT scanner. A head CT scan with contrast is obtained after the stereotactic localizer is attached to the frame, and the head is secured to the scanner table. The digital data are then transferred via Ethernet to computer workstation, and MRI and CT scan images are fused using image fusion software. The most appropriate trajectories for depth electrodes are subsequently chosen. Image fusion enables the surgeon to combine CT scan’s precision for targeting with the enhanced neuroanatomy and multiplanar imaging capabilities of MRI. However, multiple depth electrode placement with a stereotactic frame is quite cumbersome and long. For multiple depth electrodes, such as SEEG cases, the ROSA gives much more freedom both during trajectory planning and also during actual surgery. Here, we will describe hippocampal depth electrode placement using the Leksell stereotactic system, and next chapter will discuss SEEG technique. Intraparenchymal depth electrodes with 10 to 16 contacts can be placed via occipitotemporal approach in TLE to maximize the number of electrode contacts in the amygdala and hippocampus. First, target and entry points are selected, and then the trajectory is defined. The ideal trajectory avoids cortical vascular structures and does not pass through sulci or the ventricle. The target point of the most distal electrode contact is in amygdala, and the other contacts lay in hippocampal head and body (Fig. 33.4). This may not always be possible in some cases because of the individual anatomical variations, and some compromises may be needed while choosing the entry point and trajectory.12 The trajectory is checked in axial, sagittal, and coronal slices carefully using the probe’s eye view on the workstation, and the entry and target points are subsequently slightly altered, if necessary, to obtain a safer trajectory. After the most appropriate trajectory is chosen, coordinates are calculated and recorded. Simultaneously, the patient’s head is attached to the operating table using table attachment for the stereotactic frame, and the patient’s neck is slightly flexed. The head of the bed is elevated up to 45 degrees for a semisitting position. This position provides a perfect exposure to the parieto-occipital region. The surgical site is prepared and draped with a transparent adhesive cover. Then the appropriate coordinates are chosen on the frame and arc, and a guide cannula is inserted to the scalp to mark the entry point. A small linear scalp incision is made for a drill hole. The dura is coagulated and opened. Next, a rigid guide cannula and its obturator are placed to target; then, the obturator is removed, and the depth electrode is inserted through the cannula. The cannula is then removed while holding the depth electrode and its semirigid stylet steadily. Then, the stylet of the depth electrode is removed. Then the depth electrode is secured in place by tightening the bold cap or alternatively, if bolt has not been used, it is tunneled greater than 3 cm away from the incision using either a specially designed tunneler or a 14-gauge angiocath and is

tied down to scalp with 3–0 silk. Then a purse-string suture, using 3–0 silk, passing through the galea is placed around the cable to decrease risk for CSF leak. Then the electrode cable is again sutured to scalp by making a small loop.

„„ Postoperative Management Patients stay in pediatric intensive care unit for the first postoperative day and then are transferred to epilepsy monitoring unit. Patients receive pre- and postoperative intravenous antibiotics. We also use steroids for the first 3 postoperative days in subdural grid cases. Postoperative skull X-rays and a head CT scan are obtained to document the electrode positions. Digital photography and supplemental drawings of the electrode locations drawn by the surgeon are placed into the patient file as well as the neurophysiology file. We also obtain a postoperative brain MRI, especially in-depth electrode cases, to see the electrode contact locations more precisely. Although MRI is a concern because of potential heating or electrical injury to brain parenchyma, we have not encountered any issues to date; these results were confirmed in a large study showing that MRI is safe for invasive monitoring patients.20 Alternatively, postoperative CT scan can be coregistered with preoperative MRI as well. Postoperative images are significant to determine precise location of the electrode contacts, so the epilepsy surgery team can appreciate the anatomical relationships between lesion, EZ, and eloquent cortex better to plan safe and effective cortical resection. Duration of monitoring and the ideal seizure numbers necessary to complete a monitoring period are somewhat controversial. We aim to record at least three habitual seizures before removing the electrodes, and we verify the seizure types with family members to be sure these are typical seizures. If the patient does not have the sufficient number of seizures, then provocative measures such as sleep deprivation are considered along with discontinuation of medications. Patients are closely monitored for CSF leak, and if any leak is noted, additional sutures are placed. We do not use lumbar drains to prevent CSF leak. After obtaining satisfactory electrophysiological data, the patient returns to the operating room for removal of the electrodes with or without cortical resection, and all removed ­electrodes are sent for culture.

„„ Complications The most common complications related to intracranial ­electrodes, both in adult and children, are subdural ­hematoma and infection.1,​12,​14,​16,​21 Although not that common, other reported complications are CSF leak, cerebral edema, and neurological deficits. The rate of common complications varies widely depends on series and type of intracranial electrodes. Overall complication rate for subdural grid and strip electrodes was reported between 0 and 21.4%, and this rate was 2.1 to 13.6% for intraparenchymal depth electrodes.11,​22,​23,​24 Overall complication rate for intracranial electrodes is mostly around 9%, with less than 5% of patients has neurological deficits and only less than 1% being permanent deficit.22 A meta-analysis in a pooled series of 2,542 cases with subdural electrodes

321

322

IVb  Intracranial Electrode Placement for Invasive Monitoring showed 5.3% infection rate, 3% of that being superficial infections and 4% for intracranial hemorrhage.24 In another large series, authors reviewed 269 patients (both adult and children) undergoing 317 procedures for invasive monitoring with both subdural and depth electrodes.22 Overall, the complication rate was found 9.1%. Although 4.1% resulted in a neurological deficit, 3.5% of these resolved in the postoperative period, and only 0.6% exhibited permanent deficits. Clinically significant hemorrhage rate in this series was 1.9% for subdural electrodes and 0.6% for depth electrodes. Infection rate was 4.7% for subdural electrodes and 3.8% for depth electrodes.22 There are just a limited number of series reviewing complications specifically in children.1,​2,​3,​4,​5,​6,​7,​10,​21 Infection rates in these series for subdural electrodes are reported as 3.2 to 7%, and the subdural hematoma rate was 0.8 to 14%.18 The Cleveland Clinic series for invasive monitoring in children reported complication rate as 3%.5 Delalande et al reported their e ­ xperience with subdural and SEEG depth electrodes in 26 children younger than 3 years old, and complication rate was reported as 3%.9 Another series published by Cossu et al reviewed SEEG depth electrode outcome in children younger than 4 years of age, and there was no infection or hemorrhage in this series of 15 patients, but

there was one mortality secondary to profound hyponatremia and cerebral edema in a child.25 Overall, subdural electrodes appear to have higher rates of postoperative subdural hemorrhage and nonhemorrhagic extra-axial collections in postoperative radiographic studies, although vast majority of these are clinically insignificant extra-axial collections.

„„ Conclusion In conclusion, invasive monitoring constitutes a significant portion of epilepsy surgery procedures in the pediatric age group and may provide essential additional data in selected patients. Invasive monitoring techniques often include a combination of depth, strip, and grid electrodes, and the precise configuration is determined by individual patient AEC findings. The surgeon and epilepsy team must be aware of the advantages, limitations, and potential complications of each technique and must be versatile with them in order to provide the best possible electrophysiological data to define the best surgical approaches in these cases.

References 1. Johnston JM Jr, Mangano FT, Ojemann JG, Park TS, Trevathan E, Smyth MD. Complications of invasive subdural electrode monitoring at St. Louis Children’s Hospital, 1994–2005. J Neurosurg 2006;105(5, Suppl):343–347 2. Simon SL, Telfeian A, Duhaime AC. Complications of invasive monitoring used in intractable pediatric epilepsy. Pediatr Neurosurg 2003;38(1):47–52 3. Bruce DA, Bizzi JWJ. Surgical technique for the insertion of grids and strips for invasive monitoring in children with intractable epilepsy. Childs Nerv Syst 2000;16(10–11):724–730 4. Bingaman WE, Bulacio J. Placement of subdural grids in pediatric patients: technique and results. Childs Nerv Syst 2014;30(11):1897–1904 5. Gonzalez-Martinez J, Najm IM. Indications and selection criteria for invasive monitoring in children with cortical dysplasia. Childs Nerv Syst 2014;30(11):1823–1829 6. Munari C, Lo Russo G, Minotti L, et al. Presurgical strategies and epilepsy surgery in children: comparison of literature and personal experiences. Childs Nerv Syst 1999;15(4):149–157 7. Adelson PD, O’Rourke DK, Albright AL. Chronic invasive monitoring for identifying seizure foci in children. Neurosurg Clin N Am 1995;6(3):491–504 8. Blount JP, Cormier J, Kim H, Kankirawatana P, Riley KO, ­Knowlton RC. Advances in intracranial monitoring. Neurosurg Focus 2008;25(3):E18 9. Sperling MR. Clinical challenges in invasive monitoring in epilepsy surgery. Epilepsia 1997;38(Suppl 4):S6–S12 10. Taussig D, Dorfmüller G, Fohlen M, et al. Invasive explorations in children younger than 3 years. Seizure 2012;21(8):631–638 11. Diehl B, Lüders HO. Temporal lobe epilepsy: when are invasive recordings needed? Epilepsia 2000;41(Suppl 3):S61–S74 12. Blatt DR, Roper SN, Friedman WA. Invasive monitoring of limbic epilepsy using stereotactic depth and subdural strip electrodes: surgical technique. Surg Neurol 1997;48(1):74–79 13. Dubeau F, McLachlan RS. Invasive electrographic recording techniques in temporal lobe epilepsy. Can J Neurol Sci 2000;27(Suppl 1):S29–S34, discussion S50–S52

14. Salazar F, Bingaman WE. Placement of subdural grids. In: ­Luders HO, ed. Textbook of Epilepsy Surgery. London: Informa; 2008:931–937 15. Mulligan L, Vives K, Spencer D. Placement of depth electrodes. In: Luders HO, ed. Textbook of Epilepsy Surgery. London: Informa; 2008:938–944 16. Hamer HM, Morris HH, Mascha EJ, et al. Complications of invasive video-EEG monitoring with subdural grid electrodes. Neurology 2002;58(1):97–103 17. Schevon CA, Carlson C, Zaroff CM, et al. Pediatric language ­mapping: sensitivity of neurostimulation and Wada testing in epilepsy surgery. Epilepsia 2007;48(3):539–545 18. Taussig D, Montavont A, Isnard J. Invasive EEG explorations. ­Neurophysiol Clin 2015;45(1):113–119 19. Chamoun RB, Nayar VV, Yoshor D. Neuronavigation applied to epilepsy monitoring with subdural electrodes. Neurosurg Focus 2008;25(3):E21 20. Davis LM, Spencer DD, Spencer SS, Bronen RA. MR imaging of implanted depth and subdural electrodes: is it safe? Epilepsy Res 1999;35(2):95–98 21. Onal C, Otsubo H, Araki T, et al. Complications of invasive subdural grid monitoring in children with epilepsy. J Neurosurg 2003;98(5):1017–1026 22. Schmidt RF, Wu C, Lang MJ, et al. Complications of subdural and depth electrodes in 269 patients undergoing 317 procedures for invasive monitoring in epilepsy. Epilepsia 2016;57(10):1697–1708 23. Wellmer J, von der Groeben F, Klarmann U, et al. Risks and benefits of invasive epilepsy surgery workup with implanted subdural and depth electrodes. Epilepsia 2012;53(8):1322–1332 24. Arya R, Mangano FT, Horn PS, Holland KD, Rose DF, Glauser TA. Adverse events related to extraoperative invasive EEG monitoring with subdural grid electrodes: a systematic review and meta-­analysis. Epilepsia 2013;54(5):828–839 25. Cossu M, Schiariti M, Francione S, et al. Stereoelectroencephalography in the presurgical evaluation of focal epilepsy in infancy and early childhood. J Neurosurg Pediatr 2012;9(3):290–300

34

  Stereoelectroencephalography in Children: Methodology and Surgical Technique Robert A. McGovern and Jorge A. Gonzalez-Martinez

Summary Stereoelectroencephalography (SEEG) is a safe and efficient extraoperative invasive methodology designed to define the anatomical boundaries of the cortical and subcortical cerebral regions of the epileptogenic zone (EZ). Both frame-based and frameless techniques can be utilized for implantation of SEEG electrodes. While children pose some unique challenges in the planning and execution of electrode implantation, obtaining vascular imaging and attention to vascular anatomy when planning trajectories are essential to reduce the risk of hemorrhagic complications. While children have not been studied as well as adults, most studies have found that children have comparable outcomes with similar types of morbidity. Centers utilizing SEEG are typically able to localize the EZ in approximately 90% of patients. Of the patients who undergo resective surgery, 55 to 70% are seizure free at 2 to 3 years of follow-up. The main morbidity related to SEEG in children is hemorrhage, which appears to be 1 to 3%. As a result, the main clinical challenge for the near future remains in the further refinement of specific selection criteria for the different methods of invasive monitoring, with the ultimate goal of comparing and validating the results (long-term seizure-free outcome) obtained from different methods of invasive monitoring. Keywords:  epilepsy surgery, stereoelectroencephalography (SEEG), stereotaxy, morbidity, seizure outcome

„„ Introduction The main goal of epilepsy surgery is the complete resection (or complete disconnection) of the cortical areas responsible for the primary organization of epileptogenic activity. This area is also known as the EZ. As the EZ can overlap with functional cortical areas (eloquent cortex), preservation of these crucial regions is another goal of any surgical resection in patients with medically refractory epilepsy.1,​2,​3,​4,​5,​6,​7 Thus, epilepsy surgery is a balance between maximizing resection of the EZ while preserving eloquent neurological function. As successful resective epilepsy surgery relies on accurate preoperative localization of the EZ, a presurgical evaluation is necessary to obtain the widest and most accurate spectrum of information from clinical, anatomical, and neurophysiological

aspects, with the ultimate goal of performing an individualized resection for each patient. Initially, noninvasive methods (scalp electroencephalography [EEG], magnetic resonance imaging (MRI), positron emission tomography [PET], magnetoencephalography [MEG], ictal single-photon emission computed tomography [SPECT]) are used to attempt to lateralize and localize the EZ. These studies are then used to compose a hypothesis of the anatomical location of the EZ. When the noninvasive data are insufficient to define the EZ, extraoperative invasive monitoring may be indicated. SEEG is one of the extraoperative invasive methods that can be applied in patients with medically refractory focal epilepsy in order to anatomically define the EZ and the possible related functional cortical areas. The clinical aspects of SEEG methodology and technique, specifically applied to the pediatric population, will be discussed in this chapter.

„„ History and Basic Principles of the Stereoelectroencephalography Methodology The SEEG method was originally developed by Jean Talairach and Jean Bancaud during the 1950s8 and has been mostly used in France and western Europe as the method of choice for invasive mapping in refractory focal epilepsy.7,​9,​10,​11,​12,​13,​14,​15,​16,​17,​18,​19,​ 20,​21,​22,​23,​24,​25,​26,​27,​28,​29,​30,​31 In France, after the development of the stereotactic technique and frames, which were initially applied for movement disorder surgery, Jean Talairach devoted most of his activity to the field of epilepsy, along with Jean Bancaud, who joined him in 1952. The new methodology created by both physicians led them to depart very quickly from an approach limited to the superficial cortex, as favored by Wilder Penfield and his colleagues at Montreal Neurological Institute. His innovative thinking was to implement a working methodology for a comprehensive analysis of morphological and functional cerebral space. His atlas on the telencephalon, published in 1967, perfectly illustrates the new anatomical concepts for stereotaxis.32 The development of tools, adapted to a new stereotactic frame designed by Talairach and colleagues, spurred the French ­investigators to propose the functional exploration of the brain by depth electrodes, allowing the exploration of both superficial and deep cortical areas. The debut of SEEG was in 1957, when the first implantation of intracerebral electrodes for epi-

324

IVb  Intracranial Electrode Placement for Invasive Monitoring lepsy was performed at Saint Anne Hospital. By departing from the current methods of invasive monitoring, such implantations allow the exploration of activity of multiple, superficial, and deep brain structures while recording the patient’s spontaneous seizures. This new technique or method was described in 1962 as the “stereoelectroencephalography.”11,​32 The principle of SEEG methodology today remains similar to the principles originally described by Bancaud and Talairach, which is based on anatomo-electro-clinical (AEC) correlations. Thus, the main aim of this methodology is to conceptualize the three-dimensional spatial–temporal organization of the epileptic discharge within the brain.7,​11,​12,​13,​22–​31,​33,​34 As a result, the implantation strategy is individualized, with electrode placement based on preimplantation hypotheses that take into consideration the patient’s individual AEC and the anatomical relationship with a supposed epileptic lesion. For these reasons, the preimplantation AEC hypothesis formulation is the single most important element in the process of planning the placement of SEEG electrodes. If the preimplantation hypotheses are incorrect, the placement of the depth electrodes will be inadequate, and the interpretation of the SEEG recordings will not be able to fully describe the EZ.

„„ Choosing SEEG as the Appropriate Method for Extraoperative Invasive Monitoring Following the establishment of the diagnosis of pharmacoresistant epilepsy (defined as a failure to respond to two or more adequately chosen and used antiepileptic medications),35 a presurgical evaluation is indicated with two main goals: (1) mapping of the AEC network leading to the identification of the EZ and its extent, and (2) assessment of the functional status of the ­epileptogenic region(s). Achievement of both goals will lead to optimization of postresective seizure and functional outcomes following surgery. As briefly discussed above, multiple techniques may be used in order to achieve the above stated goals. Scalp video-EEG monitoring is needed to confirm the diagnosis of focal ­epilepsy (including interictal and ictal EEG recordings). The recorded clinical semiology on video monitoring along with the electrical abnormalities on EEG allows the team to identify the cortical structure of the hypothetical networks that may be involved in seizure organization and leads to the formulation of clear AEC hypotheses. Further, validation of the AEC h ­ ypothesis is achieved through structural imaging (the identification of any potential lesions on MRI), with or without metabolic imaging (including FDG-PET h ­ ypometabolism that may point to focal regions of cortical dysfunction). Other studies may include ictal SPECT, MEG, and EEG-functional MRI.6,​36,​37 These noninvasive studies identify the EZ in more than half of the patients undergoing presurgical workup (around 70% of the patients who are operated on at Cleveland Clinic in 2012, unpublished data). These patients are able to be offered a resective surgery without further testing. However, a formulation of a clear and unique AEC hypothesis may not be possible in the remaining 30% of patients. In these cases, focal or focal–regional epilepsy is likely, but there are multiple AEC hypotheses that all remain plausible after the noninvasive testing. Occasionally, there is a sound regional hypothesis but not enough arguments in favor of one hemisphere.

Finally, there are cases in which hypotheses are generated, but the exact location of the EZ, its extent, and/or its overlap with functional (eloquent) cortex remain unclear. Consequently, these children may be candidates for an invasive evaluation using different methods of evaluation, including intraoperative electrocorticography (ECoG) or extraoperative methods as subdural grids or strips, subdural grids combined with depth electrodes and SEEG.38 In summation, the main indications for an invasive evaluation in focal pharmacoresistant epilepsy are to address the challenges and limitations of various noninvasive techniques. Based on the limitations outlined above, an invasive evaluation should be considered in any one of the following cases: yy MRI negative cases: The MRI does not show a cortical lesion in a location that is concordant with the AEC hypothesis generated by the video-EEG recordings. yy Electroclinical and MRI data discordance: The anatomical location of the MRI-identified lesion (and at times the location of a clearly hypometabolic focal area on PET) is not concordant with the AEC hypothesis. These include cases of deeply seated brain lesions such periventricular nodular heterotopias or deep sulcal lesions. In addition, scalp EEG recordings in 85 to 100% of patients with focal cortical dysplasia (FCD) show interictal spikes that range in their distribution from lobar to lateralized, from difficult to localize to diffuse (including generalized spike-wave patterns in some cases of subependymal heterotopia).26,​27,​31,​39,​40,​41 The spatial distribution of interictal spikes is usually more extensive than the structural abnormality as assessed by intraoperative inspection or MRI visual analysis.42 yy Multiple, in part discordant lesions: There are two or more anatomical lesions with the location of at least one of them being discordant with the AEC hypothesis, or both lesions are located within the same functional network, and it is unclear if either or both are epileptogenic. yy Overlap with eloquent cortex: The generated AEC hypothesis (MRI-negative or MRI-identifiable lesion) involves potentially highly eloquent cortex. The identification of the EZ, mapping of its extent, and/or its relationship with potentially eloquent cortex are not typically resolved in these cases. These include patients with suspected FCD as the possible pathological substrate for epilepsy.4,​34,​40,​42,​43,​44,​45,​46 In these instances, an invasive evaluation usually leads to the formulation of a clear resective surgical strategy. In our center, approximately two-thirds of patients explored with the SEEG methodology are considered candidates for respective surgery. The recommendation for invasive monitoring and its type is made during a multidisciplinary patient management meeting that includes neurologists, neurosurgeons, neuroradiologists, and neuropsychologists. Areas and networks of coverage or sampling are determined based on a well-formulated AEC hypothesis including results of the noninvasive studies. There is no clear consensus on the best selection criteria for the type of invasive evaluation. Some epilepsy centers have applied both technical procedures in a systematic manner, but none of them have conducted definitive comparative studies. The “pro-SEEG groups” consider that this method can answer any question an invasive method can provide.7,​9–​31,​33,​34,​38,​39,​47–​68 On the contrary, the “pro-subdural groups” tend to limit SEEG indications strictly to the exploration of deep structures, for example, to distinguish unilateral or bilateral lobe epilepsy,

34  Stereoelectroencephalography in Children: Methodology and Surgical Technique and possibly to study epilepsy related to nodular heterotopia. However, differences between SEEG and subdural grids and strips are more extensive and complex than just the dichotomy between deep versus superficial mapping. As outlined in the historical section above, the philosophical underpinnings and conceptual basis of the two types of explorations are quite different and at times divergent. Subdural explorations were initially oriented toward the invasive study of lesional epilepsy, whereas SEEG evaluations are typically not as focused on anatomically localizable lesions and instead are focused on investigating functionally connected networks. We may therefore speculate that SEEG is more suitable to explore children with nonlesional MRIs for whom, in some cases, it is not at all clear that surgery should be performed.38,​60,​61,​62 In addition, SEEG allows us to explore remote and multilobar areas without the need for craniotomies and immediate surgery, allowing a prolonged reflection time for the patients and families and, consequently, a more complete informed consent process. Extraoperative mapping with the subdural method (including grids, strips, and the possible combination with depth electrodes) has the advantage of allowing an optimal anatomical and contiguous coverage and sampling of the adjacent cortex leading to accurate superficial cortex functional mapping ­exploration.69,​70 This is especially the case when there is the need to determine the extension of the EZ associated with a superficial lesion and its anatomical relation to a close functional area, in special, posterior language areas. This is not true if the lesion includes a deep-seated component where functional mapping cannot be obtained from subdural mapping. From a surgical perspective, subdural implantations are open procedures, with better management of occasional intracranial hemorrhagic complications. The main disadvantages of the subdural method are related to the inability to record and map deep structures such as the insular cortex, orbitofrontal cortex, cingulate gyrus, depths of sulci, etc., and consequently, its incapacity in analyzing the spatiotemporal dynamics of the epileptogenic network. In these scenarios, the SEEG m ­ ethodology may be considered a more adequate and safer option. SEEG has the advantages of allowing extensive and precise deep brain

recordings and stimulations (to localize seizure onset) with minimal associated morbidity.34,​54,​55,​60,​61,​62,​71 The particular feature related to low-morbidity rate is particularly appealing for the pediatric population. Consequently, based on the potential advantages and disadvantages from each method, one can consider possible specific indications to choose SEEG with respect to other methods of invasive monitoring: yy The possibility of a deep-seated or difficult-to-cover location of the EZ in areas such as the mesial structures of the temporal lobe, perisylvian areas, cingulate gyrus and mesial interhemispheric regions, ventromedial prefrontal areas, insula, and depths of sulci. yy A failure of a previous subdural invasive study to clearly outline the exact location of the seizure-onset zone. The failure to identify the EZ in these patients may be due to multiple reasons that include the lack of adequate sampling from a deep focus or a clinically silent focus upstream from the EZ. yy The need for extensive bihemispheric explorations (in particular in focal epilepsies arising from the interhemispheric or deep insular regions, or temporo-parieto-­occipital junction). yy Presurgical evaluation suggestive of extended network involvement (e.g., temporofrontal or frontoparietal) in the setting of a normal MRI (Table 34.1). A majority of patients undergoing reoperations may have failed epilepsy surgery during preceding subdural evaluations because of difficulties in accurately localizing the EZ. These patients pose a significant dilemma for further management, having relatively few options available. Further, open subdural grid evaluations may carry the risks associated with encountering dural and cortical scarring and still having limitations related to deep cortical structure recordings. A subsequent evaluation using the SEEG method may overcome these limitations, offering an additional opportunity for seizure localization and sustained seizure freedom.55The hypothetical disadvantage of the SEEG method is the more restricted capability for performing functional

Table 34.1  Selection criteria for different methods of invasive monitoring in medically refractory focal epilepsy

Clinical scenario

Method of choice

Second option

• Lesional MRI: Potential epileptogenic lesion is superficially located, near or in the proximity of eloquent cortex • Nonlesional MRI: Hypothetical EZ located in the proximity of eloquent cortex

SBG

SEEG

• Lesional MRI: Potential epileptogenic lesion is located in deep cortical and subcortical areas • Nonlesional MRI: Hypothetical EZ is deeply located or located in non-eloquent areas

SEEG

SBG with depths

• Need for bilateral explorations and/or reoperations

SEEG

SBG with depths

• After subdural grids failure

SEEG

SBG with depths

• When the AEC hypothesis suggest the involvement of a more extensive, multilobar epileptic network

SEEG

SBG with depths

• Suspected frontal lobe epilepsy in nonlesional MRI scenario

SEEG

SEEG

Abbreviations: AEC, anatomo-electro-clinical; EZ, epileptogenic zone; SBG, standards-based grading; SEEG, ­stereoelectroencephalography.

325

326

IVb  Intracranial Electrode Placement for Invasive Monitoring mapping. Because of a limited number of c­ ontacts located in the superficial cortex, contiguous mapping of eloquent brain areas cannot be obtained as in subdural mapping.34,​54,​55 It is interesting to note that functional mapping in SEEG cannot be dissociated from the electroclinical localization process and, consequently, a fair comparison between both methods cannot be performed. Lastly, the functional mapping information extracted from the SEEG method can be frequently complemented with other methods of mapping, such as diffusion tensor imaging (DTI) or intraoperative mapping in awake craniotomies,34 diminishing the relative disadvantages compared to the subdural grid method.

„„ How to Select SEEG Trajectories: Planning the Implantation? As indicated above, the development of a SEEG implantation plan requires the clear formulation of precise AEC hypotheses to be tested. These hypotheses are typically generated during the multidisciplinary patient management conference based on the results of various noninvasive tests. At Cleveland Clinic, a final tailored implantation strategy is generated during a separate presurgical implantation meeting. Depth electrodes should sample the anatomical lesion (if identified), the more likely structure(s) of ictal onset, the early and late spread regions, and the interactions with the functional (cognitive, sensorimotor, behavioral, etc.) networks. A three-dimensional “conceptualization” of the network nodes upstream and downstream from the hypothesized epileptogenic network is an essential component of the presurgical implantation strategy. Initially, by analyzing the available noninvasive data and the temporal evolution of the ictal clinical manifestations, a hypothesis of the anatomical location of the EZ is formulated.72 The implantation plan is created in collaboration with experienced epileptologists, neurosurgeons, and neuroradiologists who, together, will formulate hypotheses for EZ localization. Adequate knowledge of the possible functional networks involved in the primary organization of the epileptic activity is mandatory in order to formulate adequate hypotheses. In addition, the treating physicians will have to take into account the three-dimensional aspects of depth electrode recordings, which enables an accurate sampling of the structures along its trajectory, from the entry site to the final impact point. Thus, the trajectory is more important than the target or entry point areas. Consequently, the investigation may include lateral and mesial surfaces of the different lobes, deep-seated cortices as the depths of sulci, insula, posterior areas in the interhemispheric cortical surface, etc. The implantation should also consider the different cortical cytoarchitectonic areas involved in seizure organization patterns and their likely connectivity to other cortical and subcortical areas. It is important to emphasize that the implantation strategy focus is not to map lobes or lobules, but epileptogenic networks, which, in general, involve multiple lobes. Furthermore, the exploration strategy should also take into consideration possible alternative hypotheses of localization.57,​62,​73 Lastly, the aim to obtain all the possible information from the SEEG exploration should not be pursued at the expenses of an excessive number of electrodes, which will likely increase the morbidity of the implantation. In general, implantations that exceed 15 depth electrodes are rare. In addition, the possible involvement of eloquent regions in the ictal discharge requires

their judicious coverage, with the twofold goal to assess their role in the seizure organization and to define the boundaries of a safe surgical resection (Fig. 34.1). The SEEG implantation patterns are based on a tailored strategy of exploration, which results from the primary hypothesis of the anatomical location of the EZ, for every single case. In consequence, standard implantations for specific areas and lobes are difficult to conceptualize. Nevertheless, a number of typical patterns of coverage can be recognized, as discussed below.

Limbic Network Explorations Cases of temporal lobe epilepsy with consistent AEC findings suggesting a limbic network involvement are usually operated on after noninvasive investigation only. In general, the use of invasive monitoring is not necessary when semiological and electrophysiological studies demonstrate typical nondominant mesial temporal epilepsy, and imaging studies show clear lesion (mesial temporal sclerosis, as an example) that fits the initial localization hypothesis. Nevertheless, invasive exploration with SEEG recordings may be required in patients in whom the supposed EZs, probably involving the temporal lobes, are suspected to involve extratemporal areas as well. In these cases, the implantation pattern points to disclose a preferential spread of the discharge to the temporoinsular anterior perisylvian areas, the temporoinsular orbitofrontal areas, or the posterior temporal, posterior insula, temporobasal, parietal, and posterior cingulate areas. Consequently, sampling of extratemporal limbic areas must be wide enough to provide information to identify a possible extratemporal origin of the seizures that could not been anticipated with precision according to noninvasive methods of investigation.

Frontal–Parietal Network Explorations Because of the large volume of the frontal and parietal lobes, a high number of electrodes are required for an adequate coverage of this region. In most patients, however, excessive sampling can be avoided, and the implantation to more limited portions of the frontal and parietal lobes can be performed. The suspicion of orbitofrontal epilepsy, for instance, often requires the investigation of gyrus rectus, the frontal polar areas, the anterior cingulate gyrus, and the anterior portions of the temporal lobe (temporal pole). Similarly, seizures that are thought to arise from the mesial wall of the premotor cortex are evaluated by targeting at least the rostral and caudal part of the supplementary motor area (SMA), the pre-SMA area, different portions of the cingulate gyrus and sulcus, as well as the primary motor cortex and mesial and dorsal–lateral parietal cortex. Consequently, the hypothesis-based sampling often allows localization of the EZ in the frontal and/or parietal lobes, and in some cases may allow the identification of relatively small EZs. Eventually, frontal– parietal network explorations may be bilateral, and sometimes symmetrical, mainly when a mesial f­ rontal–parietal epilepsy is suspected and the noninvasive methods of investigation failed in lateralizing the epileptic activity. Electrodes in Rolandic regions are normally placed when there is a need to define the posterior margin of the resection in frontal network explorations or the anterior margin in ­parietal– occipital explorations, or when the EZ may be located in or near Rolandic cortex. The main goal here is to evaluate the Rolandic participation to the ictal discharge and to obtain a functional mapping by intracerebral electrical stimulation. In this location,

34  Stereoelectroencephalography in Children: Methodology and Surgical Technique

Fig. 34.1  Illustrative case: A 15-year-old teenager with intractable epilepsy and nonlesional MRI who underwent SEEG implantation and SEEGguided resection. (a) Preimplantation ictal SPECT showing hyperflow in the right temporal and orbitofrontal regions. (b) Postoperative sagittal MRI showing a right orbitofrontal/temporal resection guided by SEEG data analyses. (c) Ictal SEEG recording, showing ictal epileptiform activity in the temporal pole (TP) and orbitofrontal (OF) electrode contacts. (d, e) Surgical pathology from the orbitofrontal and temporal pole respective areas showing mild forms of cortical dysplasia. Patient has been seizure free for 5 years.

depth electrodes are particularly helpful to sample the depth of the central sulcus, as well as the descending and ascending white matter fibers associated with this region.

Posterior Quadrant Network Explorations In the posterior quadrant, placement of electrodes limited to a single lobe is extremely uncommon, due to the frequent simultaneous involvement of several occipital, parietal, and posterior temporal structures, as well as to the multidirectional spread of the discharges to supra- and infrasylvian areas. Consequently, mesial and dorsal lateral surfaces of the occipital lobes are explored, covering both infracalcarine and supracalcarine areas, in association with posterior temporal, posterior p ­ erisylvian, basal temporal–occipital areas, and posterior parietal areas including the posterior inferior parietal lobule and the posterior precuneus. In posterior quadrant epilepsies, ­bilateral explorations are generally needed due to rapid contralateral spread of ictal activity.

„„ The “Nuts and Bolts” of the SEEG Implantation Technique in Children Once the SEEG planning is finalized, the desired targets are reached using commercially available depth electrodes in

­arious lengths and number of contacts, depending on the v specific brain regions to be explored. The depth electrodes are implanted using conventional stereotactic technique or by the assistance of stereotactic robotic devices through 2.5-mm diameter drill holes. In both techniques, depth electrodes are inserted through 2.5-mm diameter drill holes, using orthogonal or oblique orientation, allowing intracranial recording from lateral, intermediate, or deep cortical and subcortical structures in a three-dimensional arrangement, thus accounting for the dynamic, multidirectional spatiotemporal organization of the epileptogenic pathways. More recently, robotic-assisted devices were applied, improving the workflow and time of implantation. Volumetric preoperative MRIs are obtained and DICOM format images are digitally transferred to the robot’s native planning software. Individual trajectories are planned within the three-dimensional imaging reconstruction according to predetermined target locations and intended trajectories. Trajectories are selected to maximize sampling from superficial and deep cortical and subcortical areas within the preselected zones of interest and are oriented orthogonally in the majority of cases to facilitate the anatomo-electrophysiological correlation during the extraoperative recording phase and to avoid possible trajectory shifts due to excessively angled entry points. Nevertheless, when multiple targets are potentially accessible via a single nonorthogonal trajectory, these multitarget trajectories are selected in order to minimize the number of implanted electrodes per patient.

327

328

IVb  Intracranial Electrode Placement for Invasive Monitoring All trajectories are evaluated for safety and target accuracy in their individual reconstructed planes (axial, sagittal, and coronal), and also along the reconstructed “probe’s eye view.” Any trajectory that appears to compromise vascular structures are adjusted appropriately without affecting the sampling from areas of interest. A set working distance of 150 mm from the drilling platform to the target are initially utilized for each trajectory, and then later adjusted in order to maximally reduce the working distance and improve the implantation accuracy. The overall implantation schemas are analyzed using the three-dimensional cranial reconstruction capabilities, and internal trajectories are checked to ensure that no trajectory collisions are present. External trajectory positions are examined for any entry sites that would be prohibitively close (< 1.5 cm distance) at the skin level. On the day of surgery, patients are placed under general anesthesia. For each patient, the head is placed into a threepoint fixation head holder. The robot is then positioned such that the working distance (distance between the base of the robotic arm and the midpoint of the cranium) is approximately 70 cm. The robot is locked into position, and the head holder device is secured to the robot. No additional position adjustments are made to the operating table during the implantation procedure. After positioning and securing the patient to the robot, image registrations are performed. Semiautomatic laser-based facial recognition is utilized to register the preoperative volumetric MRI with the patient. The laser is first calibrated using a set distance calibration tool. Preset anatomical facial landmarks are then manually selected with the laser. The areas defined by the manually entered anatomical landmarks subsequently undergo automatic registration using laser-based facial surface scanning. Accuracy of the registration process is then confirmed by correlating additional independently ­chosen

surface landmarks with the registered MRI. After successful ­registration, the planned trajectories’ accessibility with respect to the robotic arm position is automatically verified by the robot software. The patients are then prepped and draped in a standard sterile fashion. The robotic working arm is also draped with a sterile plastic cover. A drilling platform, with a 2.5-mm diameter working cannula is secured to the robotic arm. The desired trajectories are selected on the touch screen interface. After trajectory confirmation, the arm movement is initiated through the use of a foot pedal. The robotic arm automatically locks the drilling platform into a stable position once reaching the calculated position for the selected trajectory. A 2-mm diameter handheld drill (Stryker) is introduced through the platform and used to create a pinhole through the entire skull thickness. Dura is then opened with an insulated dural perforator using monopolar cautery at low settings. This step can be particularly challenging in small children, since the dura is not completely attached to the internal layers of the skull and displacement, instead of opening, can occur without been noticed. This particular step requires close attention. A guide bolt (Ad-Tech, Racine, WI) is screwed firmly into each pinhole. The distance from d ­ rilling platform to the retaining bolt is measured, and this value is subtracted from the standardized 150-mm platform to target distance. The resulting difference is recorded for later use as the final length of the electrode to be implanted. This process is repeated for each trajectory. All pinholes and retaining bolts are placed prior to beginning electrode insertion. A small stylet (2 mm in diameter) was then set to the previously recorded electrode distance and passed gently into the parenchyma, guided by the implantation bolt, followed immediately by the insertion of the premeasured electrode (Fig. 34.2).

Fig. 34.2  Robotic SEEG technique. (a) Operating room setup during left side SEEG robotic implantation, with surgeon and scrub nurse positioned on each side of the patient and the robot device placed in the middle, at the vertex. (b) Intraoperative aspect of left side frontal–temporal SEEG implantation with the guiding bolts’ final position. (c) Left side frontal– temporal SEEG implantation after the depth electrodes implantations. Final result.

34  Stereoelectroencephalography in Children: Methodology and Surgical Technique

„„ Morbidity and Seizure Outcome in the Pediatric Population Like most surgical procedures, there are many fewer studies characterizing seizure freedom and complication rates in the pediatric population undergoing SEEG when compared to adults. Our center reported a series of 30 pediatric patients undergoing SEEG with either nonlesional MRIs or equivocal MRI findings with discordant noninvasive testing. In this series, the mean age of patients was 15 years with a total of 402 depth electrodes placed (13/patient). The epileptogenic zone was localized in 26 (87%) patients, but only 18 (69%) underwent resection because the EZ was either multifocal, located solely in eloquent cortex, or had a significant improvement in seizures after the implant. Of these 18 patients, 10 (56%) were seizure free at last follow-up (2.2 years on average), 5 had seizure improvement (28%), and 3 had no improvement (17%). Complications were minimal, with one asymptomatic intraparenchymal hematoma for an implantation complication rate of 3.3% (0.2% per electrode). There are few other studies detailing the pediatric SEEG experience, but other reported data appear to be similar to our experience. A larger case series in a slightly younger ­population was published contemporaneously from an Italian center describing tailored, extratemporal resections. In this series, 53 patients with a mean age of 11 years were studied; 19 patients had concordant noninvasive testing and underwent resections with no further testing; the other 34 patients ­underwent SEEG and 32 (94%) ended up having subsequent resections. Of these patients, 22 (69%) were seizure free at last follow-up (mean: 3.7 years), 5 (16%) had significant improvement (Engel class 1), and 5 (16%) were unimproved. This study did not specifically report complications related to SEEG. Finally, a French study in a much younger cohort of 19 pediatric patients reported its experience with SEEG. These patients were between 3 and 4 years old with pharmacoresistant focal epilepsy and FCD identified on postoperative pathology. Every patient’s EZ was localized and underwent a resective surgery. Sixteen (84%) patients were seizure free at last follow-up (mean: 2.4 years), but complications specific to SEEG were again not reported. Given the similarity in seizure freedom outcomes and EZ localization rates between the pediatric and adult series, however, it may be instructive to examine complication rates in the adult literature for c­ omparison’s sake. Our center recently reported 200 patients (adults and children) undergoing 2,663 SEEG electrode implantations for the purposes of invasive intracranial EEG monitoring, in accordance with a tailored preimplantation hypothesis to investigate and anatomically characterize the extension of the EZ. Complications were minimal. They included wound infections (0.08% per electrode), hemorrhagic complications (0.08% per electrode), and a transient neurological deficit (0.04% per electrode) in a total of five patients. The total morbidity rate was 2.5%. These results parallel those of previous studies in the recent literature. Munari et al74 reported on their experience

with SEEG in 70 patients undergoing a collective total of 712 electrode implantations. In their series, specifically relating to SEEG, the authors identified one permanent complication ensuing from the procedure; this entailed the formation of an asymptomatic intracerebral hematoma following the removal of an SEEG electrode (accounting for a morbidity rate of 1.4%, or 0.1% per electrode). More recently, Guenot et al75 presented a series of 100 patients collectively undergoing 1,118 SEEG electrode implantations for invasive EEG monitoring. These authors reported on five complications (5% of cases), including two electrode site infections (0.2% per electrode), two intracranial electrode fractures (0.2% per electrode), and one intracerebral hematoma resulting in death (accounting for a mortality rate of 1% in the study). In a large series, Cossu et al reported a morbidity rate of 5.6%, with severe permanent deficits from intracerebral hemorrhage in 1%.76 In another study, Tanriverdi et al77 summarized their experience with a subgroup of 491 refractory epilepsy patients collectively undergoing 2,490 intracerebral SEEG electrode ­ implantations and 2,943 depth electrode implantations.77 Based on the authors’ experience, they identified four patients (0.8%) with an intracranial hematoma at the electrode site (0.07% per electrode) and nine patients (1.8%) with an infection arising from electrode placement (0.2% per electrode); moreover, they reported no mortalities ensuing directly from SEEG electrode placement. Finally, Cardinale and Lo Russo73 most recently presented their experience with 6,496 electrodes stereotactically implanted in 482 epilepsy patients with refractory epilepsy. These authors identified 2 patients (0.4%, or 0.03% per electrode) with permanent neurological deficits in their series; 14 patients (2.9%, or 0.2% per electrode) with hemorrhagic complication; 2 patients (0.4%, or 0.03% per electrode) with infection; and one ­mortality (0.2%) resulting from massive brain edema and concomitant hyponatremia following electrode implantation. In comparing morbidity, subdural grid electrode implantation has historically been shown to have low permanent morbidity (0–3%) compared with depth electrodes (3–6%), since there is no intraparenchymal passage.2,​37,​78,​79,​80,​81,​82,​83 Although it is difficult to compare morbidity rates between subdural grids and SEEG due to the variability in patient selection, different institutions and variable number of implanted electrodes, the clinical experience among different groups in Europe and North America suggests that the SEEG method provides at least a similar degree of safety when compared with subdural grids or strips.7,​28,​29,​33,​38,​57,​61,​64,​74,​77,​83,​84,​85,​86

„„ Conclusion In children, the main advantage of the SEEG method is that it offers the ability to study the epileptogenic neuronal network in a dynamic and three-dimensional fashion, allowing the physician to correlate abnormal electrographic activity with the clinical seizure semiology in both time and space. The main clinical challenge for the near future remains in the further refinement of specific selection criteria for the different meth-

329

330

IVb  Intracranial Electrode Placement for Invasive Monitoring ods of invasive monitoring, with the ultimate goal of comparing and validating the results (long-term seizure-free outcome) obtained from different methods of invasive monitoring. To sum up, the key points are as follows: yy In the pediatric population, SEEG is a safe and efficient extraoperative invasive methodology designed to define the anatomical boundaries of the cortical and subcortical

cerebral regions responsible for the primary generation and early propagation of epileptiform activity. yy Both frame-based and frameless techniques can be utilized for implantation of SEEG electrodes. yy Vascular imaging is fundamental for the safe implantation of SEEG electrodes. Attention to the vascular anatomy is essential to reduce the risk of hemorrhagic complications.

References 1. Rosenow F, Lüders H. Presurgical evaluation of epilepsy. Brain 2001;124(Pt 9):1683–1700 2. Wyllie E, Lüders H, Morris HH III, et al. Subdural electrodes in the evaluation for epilepsy surgery in children and adults. Neuropediatrics 1988;19(2):80–86 3. Jayakar P, Duchowny M, Resnick TJ. Subdural monitoring in the evaluation of children for epilepsy surgery. J Child Neurol 1994;9(Suppl 2):61–66 4. Adelson PD, O’Rourke DK, Albright AL. Chronic invasive monitoring for identifying seizure foci in children. Neurosurg Clin N Am 1995;6(3):491–504 5. Jayakar P. Invasive EEG monitoring in children: when, where, and what? J Clin Neurophysiol 1999;16(5):408–418 6. Winkler PA, Herzog C, Henkel A, et al. [Noninvasive protocol for surgical treatment of focal epilepsies] Nervenarzt 1999;70(12):1088–1093 7. Cossu M, Chabardès S, Hoffmann D, Lo Russo G. [Presurgical evaluation of intractable epilepsy using stereo-electro-encephalography methodology: principles, technique and morbidity] Neurochirurgie 2008;54(3):367–373 8. Bancaud J, Dell MB. [Technics and method of stereotaxic functional exploration of the brain structures in man (cortex, subcortex, central gray nuclei)] Rev Neurol (Paris) 1959;101:213–227 9. Bancaud J, Talairach J, Waltregny P, Bresson M, Morel P. [Stimulation of focal cortical epilepsies by megimide in topographic diagnosis. (Clinical EEG and SEEG study)] Rev Neurol (Paris) 1968;119(3):320–325 10. Bancaud J, Talairach J, Waltregny P, Bresson M, Morel P. Activation by Megimide in the topographic diagnosis of focal cortical epilepsies (clinical EEG and SEEG study). Electroencephalogr Clin Neurophysiol 1969;26(6):640 11. Bancaud J, Angelergues R, Bernouilli C, et al. Functional stereotaxic exploration (SEEG) of epilepsy. Electroencephalogr Clin Neurophysiol 1970;28(1):85–86 12. Bancaud J, Favel P, Bonis A, Bordas-Ferrer M, Miravet J, Talairach J. [Paroxysmal sexual manifestations and temporal lobe epilepsy. Clinical, EEG and SEEG study of a case of epilepsy of tumoral origin] Rev Neurol (Paris) 1970;123(4):217–230 13. Bancaud J, Talairach J. [Methodology of stereo EEG exploration and surgical intervention in epilepsy] Rev Otoneuroophtalmol 1973;45(4):315–328 14. Geier S, Bancaud J, Talairach J, Enjelvin M. [Radio-telemetry in EEG and SEEG. Technology and material] Rev Electroencephalogr Neurophysiol Clin 1973;3(4):353–354 15. Cabrini GP, Ettorre G, Marossero F, Miserocchi G, Ravagnati L. Surgery of epilepsy: some indications for SEEG. J Neurosurg Sci 1975;19(1–2):95–104 16. Bancaud J, Talairach J, Geier S, Bonis A, Trottier S, Manrique M. [Behavioral manifestations induced by electric stimulation of the anterior cingulate gyrus in man] Rev Neurol (Paris) 1976;132(10):705–724 17. Musolino A, Tournoux P, Missir O, Talairach J. Methodology of “in vivo” anatomical study and stereo-electroencephalographic exploration in brain surgery for epilepsy. J Neuroradiol 1990;17(2):67–102

18. Engel J Jr, Henry TR, Risinger MW, et al. Presurgical evaluation for partial epilepsy: relative contributions of chronic depth-electrode recordings versus FDG-PET and scalp-sphenoidal ictal EEG. Neurology 1990;40(11):1670–1677 19. Baucaud J, Talairach J, Munari C, Giallonardo T, Brunet P. [Introduction to the clinical study of postrolandic epileptic seizures]. Can J Neurol Sci. Le Journal Canadien des Sciences Neurologiques 1991;18(4, Suppl):566–569 20. Talairach J, Bancaud J, Bonis A, et al. Surgical therapy for frontal epilepsies. Adv Neurol 1992;57:707–732 21. Avanzini G. Discussion of stereoelectroencephalography. Acta Neurol Scand Suppl 1994;152:70–73 22. Bartolomei F, Wendling F, Bellanger JJ, Régis J, Chauvel P. Neural networks involving the medial temporal structures in temporal lobe epilepsy. Clin Neurophysiol 2001;112(9):1746–1760 23. Biraben A, Taussig D, Thomas P, et al. Fear as the main feature of epileptic seizures. J Neurol Neurosurg Psychiatry 2001;70(2):186–191 24. Wendling F, Bartolomei F, Bellanger JJ, Chauvel P. Interpretation of interdependencies in epileptic signals using a macroscopic physiological model of the EEG. Clin Neurophysiol 2001;112(7):1201–1218 25. Wendling F, Bartolomei F, Bellanger JJ, Chauvel P. [Identification of epileptogenic networks from modeling and nonlinear analysis of SEEG signals] Neurophysiol Clin 2001;31(3):139–151 26. Tassi L, Colombo N, Cossu M, et al. Electroclinical, MRI and neuropathological study of 10 patients with nodular heterotopia, with surgical outcomes. Brain 2005;128(Pt 2):321–337 27. Battaglia G, Chiapparini L, Franceschetti S, et al. Periventricular nodular heterotopia: classification, epileptic history, and genesis of epileptic discharges. Epilepsia 2006;47(1):86–97 28. Cossu M, Cardinale F, Castana L, Nobili L, Sartori I, Lo Russo G. ­Stereo-EEG in children. Childs Nerv Syst 2006;22(8):766–778 29. Sindou M, Guenot M, Isnard J, Ryvlin P, Fischer C, Mauguière F. Temporo-mesial epilepsy surgery: outcome and complications in 100 consecutive adult patients. Acta Neurochir (Wien) 2006;148(1):39–45 30. Guenot M, Isnard J. [Epilepsy and insula] Neurochirurgie 2008;54(3):374–381 31. Guenot M, Isnard J. [Multiple SEEG-guided RF-thermolesions of epileptogenic foci] Neurochirurgie 2008;54(3):441–447 32. Talairach J, Bancaud J, Bonis A, Tournoux P, Szikla G, Morel P. [Functional stereotaxic investigations in epilepsy. Methodological remarks concerning a case] Rev Neurol (Paris) 1961;105:119–130 33. Devaux B, Chassoux F, Guenot M, et al. [Epilepsy surgery in France] Neurochirurgie 2008;54(3):453–465 34. Gonzalez-Martinez J, Bulacio J, Alexopoulos A, Jehi L, Bingaman W, Najm I. Stereoelectroencephalography in the “difficult to localize” refractory focal epilepsy: early experience from a North American epilepsy center. Epilepsia 2013;54(2):323–330 35. Kwan P, Brodie MJ. Definition of refractory epilepsy: defining the indefinable? Lancet Neurol 2010;9(1):27–29 36. Najm IM, Naugle R, Busch RM, Bingaman W, Lüders H. Definition of the epileptogenic zone in a patient with non-lesional tempo-

34  Stereoelectroencephalography in Children: Methodology and Surgical Technique ral lobe epilepsy arising from the dominant hemisphere. Epileptic Disord 2006;8(Suppl 2):S27–S35

grid placement for difficult to localize epilepsy. Neurosurgery 2013;72(5):723–729, discussion 729

37. Nair DR, Burgess R, McIntyre CC, Lüders H. Chronic subdural electrodes in the management of epilepsy. Clin Neurophysiol 2008;119(1):11–28

56. Wang S, Wang IZ, Bulacio JC, et al. Ripple classification helps to localize the seizure-onset zone in neocortical epilepsy. Epilepsia 2013;54(2):370–376

38. Gonzalez-Martinez J, Najm IM. Indications and selection criteria for invasive monitoring in children with cortical dysplasia. Childs Nerv Syst 2014;30(11):1823–1829

57. Cardinale F, Cossu M, Castana L, et al. Stereoelectroencephalography: surgical methodology, safety, and stereotactic application accuracy in 500 procedures. Neurosurgery 2013;72(3):353–366, discussion 366

39. Marnet D, Devaux B, Chassoux F, et al. [Surgical resection of focal cortical dysplasias in the central region] Neurochirurgie 2008;54(3):399–408 40. Russo GL, Tassi L, Cossu M, et al. Focal cortical resection in malformations of cortical development. Epileptic Disord 2003;5(Suppl 2):S115–S123 41. Lüders H, Schuele SU. Epilepsy surgery in patients with malformations of cortical development. Curr Opin Neurol 2006;19(2):169–174 42. Kellinghaus C, Möddel G, Shigeto H, et al. Dissociation between in vitro and in vivo epileptogenicity in a rat model of cortical dysplasia. Epileptic Disord 2007;9(1):11–19 43. González-Martínez JA, Srikijvilaikul T, Nair D, Bingaman WE. Long-term seizure outcome in reoperation after failure of epilepsy surgery. Neurosurgery 2007;60(5):873–880, discussion 873–880 44. Tassi L, Colombo N, Garbelli R, et al. Focal cortical dysplasia: neuropathological subtypes, EEG, neuroimaging and surgical outcome. Brain 2002;125(Pt 8):1719–1732 45. Srikijvilaikul T, Najm IM, Hovinga CA, Prayson RA, ­Gonzalez-Martinez J, Bingaman WE. Seizure outcome after temporal lobectomy in temporal lobe cortical dysplasia. Epilepsia 2003;44(11):1420–1424 46. Francione S, Kahane P, Tassi L, et al. Stereo-EEG of interictal and ictal electrical activity of a histologically proved heterotopic gray matter associated with partial epilepsy. Electroencephalogr Clin Neurophysiol 1994;90(4):284–290 47. Catenoix H, Mauguière F, Guénot M, et al. SEEG-guided thermocoagulations: a palliative treatment of nonoperable partial epilepsies. Neurology 2008;71(21):1719–1726 48. Abraham G, Zizzadoro C, Kacza J, et al. Growth and differentiation of primary and passaged equine bronchial epithelial cells under conventional and air-liquid-interface culture conditions. BMC Vet Res 2011;7:26 49. Kerr MS, Burns SP, Gale J, Gonzalez-Martinez J, Bulacio J, Sarma SV. Multivariate analysis of SEEG signals during seizure. Conf Proc IEEE Eng Med Biol Soc 2011;2011:8279–8282 50. Centeno RS, Yacubian EM, Caboclo LO, Júnior HC, Cavalheiro S. Intracranial depth electrodes implantation in the era of imageguided surgery. Arq Neuropsiquiatr 2011;69(4):693–698 51. Kakisaka Y, Kubota Y, Wang ZI, et al. Use of simultaneous depth and MEG recording may provide complementary information regarding the epileptogenic region. Epileptic Disord 2012;14(3):298–303 52. Yaffe R, Burns S, Gale J, et al. Brain state evolution during seizure and under anesthesia: a network-based analysis of stereotaxic EEG activity in drug-resistant epilepsy patients. Conf Proc IEEE Eng Med Biol Soc 2012;2012:5158–5161 53. Antony AR, Alexopoulos AV, González-Martínez JA, et al. Functional connectivity estimated from intracranial EEG predicts surgical outcome in intractable temporal lobe epilepsy. PLoS One 2013;8(10):e77916 54. Vadera S, Marathe AR, Gonzalez-Martinez J, Taylor DM. Stereoelectroencephalography for continuous two-dimensional cursor control in a brain-machine interface. Neurosurg Focus 2013;34(6):E3 55. Vadera S, Mullin J, Bulacio J, Najm I, Bingaman W, Gonzalez-­ Martinez J. Stereoelectroencephalography following subdural

58. Enatsu R, Bulacio J, Nair DR, Bingaman W, Najm I, Gonzalez-Martinez J. Posterior cingulate epilepsy: clinical and neurophysiological analysis. J Neurol Neurosurg Psychiatry 2014;85(1):44–50 59. Enatsu R, Bulacio J, Najm I, et al. Combining stereo-electroencephalography and subdural electrodes in the diagnosis and treatment of medically intractable epilepsy. J Clin Neurosci 2014;21(8):1441–1445 60. Gonzalez-Martinez J, Lachhwani D. Stereoelectroencephalography in children with cortical dysplasia: technique and results. Childs Nerv Syst 2014;30(11):1853–1857 61. Gonzalez-Martinez J, Mullin J, Bulacio J, et al. Stereoelectroencephalography in children and adolescents with difficult-to-­ localize refractory focal epilepsy. Neurosurgery 2014;75(3):258–268, ­discussion 267–268 62. Gonzalez-Martinez J, Mullin J, Vadera S, et al. Stereotactic placement of depth electrodes in medically intractable epilepsy. J Neurosurg 2014;120(3):639–644 63. Johnson MA, Thompson S, Gonzalez-Martinez J, et al. Performing behavioral tasks in subjects with intracranial electrodes. J Vis Exp 2014(92):e51947 64. Serletis D, Bulacio J, Bingaman W, Najm I, González-Martínez J. The stereotactic approach for mapping epileptic networks: a prospective study of 200 patients. J Neurosurg 2014;121(5):1239–1246 65. Vadera S, Burgess R, Gonzalez-Martinez J. Concomitant use of stereoelectroencephalography (SEEG) and magnetoencephalographic (MEG) in the surgical treatment of refractory focal epilepsy. Clin Neurol Neurosurg 2014;122:9–11 66. Cardinale F, Cossu M. Letter to the Editor: SEEG has the lowest rate of complications. Journal of Neurosurgery 2014:1–3 67. Cossu M, Fuschillo D, Cardinale F, et al. Stereo-EEG-guided radio-frequency thermocoagulations of epileptogenic grey-­ matter nodular heterotopy. J Neurol Neurosurg Psychiatry 2014;85(6):611–617 68. Enatsu R, Gonzalez-Martinez J, Bulacio J, et al. Connections of the limbic network: a corticocortical evoked potentials study. Cortex 2015;62:20–33 69. Najm IM, Bingaman WE, Lüders HO. The use of subdural grids in the management of focal malformations due to abnormal cortical development. Neurosurg Clin N Am 2002;13(1):87– 92, viii–ix 70. Widdess-Walsh P, Jeha L, Nair D, Kotagal P, Bingaman W, Najm I. Subdural electrode analysis in focal cortical dysplasia: predictors of surgical outcome. Neurology 2007;69(7):660–667 71. Kovac S, Kahane P, Diehl B. Seizures induced by direct electrical cortical stimulation—mechanisms and clinical considerations. Clin Neurophysiol 2016;127(1):31–39 72. Chauvel P, McGonigal A. Emergence of semiology in epileptic seizures. Epilepsy Behav 2014;38:94–103 73. Cardinale F, Lo Russo G. Stereo-electroencephalography safety and effectiveness: some more reasons in favor of epilepsy surgery. Epilepsia 2013;54(8):1505–1506 74. Munari C, Hoffmann D, Francione S, et al. Stereo-electroencephalography methodology: advantages and limits. Acta Neurol Scand Suppl 1994;152:56–67, discussion 68–69 75. Guenot M, Isnard J, Ryvlin P, et al. Neurophysiological monitoring for epilepsy surgery: the Talairach SEEG method. StereoElec-

331

332

IVb  Intracranial Electrode Placement for Invasive Monitoring troEncephaloGraphy. Indications, results, complications and therapeutic applications in a series of 100 consecutive cases. Stereotact Funct Neurosurg 2001;77(1–4):29–32

81. Onal C, Otsubo H, Araki T, et al. Complications of invasive subdural grid monitoring in children with epilepsy. J Neurosurg 2003;98(5):1017–1026

76. Cossu M, Cardinale F, Colombo N, et al. Stereoelectroencephalography in the presurgical evaluation of children with drug-resistant focal epilepsy. J Neurosurg 2005;103 (4, Suppl):333–343

82. González Martínez F, Navarro Gutiérrez S, de León Belmar JJ, Valero Serrano B. [Electrocardiographic disorders associated to recent onset epilepsy] Neurologia 2005;20(10):698–701

77. Tanriverdi T, Ajlan A, Poulin N, Olivier A. Morbidity in epilepsy ­surgery: an experience based on 2449 epilepsy surgery procedures from a single institution. J Neurosurg 2009;110(6): 1111–1123 78. Lee WS, Lee JK, Lee SA, Kang JK, Ko TS. Complications and results of subdural grid electrode implantation in epilepsy surgery. Surg Neurol 2000;54(5):346–351 79. Rydenhag B, Silander HC. Complications of epilepsy surgery after 654 procedures in Sweden, September 1990–1995: a multicenter study based on the Swedish National Epilepsy Surgery Register. Neurosurgery 2001;49(1):51–56, discussion 56–57 80. Hamer HM, Morris HH, Mascha EJ, et al. Complications of ­invasive video-EEG monitoring with subdural grid electrodes. Neurology 2002;58(1):97–103

83. Ozlen F, Asan Z, Tanriverdi T, et al. Surgical morbidity of invasive monitoring in epilepsy surgery: an experience from a single institution. Turk Neurosurg 2010;20(3):364–372 84. Afif A, Chabardes S, Minotti L, Kahane P, Hoffmann D. Safety and usefulness of insular depth electrodes implanted via an oblique approach in patients with epilepsy. Neurosurgery 2008;62(5, Suppl 2):ONS471–ONS479, discussion 479–480 85. Nobili L, Cardinale F, Magliola U, et al. Taylor’s focal cortical dysplasia increases the risk of sleep-related epilepsy. Epilepsia 2009;50(12):2599–2604 86. Serletis D, Bulacio J, Alexopoulos A, Najm I, Bingaman W, González-Martínez J. Tailored unilobar and multilobar resections for ­orbitofrontal-plus epilepsy. Neurosurgery 2014;75(4): 388–397, discussion 397

35

  Mesial Temporal Sclerosis in Pediatric Epilepsy Rafael Uribe, George I. Jallo, and Caitlin Hoffman

Summary The temporal lobe is the most epileptogenic region and the most common target for resective surgery, with mesial temporal sclerosis (MTS) representing the most common etiology in adults. In children, however, most localization-related epilepsies have an extratemporal origin, with pure mesial temporal lobe epilepsy (MTLE) and MTS being less frequent. The mesial temporal structures include the hippocampus, amygdala, parahippocampal gyrus, and entorhinal cortex. MTS refers to atrophy and gliosis of the hippocampus with or without involvement of the amygdala. While early surgical treatment in appropriate cases has a positive impact on seizure control and developmental outcomes, MTS in children is more commonly associated with dual pathology than in adults. Exploring this distinction is important with respect to the prevalence in children, as development of MTS as a secondary process would further support the lower incidence in young children. Current understanding of etiology, however, remains multifactorial. Despite its lesser frequency and difficulty of diagnosis in pediatric patients, MTS remains an important cause of refractory epilepsy in the pediatric population with successful treatment options that impact development and therefore merits review and discussion. Keywords:  temporal lobe, mesial, sclerosis, hippocampus

„„ Introduction Refractory epilepsy can have significant deleterious effects on the developing brain. The temporal lobe is the most epileptogenic region and the most common target for resective surgery, with MTS representing the most common etiology in adults.1 In children, however, most localization-related epilepsies have an extratemporal origin, with pure mesial TLE and MTS being less frequent.1 While early surgical treatment in appropriate cases has a positive impact on seizure control and developmental outcomes, MTS in children is more commonly associated with dual pathology than in adults.2 Incomplete myelination patterns in children permit faster spread and secondary generalization, making the diagnosis of temporal onset based on noninvasive modalities arduous. Additionally, the etiology of MTS is a source of discussion with respect to whether the hippocampal insult and atrophy represent the primary etiology or if sclerosis

develops over time as a result of refractory seizures. Exploring this distinction is important with respect to the prevalence in children, as development of MTS as a secondary process would further support the lower incidence in young children. Current understanding of etiology, however, remains multifactorial. Despite its lesser frequency and difficulty of diagnosis in pediatric patients, MTS remains an important cause of refractory epilepsy in the pediatric population with successful treatment options that impact development and therefore merits review and discussion.

„„ Epidemiology Mesial temporal structures include the hippocampus, amygdala, parahippocampal gyrus, and entorhinal cortex. MTS refers to atrophy and gliosis of the hippocampus with or without involvement of the amygdala.3 MTS is the cause for TLE in 50 to 65% of all patients undergoing temporal lobectomy but only in 15% of pediatric patients who undergo epilepsy surgery.4,​5 The prevalence of MTS among children with newly diagnosed seizure disorders is much lower than in adults, representing approximately 1% of refractory pediatric epilepsy.6,​7,​8 Nickels et al9 reported the results of a 30-year cohort of pediatric patients with newly diagnosed epilepsy and estimated that the prevalence of TLE was approximately 8%. In their sample, out of a total number of 468 patients with nonidiopathic focal epilepsy, only 10 patients had MRI findings consistent with MTS (2.1%).9 Ng et al studied the baseline prevalence of childhood MTS.10 A review of 3,100 consecutive brain MRI reports from children of 14 years old or younger demonstrated a 0.77% (24/3,100) incidence of MTS. All patients with MTS in this series initially presented with seizures, so the authors concluded that although MTS is uncommon in children, it invariably presents with seizures. These findings, however, could be affected by selection bias making them hard to extrapolate to the general population. It is logical to infer from those results that the prevalence of pediatric MTS in the general population is low. The incidence of MTS among seizure patients based on histopathological findings was reported by Blumcke et al11 based on tissue specimens resected during epilepsy surgery. They found that MTS was present in 44.5% of adult specimens but only in 15% of samples obtained from pediatric patients.11

334

IVc  Temporal Lobe Epilepsy and Surgical Approaches

„„ Histopathology Pathological variants observed in MTS include hippocampal neuronal loss and gliosis, neuronogenesis, and axonal reorganization.12 The International League Against Epilepsy (ILAE) ­published a consensus statement to unify definitions and terminology of hippocampal sclerosis. The task force proposed a four-group ­system based on cell loss patterns. Type 1 consists of severe cell loss in all layers of the hippocampus, but most predominantly in CA1. Type 2 consists predominantly of cell loss in CA1, while other sections show mild cell loss. Type 3 consists predominantly of cell loss in CA4 and the dentate gyrus. A fourth group was labeled “no hippocampal sclerosis,” gliosis only, due to the fact that up to 20% of samples of patients with MTLE will show reactive gliosis without the classic findings of sclerosis.13 The clinical significance of this classification, however, is undefined. Gales et al14 published the experience from the group at Cleveland Clinic and concluded that there was no relationship between ILAE-defined histological subtypes and surgical outcome related to seizure control.

„„ Pathophysiology Certain types of medically refractory epilepsy may be the result of excitotoxicity secondary to excessive glutamatergic activity. Elevated extracellular glutamate levels, glutamate receptor upregulation, and loss of glutamine synthetase (a glutamatemetabolizing enzyme) have all been demonstrated in affected brain tissue of patients with MTS.15 This observation has been referred to as the “glutamate hypothesis” of MTS pathogenesis. Astrocytes, with their important role in glutamate reuptake and metabolism, probably play an important role in this process. In support of this, excessive astrocyte proliferation, accumulation, and release of astrocytic glutamate have been demonstrated in MTS.15,​16 Prolonged febrile seizures, head injury, nonfebrile status epilepticus, encephalitis, hypertensive encephalopathy, and viruses have all been implicated as potential underlying causes of MTS in children. In these cases, however, it is more likely that the inciting injury leads to seizures, which over time result in sclerosis. Supporting this hypothesis is the fact that MRIs in patients with these underlying etiologies on presentation do not initially demonstrate MTS.17,​18,​19,​20 MTS can also be a late complication of posttransplant cyclosporine-A (CSA) neurotoxicity (including in children). The presentation of those cases involves the development of focal or generalized seizures after treatment with CSA with later development of MTS. It appears then that CSA has a neurotoxic effect that induces seizures, which later lead to the development of sclerosis.21,​22

MTS and Febrile Seizures Febrile convulsions (FCs) occur between 6 months and 5 years of age in relation to an episode of fever (T > 38.3°C) not related to infection of the central nervous system. Simple febrile seizures last less than 15 minutes and have no associated postictal neurological deficits.23,​24 Patients with a history of complex febrile seizures (defined in the National Collaborative Perinatal Project as those lasting greater than 15 minutes or more than two seizures in a 24-hour period)25 have an increased incidence of epilepsy and MTS.25,​26,​27,​28,​29,​30,​31,​32,​33 MRI studies have suggested a link

between MTS and prolonged, focal FCs.28,​29,​30 Up to 30% of TLE patients with MTS in surgical series have a history of prolonged FCs and status epilepticus and 3.5% of patients with a history of febrile seizures later develop epilepsy.25,​30,​31,​34 However, more recent studies with long-term follow-up have questioned the strength of this association. Tarkka et al35 followed 24 patients with prolonged febrile seizures, 8 with an unprovoked seizure after the first event and 32 control subjects with a single simple febrile seizure over a mean of 12.3 years and found that none fit MRI criteria for MTS at the time of follow-up. Three potential hypotheses exist to explain the association with MTS and FCs. One is that FCs cause MTS through acute hippocampal injury, which then later results in the emergence of TLE. The second hypothesis is that FCs and MTS are both a consequence of another abnormality that ultimately results in TLE. Finally, MTS may precede FCs, indicating that the sclerotic hippocampus was the result of an independent insult, and served as the subsequent seizure focus.32,​36

Human Herpes Virus 6 Human herpes virus 6 (HHV-6) is a ubiquitous betaherpesvirus associated with roseola infantum. HHV-6 has been shown to cause limbic encephalitis in immunocompromised hosts and is thought to cause seizures, meningitis, and multiple sclerosis in otherwise healthy individuals.19,​20 Theodore et al examined TLE specimens and discovered that hippocampal astrocytes contained active HHV-6B in approximately ­two-thirds of MTS patients.20 They proposed that HHV-6B may cause glutamatergic excitotoxicity by inhibiting transport of excitatory amino acids within astrocytes, leading to sclerosis.

Dual Pathology Postsurgical pathological evaluation of temporal lobectomy specimens in patients with MTS has often revealed not only ­ hippocampal sclerosis, but also coexistent malformations of cortical development (MCD). Mohamed et al37 reported mildto-moderate MCD in 79% of specimens received from 34 children and adolescents who underwent anteromesial temporal resection for MTS. Although the interictal electroencephalogram (EEG) in these patients was not as localizing as patients with MTS alone, the finding of dual pathology did not predict a worse ­postsurgical outcome. In contrast, Kan et al38 identified 19 patients with MTS, 3 of which had dual pathology (16%) and of those, only one (33%) was seizure free after surgical resection.

„„ Differential Diagnosis Pathological abnormalities associated with temporal lobe seizures in infancy and childhood include MCD, migrational ­ ­disorders, hamartomas, low-grade brain tumors (such as astrocytomas, gangliogliomas, and dysembryoplastic neuroepithelial tumors), and vascular malformations. These abnormalities have been identified pathologically or radiographically in 15 to 20% of children with TLE.5,​39,​40 Of these, glial and neuronal neoplasms account for 10 to 15%.6 These lesions may not be apparent even on high-resolution MRI until later in childhood and are therefore in the differential diagnosis of nonlesional TLE, the other significant differential diagnosis of which is MTS. Mesial temporal lesions,

35  Mesial Temporal Sclerosis in Pediatric Epilepsy sclerotic, migrational, and proliferative, may be undetectable radiographically and only diagnosed once a specimen has been obtained for histopathology. The surgical approach therefore is dictated by seizure semiology, EEG, positron emission tomography (PET), and sometimes subdural electrodes in these cases.

„„ Presentation TLE in infants and toddlers can be challenging to diagnose based on semiology alone. There is an inverse relationship between age and ictal motor manifestations. Motor manifestations can include spasms, tonic, clonic, myoclonic, and hypermotor seizures. The increased motor manifestations, which tend to be bilateral, are thought to be the product of incomplete myelination. Automatisms in the younger population can occur, but are typically limited by the small repertoire of voluntary fine motor skills in this population. More commonly, temporal seizures in children present with behavioral arrest, staring, and lip cyanosis.9 Bourgeois6 described the distinctive features of complex partial seizures in infants with temporal lobe epilepsy as follows: “(1) a predominance of behavioral arrest with possible impairment of consciousness; (2) no identifiable aura; (3) automatisms that are discrete and mostly orofacial; (4) more prominent convulsive activity; and (5) a longer duration (> 1 minute).” ­ Brockhaus and Elger studied 29 children with TLE and found that symmetrical limb motor signs, posturing (as expected in frontal lobe seizures), and head nodding were the most common signs.41 Early school-age children will have more lateralizing signs including dystonic posturing, versive head turning, and eye deviation, all of which can be highly localizing to the contralateral hemisphere. Auras can be hard to assess at that age; the clinician must assess any abnormal behavior such as a child crying and running to the parent (a possible expression of ictal fear or response to an unexpected sensation).9 Older patients and adolescents will present in a way similar to what has been classically described in adults, including rising epigastric sensation, déjà vu, psychic auras (e.g., fear, anxiety, or other strong emotions), olfactory auras, or autonomic changes (e.g., tachycardia, pallor, mydriasis). The ictal event typically lasts 1 to 2 minutes and consists clinically of staring and automatisms such as lip smacking, puckering, chewing, or swallowing, hand picking, rubbing, or fumbling. Importantly, hand automatisms are seen more frequently ipsilateral to the MTS with contralateral dystonic posturing.42 Patients can behave in a semi-purposeful manner during these episodes but do not retain full awareness of the event. The postictal period may be of variable duration and may last up to several hours.43 Absence of clinical findings typical of older children and adults should not rule out the diagnosis, however, as semiology can represent rapid spread to extratemporal regions.44 Adults with MTS often have memory impairment specific to the hemisphere involved (i.e., verbal memory impairment with dominant hemisphere disease and nonverbal learning impairment in nondominant MTS). However, children tend to have less specific neuropsychological deficits with impairment in longterm memory as well as both verbal and nonverbal learning.45 Patients with evidence of early bilateral MTS are at increased risk for more severe impairments in learning and memory.46

„„ Evaluation Detailed neurological examination should focus on evidence of memory and language dysfunction and additional signs of focal neurologic deficits. The standard of care in evaluating a patient for the cause of a first seizure is to obtain a routine 30-minute EEG.47,​48 Nevertheless, the probability of capturing a seizure during a spot-video-EEG in a patient with a mean seizure frequency of once a week is estimated to be approximately 1%.49 Therefore, most patients will require long-term video-EEG monitoring.

Electroencephalography and Electrocorticography The interictal EEG in patients with MTS can take the form of nonepileptiform abnormalities, epileptiform discharges, or a combination of both. Nonepileptiform changes include focal dysrhythmia or slowing in the form of unilateral or bilateral theta and delta activity. These findings can be intermittent or persistent. The presence of temporal intermittent rhythmic delta activity (TIRDA) can be seen in up to 25% of patients with TLE and is more commonly associated with epileptiform discharges.50 This finding, however, is not specific to MTS. Epileptiform interictal changes include unilateral or bilateral independent anterior temporal sharp waves and spikes (Fig. 35.1a).51,​52,​53 Anterior temporal spikes are present in up to 95% of patients with MTS.50 Negative spikes in the middle and posterior temporal electrodes (T3–T4, T5–T6) most likely originate from the temporal neocortex. Interictal spikes can increase during sleep, which stresses the importance of longterm video-EEG monitoring for adequate localization. Lateralized interictal epileptiform discharges (IEDs) correlate well with seizure onset. However, patients with bilateral discharges can still be candidates for surgery and have a good surgical outcome depending on the degree to which IEDs lateralize to the side of the pathology.50 The ictal EEG can be unremarkable during an aura or even at the beginning of the clinical seizure. The typical ictal pattern consists of theta (5–7 Hz) to low alpha (8–9 Hz) frequency rhythmic sharp activity originating over the anterior temporal region either at the time of ictal onset (initial focal onset) or within 30 seconds of ictal onset (delayed focal onset) (Fig. 35.1b).54,​55 This characteristic rhythm occurs in almost 90% of patients with MTS and has a high lateralizing specificity.50 In children, the initial ictal pattern is often preceded by a brief period of low-voltage fast activity and can involve more diffuse generalized or bilateral activity.37,​41 Patients with ictal and interictal patterns on scalp recording that are inconclusive or with no evidence of MTS on routine neuroimaging may require intracranial monitoring with subdural grid, strips, or depth electrodes to further differentiate MTS from neocortical or extratemporal seizure foci. Studies reviewing these recordings have shown that patients with clear radiographic hippocampal sclerosis can have seizure foci distant from the hippocampus, including the temporal pole,56 amygdala,57 or perisylvian cortex.58

335

336

IVc  Temporal Lobe Epilepsy and Surgical Approaches

Fig. 35.1  (a) Interictal left anterior temporal sharp wave (longitudinal bipolar montage). (b) Onset of a seizure showing alpha-frequency rhythmic spiking, maximal over the right posterior temporal head region (longitudinal bipolar montage).

Magnetic Resonance Imaging A variety of neuroimaging modalities have been used to detect MTS. Routine MRI is relatively insensitive in detecting MTS.59 MRI with thin, coronal, oblique sections through the temporal lobes, hippocampi, and amygdala using high-resolution, T1-weighted images with inversion recovery, and T2-weighted images with spin echo or fast spin echo is considered the gold standard in radiographically diagnosing MTS. Sensitivity of high-resolution MRI in detecting MTS is as high as 97% with a specificity of 83%.60,​61 Classic findings of MTS on MRI include decreased hippocampal volume and increased T2 signal of the mesial temporal structures (Fig. 35.2a, b). T2 hyperintensity represents gliosis and increased free water content within the hippocampus. Associated findings

include atrophy of the adjacent fornix, mammillary body, or other limbic system structures.62 Fluid-attenuated inversion recovery imaging can demonstrate abnormal signal intensity in the hippocampus, taking into account that normal limbic structures are slightly hyperintense relative to the neocortex. Hippocampal volume calculations also can be performed through volumetric thin-section T1-weighted imaging analysis (volumetric MRI) and provide significant improvement in detection of mild unilateral MTS.63,​64 The significance of these findings varies with age, because the hippocampal volume of children younger than 6 years increases linearly with age.65 Magnetic resonance spectroscopy can be helpful in identifying a mesial temporal etiology or in lateralizing the onset zone as it demonstrates the anatomical distribution of biological metabolites, which are regionally abnormal within epileptogenic foci.66,​67

35  Mesial Temporal Sclerosis in Pediatric Epilepsy

Fig. 35.2  (a) Fluid-attenuated inversion recovery (FLAIR) sequence of coronal MRI with thin cuts through the hippocampus reveals signal hyperintensity in the left hippocampus. (b) FLAIR sequence of coronal MRI demonstrates decreased volume of the left hippocampus. (c) Axial 2-deoxy-2[18F] fluorodglucose-positron emission tomography scan demonstrating left anterior temporal hypometabolism.

Positron Emission Tomography and Single Proton Emission Tomography PET scanning with [18F] fluorodeoxyglucose reveals hypometabolism in the mesial temporal structures during the i­nterictal period in patients with MTS (Fig. 35.2c). The temporal pole and lateral temporal cortex also may be hypometabolic in MTS patients as well as extratemporal regions such as the parietal cortex, orbitofrontal cortex, and insula depending on the area of electrical spread and ultimate seizure network.68,​69 Hypometabolism in the temporopolar region predicts a favorable postoperative outcome with respect to seizure control.70 MTS patients without radiographic evidence of hippocampal sclerosis can also be i­dentified on the basis of PET results revealing ipsilateral hippocampal and adjacent cortical hypometabolism.71 Single-photon emission computed tomography (SPECT) scans are performed after injection of a radiotracer ictally and interictally to reveal differences in distribution of blood flow. During a seizure, there is relative hyperperfusion at the site of ictal onset and, therefore, greatest uptake of radiotracer. Interictally, there is relative hypoperfusion and hence a decrease in radiotracer uptake compared with the surrounding unaffected brain tissue. In a meta-analysis of SPECT imaging in localization of epileptic foci, Devous et al found the overall sensitivity for SPECT in seizure localization for patients with TLE was 44% interictally, 75% postictally, and 97% ictally.72 While SPECT is therefore a good tool in identifying mesial temporal onset in the refractory epilepsy population in general, few studies have confirmed efficacy of ictal– interictal SPECT in children.73,​74

Magnetoencephalography Magnetoencephalography (MEG) has evolved to become an important adjunct to the presurgical evaluation of patients with refractory epilepsy. It is a noninvasive diagnostic test that detects magnetic fields associated with electrical currents that emanate from neuronal activity. Magnetic dipoles can be coregistered with MRI to take advantage of the temporal resolution of MEG and the spatial resolution of MR. The fusion of these two

­ iagnostic modalities is known as magnetic source localization. d There is a paucity of data regarding the clinical utility of magnetic source localization in pediatric patients with refractory epilepsy secondary to MTS. Iida et al reviewed their surgical experience in 16 pediatric patients who underwent MEG as part of their presurgical workup. Their results were limited to a small sample and were not exclusively focused on temporal lobe epilepsy. They classified spike sources into classes depending on the degree of dipole clustering. Their results suggest that class I clusters (those with 20 or more spike sources with 1 cm or less between sources) indicate the presence of an epileptogenic zone and require complete excision. The resection of areas with a dense cluster of spike sources correlates with good surgical outcomes.75 Large series analyzing the outcomes of MEG-guided resections in pediatric patients exclusively with MTLE have not been published. Observational data suggest that MEG correctly identifies the irritative zone in MTS74 and helps predict response to surgical management.76

„„ Treatment Antiepileptic Medications Patients who have a single, unprovoked seizure have an approximate subsequent seizure risk of 50%.40 Once the patient has suffered a second unprovoked seizure, the risk of further seizures within the next 4 years increases to 75%.77 Of patients newly diagnosed with epilepsy, roughly 60% will respond to initial monotherapy and 40% will ultimately be considered refractory to pharmacological treatment.78 Stephen et al78 studied 73 patients with newly diagnosed MTS and found that 42% responded to pharmacotherapy. Only 48% of these patients required more than one agent to become seizure free, suggesting an important role of anticonvulsants as initial management in the treatment of TLE in MTS.78 Aggressive anticonvulsant management at the time of diagnosis is important to help prevent psychosocial dysfunction and possibly limit cognitive decline and interference with

337

338

IVc  Temporal Lobe Epilepsy and Surgical Approaches school performance. Medication titration should be implemented with the goal of achieving seizure freedom with minimal side effects. If the initial medication is ineffective, the agent should be changed or a second agent added with a different mechanism of action from the first. Overall, success in obtaining seizure freedom with multiple agents in MTS once monotherapy fails is poor.79,​80 Previous studies support the use of sodium channel blockers (e.g., phenytoin, carbamazepine), although all commercially available anticonvulsants (with the exception of ethosuximide) have been found to be useful.81,​82 Of the newer agents, only topiramate carries a level A recommendation as adjunctive therapy in children with refractory partial epilepsy.83,​84

Laser Interstitial Thermal Therapy Laser interstitial thermal therapy (LITT) has gained popularity as a less invasive way of treating patients with refractory epilepsy. LITT has the potential to focally ablate epileptogenic tissue while minimizing disruption to the normal surrounding parenchyma. The advantages of this approach are the decrease in operative and recovery time and the potential for less neuropsychological morbidity by sparing the lateral neocortex. However, long-term data to support the neuropsychological and seizure outcomes in comparison to open resection in children with MTS are lacking.85 Experience with this technique in pediatric patients is increasing.86,​87 Large series of pediatric patients with MTS treated with LITT are still required and longer follow-up will be needed from the initial series to compare outcomes with open surgical resection.

Resective Surgery Once patients have failed, optimal treatment with multiple anticonvulsants, surgical intervention becomes a consideration.88 Delaying surgery can worsen long-term outcomes, especially in the pediatric population; intractable epilepsy has a negative impact on cognition and behavior.15,​45,​78 Functional MRI is performed in select patients to localize sensorimotor and language function prior to resection, and the Wada test (intracarotid amobarbital) is used to lateralize language and memory. Patients may undergo intraoperative mapping and electrocorticography, or extraoperative invasive recordings, to more precisely identify an epileptogenic focus. Ideally, these measures yield concordant information regarding the localization of the epileptogenic zone and triangulate on the temporal lobe in MTS. However, some studies in the pediatric population have demonstrated favorable postsurgical outcomes even when

results are discordant.41,​89 With subdural grids or strips in place, patients may undergo functional localization mapping with electrocortical stimulation. The goal of mapping is to define eloquent cortex that should be avoided during a surgical resection. Several surgical strategies have been used to treat patients with TLE including anterior mesial temporal lobectomy and selective amygdalohippocampectomy. The latter has been proposed as a way to minimize neocortical disruption. In a review of previous literature addressing seizure outcome after temporal lobe resection in children, Wyllie reported an overall 74 to 82% rate of seizure freedom.90 Clusmann et al91 evaluated 89 children with TLE, ages 1 to 18 years for an average of 46 months after surgery and found that maximum seizure control was obtained with anteromesial temporal resection (AMTR) (95%) compared with subarachnoid hemorrhage (SAH) (75%), although this study included TLE patients with and without MTS. Cohen-Gadol et al92 reported 2-year postoperative outcomes in unilateral MTS patients diagnosed on the basis of surgical pathology results that underwent anterior temporal lobectomy with hippocampectomy. They evaluated the percentage of patients that remained Engel class 1 and found that 86% were seizure free at 6 months, 83% at 1 year, 80% at 2 years, and 79% at 5 and 10 years. Baldauf et al93 reported a greater than 90% remission rate in 41 patients diagnosed with unilateral MTS on the basis of four or more interictal EEG recordings and diagnostic MRI results alone, followed for more than 3 years after cortico-amygdalohippocampectomy. In a pediatric population of carefully selected patients diagnosed with MTS on the basis of ictal semiology, EEG and MRI findings who undergo temporal lobectomy, 73 to 100% receive benefit from surgery.5 Surgical outcomes are predictably not as favorable for patients with bilateral disease as those with clear unilateral MTS,37 and other treatment modalities must be considered.

„„ Conclusion MTS is not as frequent in children as it is in adults. Pediatric patients have a higher rate of dual pathology and delayed diagnosis due to variability in presentation and difficulty in interpreting semiology in this age group. Large series focusing on outcomes of pediatric patients with MTS are lacking. The gold standard for treatment of refractory MTS in children remains AMTR. Less invasive forms of treatment, such as LITT, have emerged but sufficient evidence to recommend these modalities over open surgical approaches is currently lacking.

References 1. Bourgeois M, Sainte-Rose C, Lellouch-Tubiana A, et al. Surgery of epilepsy associated with focal lesions in childhood. J Neurosurg 1999;90(5):833–842 2. Lee YJ, Lee JS. Temporal lobe epilepsy surgery in children versu adults: from etiologies to outcomes. Korean J Pediatr 2013;56(7):275–281 3. Thom M. Review: hippocampal sclerosis in epilepsy: a neuropathology review. Neuropathol Appl Neurobiol 2014;40(5):520–543

4. National Institutes of Health Consensus Conference. Surgery for epilepsy. JAMA 1990;264(6):729–733 5. Blume WT, Hwang PA. Pediatric candidates for temporal lobe epilepsy surgery. Can J Neurol Sci 2000;27(Suppl 1):S14–S19, discussion S20–S21 6. Bourgeois BF. Temporal lobe epilepsy in infants and children. Brain Dev 1998;20(3):135–141 7. King MA, Newton MR, Jackson GD, et al. Epileptology of the first-seizure presentation: a clinical, electroencephalographic,

35  Mesial Temporal Sclerosis in Pediatric Epilepsy and magnetic resonance imaging study of 300 consecutive patients. Lancet 1998;352(9133):1007–1011 8. Berg AT, Testa FM, Levy SR, Shinnar S. Neuroimaging in children with newly diagnosed epilepsy: a community-based study. Pediatrics 2000;106(3):527–532 9. Nickels KC, Wong-Kisiel LC, Moseley BD, Wirrell EC. Temporal lobe epilepsy in children. Epilepsy Res Treat 2012;2012:849540

30. Davies KG, Hermann BP, Dohan FC Jr, Foley KT, Bush AJ, Wyler AR. Relationship of hippocampal sclerosis to duration and age of onset of epilepsy, and childhood febrile seizures in temporal lobectomy patients. Epilepsy Res 1996;24(2):119–126 31. Verity CM, Golding J. Risk of epilepsy after febrile convulsions: a national cohort study. BMJ 1991;303(6814):1373–1376 32. Lewis DV. Febrile convulsions and mesial temporal sclerosis. Curr Opin Neurol 1999;12(2):197–201

10. Ng YT, McGregor AL, Duane DC, Jahnke HK, Bird CR, Wheless JW. Childhood mesial temporal sclerosis. J Child Neurol 2006;21(6):512–517

33. Annegers JF, Hauser WA, Shirts SB, Kurland LT. Factors prognostic of unprovoked seizures after febrile convulsions. N Engl J Med 1987;316(9):493–498

11. Blumcke I, Spreafico R, Haaker G, et al; EEBB Consortium. Histopathological findings in brain tissue obtained during epilepsy surgery. N Engl J Med 2017;377(17):1648–1656

34. Cendes F. Febrile seizures and mesial temporal sclerosis. Curr Opin Neurol 2004;17(2):161–164

12. Velísek L, Moshé SL. Temporal lobe epileptogenesis and epilepsy in the developing brain: bridging the gap between the laboratory and the clinic. Progression, but in what direction? Epilepsia 2003;44(Suppl 12):51–59 13. Blümcke I, Thom M, Aronica E, et al. International consensus classification of hippocampal sclerosis in temporal lobe epilepsy: a Task Force report from the ILAE Commission on Diagnostic Methods. Epilepsia 2013;54(7):1315–1329 14. Gales JM, Jehi L, Nowacki A, Prayson RA. The role of histopathologic subtype in the setting of hippocampal sclerosis-associated mesial temporal lobe epilepsy. Hum Pathol 2017;63:79–88 15. Eid T, Williamson A, Lee TS, Petroff OA, de Lanerolle NC. Glutamate and astrocytes—key players in human mesial temporal lobe epilepsy? Epilepsia 2008;49(Suppl 2):42–52 16. Liu Z, Mikati M, Holmes GL. Mesial temporal sclerosis: pathogenesis and significance. Pediatr Neurol 1995;12(1):5–16 17. Scott RC, Gadian DG, Cross JH, Wood SJ, Neville BG, Connelly A. Quantitative magnetic resonance characterization of mesial temporal sclerosis in childhood. Neurology 2001;56(12):1659–1665 18. Solinas C, Briellmann RS, Harvey AS, Mitchell LA, Berkovic SF. Hypertensive encephalopathy: antecedent to hippocampal sclerosis and temporal lobe epilepsy? Neurology 2003;60(9):1534–1536 19. Donati D, Akhyani N, Fogdell-Hahn A, et al. Detection of human herpesvirus-6 in mesial temporal lobe epilepsy surgical brain resections. Neurology 2003;61(10):1405–1411 20. Theodore WH, Epstein L, Gaillard WD, Shinnar S, Wainwright MS, Jacobson S. Human herpes virus 6B: a possible role in epilepsy? Epilepsia 2008;49(11):1828–1837 21. Faraci M, Lanino E, Dallorso S, et al. Mesial temporal sclerosis—a late complication in four allogeneic pediatric recipients with persistent seizures after an acute episode of cyclosporine-A neurotoxicity. Bone Marrow Transplant 2003;31(10):919–922 22. Gaggero R, Haupt R, Paola Fondelli M, et al. Intractable epilepsy secondary to cyclosporine toxicity in children undergoing allogeneic hematopoietic bone marrow transplantation. J Child Neurol 2006;21(10):861–866 23. Gupta A. Febrile seizures. Continuum (Minneapolis, Minn) 2016;22(1 Epilepsy):51–5–9

35. Tarkka R, Pääkkö E, Pyhtinen J, Uhari M, Rantala H. Febrile seizures and mesial temporal sclerosis: no association in a longterm follow-up study. Neurology 2003;60(2):215–218 36. Arzimanoglou A, Guerrini R, Aicardi J. Aicardi’s Epilepsy in Children. 3rd ed. Philadelphia, PA: Lippincott, Williams & Wilkins; 2004 37. Mohamed A, Wyllie E, Ruggieri P, et al. Temporal lobe epilepsy due to hippocampal sclerosis in pediatric candidates for epilepsy surgery. Neurology 2001;56(12):1643–1649 38. Kan P, Van Orman C, Kestle JR. Outcomes after surgery for focal epilepsy in children. Childs Nerv Syst 2008;24(5):587–591 39. Wyllie E, Comair YG, Kotagal P, Bulacio J, Bingaman W, Ruggieri P. Seizure outcome after epilepsy surgery in children and adolescents. Ann Neurol 1998;44(5):740–748 40. Ray A, Wyllie E. Treatment options and paradigms in childhood temporal lobe epilepsy. Expert Rev Neurother 2005;5(6):785–801 41. Brockhaus A, Elger CE. Complex partial seizures of temporal lobe origin in children of different age groups. Epilepsia 1995;36(12):1173–1181 42. Kotagal P, Lüders H, Morris HH, et al. Dystonic posturing in complex partial seizures of temporal lobe onset: a new lateralizing sign. Neurology 1989;39(2, Pt 1):196–201 43. Engel J Jr. Mesial temporal lobe epilepsy: what have we learned? Neuroscientist 2001;7(4):340–352 44. Cendes FKP, Brodie M, Andermann F. The Mesio-Temporal Lobe Epilepsy Syndrome. 3rd ed. Eastleigh: John Libbey; 2002 45. Aldenkamp AP, Alpherts WC, Dekker MJ, Overweg J. Neuropsychological aspects of learning disabilities in epilepsy. Epilepsia 1990;31(Suppl 4):S9–S20 46. DeLong GR, Heinz ER. The clinical syndrome of early-life bilateral hippocampal sclerosis. Ann Neurol 1997;42(1):11–17 47. Hirtz D, Berg A, Bettis D, et al; Quality Standards Subcommittee of the American Academy of Neurology. Practice Committee of the Child Neurology Society. Practice parameter: treatment of the child with a first unprovoked seizure: report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society. Neurology 2003;60(2):166–175

25. Nelson KB, Ellenberg JH. Predictors of epilepsy in children who have experienced febrile seizures. N Engl J Med 1976;295(19):1029–1033

48. Krumholz A, Wiebe S, Gronseth G, et al; Quality Standards Subcommittee of the American Academy of Neurology. American Epilepsy Society. Practice parameter: evaluating an apparent unprovoked first seizure in adults (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology 2007;69(21):1996–2007

26. Cendes F, Li LM, Andermann F, et al. Dual pathology and its clinical relevance. Adv Neurol 1999;81:153–164

49. Sundaram M, Sadler RM, Young GB, Pillay N. EEG in epilepsy: current perspectives. Can J Neurol Sci 1999;26(4):255–262

27. Falconer MA, Serafetinides EA, Corsellis JA. Etiology and pathogenesis of temporal lobe epilepsy. Arch Neurol 1964;10:233–248

50. Javidan M. Electroencephalography in mesial temporal lobe epilepsy: a review. Epilepsy Res Treat 2012;2012:637430

28. Scott RC, Gadian DG, King MD, et al. Magnetic resonance imaging findings within 5 days of status epilepticus in childhood. Brain 2002;125(Pt 9):1951–1959

51. Gambardella A, Gotman J, Cendes F, Andermann F. The relation of spike foci and of clinical seizure characteristics to different patterns of mesial temporal atrophy. Arch Neurol 1995;52(3):287–293

24. Knudsen FU. Febrile seizures: treatment and prognosis. Epilepsia 2000;41(1):2–9

29. VanLandingham KE, Heinz ER, Cavazos JE, Lewis DV. Magnetic resonance imaging evidence of hippocampal injury after prolonged focal febrile convulsions. Ann Neurol 1998;43(4):413–426

52. Cascino GD, Trenerry MR, So EL, et al. Routine EEG and temporal lobe epilepsy: relation to long-term EEG monitoring, quantitative MRI, and operative outcome. Epilepsia 1996;37(7):651–656

339

340

IVc  Temporal Lobe Epilepsy and Surgical Approaches 53. Gilliam F, Bowling S, Bilir E, et al. Association of combined MRI, interictal EEG, and ictal EEG results with outcome and pathology after temporal lobectomy. Epilepsia 1997;38(12):1315–1320

74. Pataraia E, Lindinger G, Deecke L, Mayer D, Baumgartner C. Combined MEG/EEG analysis of the interictal spike complex in mesial temporal lobe epilepsy. Neuroimage 2005;24(3):607–614

54. Ebersole JS, Pacia SV. Localization of temporal lobe foci by ictal EEG patterns. Epilepsia 1996;37(4):386–399

75. Iida K, Otsubo H, Matsumoto Y, et al. Characterizing magnetic spike sources by using magnetoencephalography-guided neuronavigation in epilepsy surgery in pediatric patients. J Neurosurg 2005;102(2, Suppl):187–196

55. Assaf BA, Ebersole JS. Visual and quantitative ictal EEG predictors of outcome after temporal lobectomy. Epilepsia 1999;40(1):52–61 56. Kahane P, Chabardès S, Minotti L, Hoffmann D, Benabid AL, Munari C. The role of the temporal pole in the genesis of temporal lobe seizures. Epileptic Disord 2002;4(Suppl 1):S51–S58 57. Spanedda F, Cendes F, Gotman J. Relations between EEG seizure morphology, interhemispheric spread, and mesial temporal ­ atrophy in bitemporal epilepsy. Epilepsia 1997;38(12):1300–1314 58. Kahane PHJ, Hoffman D. Perisylvian Cortex Involvement in Seizures Affecting the Temporal Lobe. London: John Libbey; 2001 59. McBride MC, Bronstein KS, Bennett B, Erba G, Pilcher W, Berg MJ. Failure of standard magnetic resonance imaging in patients with refractory temporal lobe epilepsy. Arch Neurol 1998;55(3):346–348 60. Berkovic SF, McIntosh AM, Kalnins RM, et al. Preoperative MRI predicts outcome of temporal lobectomy: an actuarial analysis. Neurology 1995;45(7):1358–1363 61. Jack CR Jr, Rydberg CH, Krecke KN, et al. Mesial temporal sclerosis: diagnosis with fluid-attenuated inversion-recovery versus spin-echo MR imaging. Radiology 1996;199(2):367–373 62. Chan S, Erickson JK, Yoon SS. Limbic system abnormalities associated with mesial temporal sclerosis: a model of chronic cerebral changes due to seizures. Radiographics 1997;17(5):1095–1110 63. Ho SS, Kuzniecky RI, Gilliam F, Faught E, Morawetz R. Temporal lobe developmental malformations and epilepsy: dual pathology and bilateral hippocampal abnormalities. Neurology 1998;50(3):748–754 64. Free SL, Li LM, Fish DR, Shorvon SD, Stevens JM. Bilateral hippocampal volume loss in patients with a history of encephalitis or meningitis. Epilepsia 1996;37(4):400–405 65. Szabó CA, Wyllie E, Siavalas EL, et al. Hippocampal volumetry in children 6 years or younger: assessment of children with and without complex febrile seizures. Epilepsy Res 1999;33(1):1–9 66. Capizzano AA, Vermathen P, Laxer KD, et al. Temporal lobe epilepsy: qualitative reading of 1H MR spectroscopic images for presurgical evaluation. Radiology 2001;218(1):144–151 67. Li LM, Cendes F, Antel SB, et al. Prognostic value of proton magnetic resonance spectroscopic imaging for surgical outcome in patients with intractable temporal lobe epilepsy and bilateral hippocampal atrophy. Ann Neurol 2000;47(2):195–200 68. Semah F, Baulac M, Hasboun D, et al. Is interictal temporal hypometabolism related to mesial temporal sclerosis? A positron emission tomography/magnetic resonance imaging confrontation. Epilepsia 1995;36(5):447–456 69. Arnold S, Schlaug G, Niemann H, et al. Topography of interictal glucose hypometabolism in unilateral mesiotemporal epilepsy. Neurology 1996;46(5):1422–1430 70. Dupont S, Semah F, Clémenceau S, Adam C, Baulac M, Samson Y. Accurate prediction of postoperative outcome in mesial temporal lobe epilepsy: a study using positron emission tomography with 18fluorodeoxyglucose. Arch Neurol 2000;57 (9):1331–1336 71. Carne RP, O’Brien TJ, Kilpatrick CJ, et al. ‘MRI-negative PET-positive’ temporal lobe epilepsy (TLE) and mesial TLE differ with quantitative MRI and PET: a case control study. BMC Neurol 2007;7:16 72. Devous MD Sr, Thisted RA, Morgan GF, Leroy RF, Rowe CC. SPECT brain imaging in epilepsy: a meta-analysis. J Nucl Med 1998;39(2):285–293 73. Harvey AS, Bowe JM, Hopkins IJ, Shield LK, Cook DJ, Berkovic SF. Ictal 99mTc-HMPAO single photon emission computed tomography in children with temporal lobe epilepsy. Epilepsia 1993;34(5):869–877

76. Assaf BA, Karkar KM, Laxer KD, et al. Magnetoencephalography source localization and surgical outcome in temporal lobe epilepsy. Clin Neurophysiol 2004;115(9):2066–2076 77. Hauser WA, Rich SS, Lee JR, Annegers JF, Anderson VE. Risk of recurrent seizures after two unprovoked seizures. N Engl J Med 1998;338(7):429–434 78. Stephen LJ, Kwan P, Brodie MJ. Does the cause of localisationrelated epilepsy influence the response to antiepileptic drug treatment? Epilepsia 2001;42(3):357–362 79. Semah F, Picot MC, Adam C, et al. Is the underlying cause of epilepsy a major prognostic factor for recurrence? Neurology 1998;51(5):1256–1262 80. Kim WJ, Park SC, Lee SJ, et al. The prognosis for control of seizures with medications in patients with MRI evidence for mesial temporal sclerosis. Epilepsia 1999;40(3):290–293 81. Brodie MJ, Dichter MA. Antiepileptic drugs. N Engl J Med 1996;334(3):168–175 82. Brodie MJ. Management strategies for refractory localizationrelated seizures. Epilepsia 2001;42(Suppl 3):27–30 83. French JA, Kanner AM, Bautista J, et al; Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Quality Standards Subcommittee of the American Academy of Neurology. American Epilepsy Society. Efficacy and tolerability of the new antiepileptic drugs I: treatment of new onset epilepsy: report of the Therapeutics and Technology Assessment Subcommittee and Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology 2004;62(8):1252–1260 84. French JA, Kanner AM, Bautista J, et al; Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Quality Standards Subcommittee of the American Academy of Neurology. American Epilepsy Society. Efficacy and tolerability of the new antiepileptic drugs II: treatment of refractory epilepsy: report of the Therapeutics and Technology Assessment Subcommittee and Quality Standards Subcommittee of the American Academy of Neurology and the American Epilepsy Society. Neurology 2004;62(8):1261–1273 85. Liscak R, Malikova H, Kalina M, et al. Stereotactic radiofrequency amygdalohippocampectomy in the treatment of mesial temporal lobe epilepsy. Acta Neurochir (Wien) 2010;152(8):1291–1298 86. Lewis EC, Weil AG, Duchowny M, Bhatia S, Ragheb J, Miller I. MR-guided laser interstitial thermal therapy for pediatric drug-­ resistant lesional epilepsy. Epilepsia 2015;56(10):1590–1598 87. Curry DJ, Gowda A, McNichols RJ, Wilfong AA. MR-guided stereotactic laser ablation of epileptogenic foci in children. Epilepsy Behav 2012;24(4):408–414 88. Brodie MJ, French JA. Management of epilepsy in adolescents and adults. Lancet 2000;356(9226):323–329 89. Castro LH, Serpa MH, Valério RM, et al. Good surgical outcome in discordant ictal EEG-MRI unilateral mesial temporal sclerosis patients. Epilepsia 2008;49(8):1324–1332 90. Wyllie E. Surgical treatment of epilepsy in children. Pediatr Neurol 1998;19(3):179–188 91. Clusmann H, Kral T, Gleissner U, et al. Analysis of different types of resection for pediatric patients with temporal lobe epilepsy. Neurosurgery 2004;54(4):847–859, discussion 859–860 92. Cohen-Gadol AA, Wilhelmi BG, Collignon F, et al. Long-term outcome of epilepsy surgery among 399 patients with nonlesional seizure foci including mesial temporal lobe sclerosis. J Neurosurg 2006;104(4):513–524 93. Baldauf CM, Cukiert A, Argentoni M, et al. Surgical outcome in patients with refractory epilepsy associated to MRIdefined unilateral mesial temporal sclerosis. Arq Neuropsiquiatr 2006;64(2B):363–368

36

  Anteromesial Temporal Lobectomy Oğuz Çataltepe

Summary Anteromesial temporal lobectomy (AMTL) is the most commonly performed procedure for surgical treatment of the patients with temporal lobe epilepsy (TLE). In this technique, anterior temporal neocortex is resected along the mesial temporal structures, including amygdala and hippocampus. This chapter gives a step-by-step description of the surgical technique we use for AMTL. However, surgical technique may change based on the underlying lesion and extent of the epileptogenic zone, especially in patients with cortical dysplasia. In general, our temporal lobe resection includes the anterior 3.5 cm of the temporal neocortex in the dominant hemisphere by sparing most of the superior temporal gyrus. Our resection also includes the uncus, amygdala, and an approximately 3 cm length of the hippocampus–parahippocampus as an en bloc specimen. This technique may be modified depending on the patient’s age, imaging, and electrophysiological characteristics. The most common neuropathological substrates in children are cortical dysplasia and neoplasms followed by gliosis and mesial temporal sclerosis (MTS). Temporal lobe resection is a safe and effective surgical technique in the management of TLE with reported seizure control rates between 60 and 80%. Keywords:  anteromesial temporal lobectomy, hippocampal sclerosis, hippocampectomy, cortical dysplasia

„„ Introduction AMTL is the most commonly performed procedure for surgical treatment of the patients with TLE. Although TLE is more common in adults, AMTL is performed in children as well. AMTL constitutes 30 to 44% of all surgical resections in published pediatric epilepsy surgery series compared to 62 to 73% of cases in adult epilepsy surgery series.1,​2,​3,​4,​5,​6 The main reason for this discrepancy is related to the differences in neuropathological substrates causing epilepsy in children and adults. MTS is the most common reason for AMTL, and it is seen more frequently in adult epilepsy patients compare to children. AMTL is a very effective surgical intervention, and its efficiency in the treatment of children with intractable TLE has been demonstrated in many surgical series.1,​2,​7,​8,​9,​10,​11 This chapter gives a step-by-step description of the surgical technique we use for AMTL. However, surgical technique may change based on the underlying lesion and extent of the epileptogenic zone, especially in patients with cortical­

dysplasia. All patients undergo a comprehensive presurgical assessment by pediatric epilepsy team. Preoperative assessment includes a detailed clinical examination, MRI with epilepsy protocol, electroencephalography (EEG), and long-term video-EEG monitoring to obtain ictal and interictal electrophysiological data. Positron emission tomography (PET) or single-photon emission computed tomography (SPECT), neuropsychological assessment, and intracarotid amobarbital procedure (the Wada test) are among the other commonly used diagnostic modalities and tests. Despite all of these tests, locating the epileptogenic zone remains problematic in a significant number of children with TLE, and these patients frequently are candidates for invasive monitoring. Further details on patient selection criteria, preoperative workup, and surgical techniques for other pathologies can be found in other related chapters in this book.

„„ Historical Evolution of the Surgical Technique Temporal lobe resection in epilepsy surgery is not a standard technique.12,​13 Because temporal lobectomy suggests removal of the entire temporal lobe, anterior temporal lobectomy and anteromesial temporal lobectomy are more appropriate terms for this technique. Although Penfield performed his first temporal lobectomy as early as 1928, temporal lobe seizures were defined as a distinct entity in Montreal Neurological Institute (MNI) during the period between 1945 and 1955. Early variations of the surgical technique for temporal lobectomy were developed during this decade.13,​14,​15,​16,​17,​18 The first application of the technique was temporal neocortical resection without removing mesial temporal structures. Thereafter, Wilder Penfield and his colleagues started to resect the hippocampus and uncus along the temporal neocortex and reported better results with this approach. They published their classical report describing anterior temporal lobectomy that included amygdala and hippocampus in 1952.14 Gradually, resection of anteromesial structures became an established surgical approach for TLE.14,​15,​16,​17 After the initial studies regarding the role of hippocampus in the memory function, the surgical technique evolved toward electrophysiologically tailored temporal lobectomy aiming to preserve the hippocampus as much as possible.12,​13,​18,​19 In the mid-1950s, Niemeyer described a new technique: selective transcortical amygdalohippocampectomy.20 Later, Yaşargil and colleagues modified this technique and developed a selective

342

IVc  Temporal Lobe Epilepsy and Surgical Approaches ­amygdalohippocampectomy technique through the transsylvian approach. They reported impressive seizure control rates without removing temporal neocortex.21 The details of this approach are discussed in related chapters of this book. At the MNI, Rasmussen performed anterior temporal lobectomy by including uncus and amygdala and used electrocorticography to determine the extent of hippocampal resection. His approach was to remove just the anterior 1 to 1.5 cm of the hippocampus.22,​23,​24,​25,​26 Conversely, Feindel and colleagues in the same institution (MNI) routinely avoided the removal of hippocampus to preserve memory functions while aggressively resecting the amygdala.27,​28,​29 Then Goldring and colleagues described an anterior temporal lobectomy technique that spares the amygdala.30 All these techniques were reported with variable success rates. Today, the most commonly used surgical technique for TLE is the resection of anterior temporal neocortex along the mesial temporal structures, including amygdala and hippocampus. Even this technique has some variations, including en bloc resection of both neocortex and mesial temporal structures that was first described by Falconer and later applied by Polkey and Crandal.31 Another modification of the technique was described by Spencer at Yale.32 Spencer’s technique is the most commonly used technique today. One of the main differences among these techniques is resection length of anterior temporal lobe (from the tip of anterior temporal lobe). The majority of epilepsy surgeons do not exceed a 4-cm neocortical resection length in the dominant hemisphere, whereas the length of resection may increase up to 5.5 to 6 cm in the nondominant hemisphere. Another difference among surgeons is the intent to spare the superior temporal gyrus during temporal neocortical resection. Many epilepsy surgeons spare the superior temporal gyrus partially or fully to decrease the risk for visual field defect. The extent of the hippocampal resection is also controversial. Although some find it sufficient to remove the anterior 1.5 cm of the hippocampus, others extend their hippocampal resection up to 3 cm by reaching back to the posterior part of the tail.12,​13 Visual field cut is a common problem in temporal lobe resection cases. Visual field cut is closely related to anterior limit of Meyer’s loop (ML). In one study, up to 19-mm interindividual variation in the degree of the anterior extent of ML is reported.33 In this study, average upper limit of temporal loop–ML distance was found 35.4 mm on the left and 38.1 mm on the right side in normal subjects. In the same study, upper limit of temporal lobe resection to avoid complete quadrantanopia was found to be 46.2 mm on the left and 49.7 mm on the right side.

„„ Surgical Technique Here, we will describe the anteromesial temporal lobectomy technique we use at the University of Massachusetts Medical Center (Video 36.1). In general, our temporal lobe resection includes the anterior 3.5 cm of the temporal neocortex in the dominant hemisphere with most of the superior temporal gyrus spared, as described by Spencer et al.32 Our resection also includes the uncus, amygdala, and an approximately 3 cm length of the hippocampus–parahippocampus as an en bloc

Video 36.1 Anterior temporal lobectomy and Amygdalohippocampectomy. (This video is provided courtesy of Oğuz Çataltepe.) ht t ps://www.thieme.de/de/q.ht m?p=opn/ tp/255910102/9781626238176_c036_v001&t=video specimen. The neocortical resection is extended up to 5 cm in the nondominant hemisphere if needed. Mesial structures are resected in the same manner in both dominant and nondominant hemispheres if neuropsychological assessment and the Wada test results are reassuring. This technique may be modified depending on the patient’s age, imaging, and electrophysiological characteristics. If there is radiologically defined dysplastic cortex or an electrophysiologically more extensive abnormality, then our neocortical resection borders are redefined and may be extended further. If the epileptogenic zone is limited to a certain part of the temporal neocortex based on the invasive monitoring data, then the resection may be tailored accordingly. In these cases, mesial structures can be spared, especially in some lesional epilepsy cases. Alternatively, if the radiological finding of hippocampal sclerosis is pronounced at the hippocampal tail as well, then we extend our resection of the hippocampal tail much further than our standard limits. The surgical plan is extensively discussed in advance with the pediatric epilepsy team in a multidisciplinary epilepsy surgery conference, and the extent of resection is predetermined based on the aforementioned considerations.

Positioning the Patient The patient is placed in supine position, and the head is placed in a pin head holder if the patient is older than 3 years. The horseshoe head holder is used for younger patients. A gel roll is placed under the ipsilateral shoulder, and the head is turned to the contralateral side approximately 60 degrees. The neck is slightly extended by lowering the vertex approximately 15 degrees downward, just enough to bring the zygoma to the surgeon’s eyeline and to make the zygoma the most prominent point on the midline. Lastly, the occiput is tilted slightly toward the ipsilateral shoulder (Fig. 36.1). This head position places base of the temporal fossa perpendicular to the horizontal plane. The surface of the lateral temporal lobe is in a horizontal position, and long axis of the hippocampus is oriented vertically relative to the surgeon with this positioning. The purpose is having a head position with a good alignment of the mesial structures to the surgeon’s eyeline. This position provides an excellent exposure to the uncus–amygdala complex, the whole length of hippocampus, and the lateral–basal temporal neocortex.

Scalp Incision A smoothly curved, question mark–shaped scalp incision is drawn starting just above the zygoma and approximately 1 cm anterior to tragus, based on the location of palpated superficial temporal artery. Then, the incision extends upward such that it makes a smooth anterior turn at the upper point of the pinna by following the superior temporal line toward the keyhole. It

36  Anteromesial Temporal Lobectomy

Fig. 36.1  Head positioning of the patient. (a) The neck is extended by lowering the vertex approximately 15 degrees downward with a slight occipital tilt toward the ipsilateral shoulder and making the zygoma the most prominent point on the midline. (b) The head is turned to the contralateral side approximately 60 degrees. A question mark–shaped incision starts just above the zygoma and extends anteriorly toward the keyhole by ending just behind the hairline.

ends approximately 3 to 4 cm behind the keyhole, depending on the patient’s hairline (Fig. 36.1). Then the ­incision is infiltrated with 0.5% bupivacaine hydrochloride (Marcaine) diluted in 1:200,000 epinephrine solution. The superficial ­temporal artery is palpated and protected during the scalp incision. Some small branches of superficial temporal artery may be occasionally sacrificed, but main arterial branch can be protected by dissecting and mobilizing it. Then, the incision of the temporal fascia, muscle, and periosteum is also completed sharply by cutting these layers parallel to the scalp incision. Scalp, temporal fascia, muscle, and underlying periosteum are dissected subperiosteally to create a single musculocutaneous flap. The lower part of the incision is extended down to the zygoma. Having an exposure down to the zygomatic root is critical for satisfactory access to base of the temporal fossa during the neocortical resection. The other critical point at this stage is exposure of the orbital–zygomatic ridge or “keyhole.” Keyhole can be palpated, and the temporal muscle is cut and dissected from the keyhole by retracting the scalp further and working beneath it. Then, the temporal muscle is dissected subperiosteally using sharp periosteal elevators. The periosteum should be kept attached to the temporal muscle as much as possible to preserve muscle innervation and vascular supply. Monopolar cautery should not be used during this dissection for the same reason. Strict adherence to this technique is critical to prevent temporal muscle atrophy. Although application of this technique may be difficult in elderly patients, it is much easier to have an excellent subperiosteal dissection by keeping the periosteum intact and attached to temporal muscle in the pediatric age group. Then, fish hooks are used to reflect the musculocutaneous flap anterolaterally to expose the temporal bone widely.

Craniotomy Two burr holes are placed on the keyhole and just above the zygoma. A free bone flap is removed after dissecting the dura with Penfield’s dissectors. The sphenoid ridge is removed with rongeurs to create a smooth anterior–medial bony wall. This maneuver has critical significance to have a good exposure for uncus–amygdala resection. Further bone removal is needed along the floor of the temporal fossa down to the root of the zygoma and toward the temporal tip. This will provide a comfortable access to inferobasal neocortical region and temporal pole during the resection. Dural tack-up sutures are placed at this stage, and surrounding epidural space is filled with an injectable hemostatic agent, such as Surgifoam

Fig. 36.2  Exposed surgical field includes anterior part of the inferior frontal gyrus, sylvian vein, and superior and middle temporal gyri. The green line marks the surgical incision lines. First incision line (a–b) stays parallel to the sylvian fissure, and second incision line (b–c) stays perpendicular to the first incision line. The first incision line starts from the most anteromedial part of the temporal pole and extends posteriorly approximately 2 cm by following the sylvian vein and staying just a few millimeters below the vein. Then, the incision makes a smooth curve toward the superior temporal sulcus to preserve the superior temporal gyrus and follows the sulcus until the posterior resection line. The second incision line starts from the most posterior point of the first incision line and extends toward the floor of the temporal fossa by traversing the middle and inferior temporal gyri.

(Johnson & Johnson, Gateway, NJ). Then the dura is opened C-shaped, starting from the keyhole site on frontal region and ending at temporal pole by following the craniotomy edges. The dura is folded and tacked up with 4–0 N ­ urolon sutures to the muscle flap over the sphenoid wing. At this stage, the exposed area in the surgical field includes the full extent of the sylvian fissure or vein, superior and middle temporal gyri, and the inferior temporal gyrus (Fig. 36.2).

343

344

IVc  Temporal Lobe Epilepsy and Surgical Approaches

Neocortical Resection The previously planned resection length of the lateral temporal neocortex is measured and marked on the cortex as a first step. The tip of the temporal pole can be seen easily with the help of a cortical ribbon placed over the middle temporal gyrus. Previously planned resection length (≥ 3.5; depends on being on the dominant or nondominant side) from the tip of the temporal lobe is measured along the middle temporal gyrus and marked on the cortex. The resection line starts at the medial edge of the temporal pole and turns toward the middle temporal gyrus approximately 2 cm behind the temporal tip (Fig. 36.2). The remaining part of the incision continues along the upper border of the middle temporal gyrus to spare most of the superior temporal gyrus posteriorly. This resection line is marked on the pia–arachnoid of the superior and middle temporal gyri with a fine-tip bipolar coagulator staying parallel and 5 to 6 mm below the sylvian vein or superior temporal sulcus. After coagulation of the pia–arachnoid over the gyrus, it is incised with microscissors along the marked incision line. After completing the incision, the pia–arachnoid edge adjacent to the sylvian vein is coagulated thoroughly to create an appropriate handle to hold during the subpial dissection of the superior and middle temporal gyri. Then, cortex is subpially dissected from pia of the sylvian fissure anteriorly and from the superior temporal sulcus posteriorly. Meticulous subpial dissection technique is used to avoid injury to the middle cerebral artery (MCA) branches in the sylvian fissure (Fig. 36.3) and to protect the vascular supply of the unresected part of superior temporal gyrus by leaving both pial layers of the superior temporal sulcus undisrupted on the lower bank of the superior temporal gyrus. Some bleeding is generally encountered while peeling the cortex from pia that can be easily controlled by placing cottonoid patties. Subpial dissection is much more challenging in pediatric patients than adults because of the very thin and fragile nature of the pia at this age. Appropriate application of this technique may not be feasible in very young children. The next critical step is finding the temporal horn. There are several approaches for this and a close review of the patient’s MRI, especially coronal spoiled gradient recalled cuts, will be helpful to determine the best approach. The temporal horn starts approximately 3 cm behind the temporal tip, and the average distance between the surface of superior t­emporal

gyrus and the ventricle is approximately 31 to 34 mm.34,​35 We prefer to perform our dissection to reach the temporal horn at a point on the superior temporal sulcus approximately 3.5 cm behind the tip of the temporal pole. Frequently, the T1 sulcus (superior temporal sulcus) directly brings the surgeon into the temporal horn. This can be done through an intrasulcal approach or by remaining subpial and following either the inferior wall of the superior temporal gyrus or superior wall of the middle temporal gyrus, which we prefer. Bottom of the sulcus can be easily recognized by visualizing the end of the pial bank. Then, the ependyma can be appreciated after deepening the same i­ncision a ­ pproximately 10 mm further.35 Actual distance can be measured case by case on patient’s MRI coronal cuts easily. The ependyma can be opened with Penfield #4 dissector (­ Codman, MA) and cerebrospinal fluid will verify the intraventricular location. If the surgeon passes the estimated distance and the temporal horn is not in sight, then the best strategy is to redirect the dissection. The most common two reasons for not being able to find the ventricle are either placing the entry point of the dissection too anteriorly or directing the dissection either too medially or too laterally. At this stage, the appropriate strategy is to redirect the dissection toward the floor of the middle fossa but not medially. The dissection is then deepened toward the floor of the middle fossa until gray matter is encountered on the adjacent occipitotemporal (or fusiform) gyrus. Then, the dissection is redirected again, this time medially into the white matter until temporal horn is entered. Deepening the dissection medially to search the temporal horn without taking the aforementioned strategies may easily lead the surgeon into the temporal stem and basal ganglia and may cause significant complications. Therefore, redirecting the dissection intentionally too laterally first is a much safer approach, as defined very clearly by Wen et al.34 When we enter into the ventricle, we place a cottonoid micropatty in it to prevent blood contamination and then subpially dissect first the superior wall of the medial temporal gyrus and then the sylvian sulcus pia anteriorly to the temporal pole using microsuction in a low setting and a Penfield dissector. This subpial dissection is performed down to the ependymal level throughout the sulcus. Then, the ependyma is opened using a bipolar coagulator, and the temporal horn is unroofed all the way to its tip, and a small cotton ball is placed into the temporal horn toward the atrium to avoid intraventricular dissemination of blood products.

Fig. 36.3  (a) Anterior temporal neocortex is subpially dissected from the sylvian fissure by keeping the pia intact to avoid any risk of injury to the middle cerebral artery branches. (b) Anterior temporal neocortex is removed en bloc by exposing the tentorium and mesial temporal structures.

36  Anteromesial Temporal Lobectomy Several other approaches to the temporal horn exist. One is to follow the collateral sulcus. This approach is only feasible after completing the second cortical incision, which will be described in the following paragraph. Alternatively, the temporal horn can be found after completing the resection of the anterolateral temporal lobe without locating the temporal horn. In this case, the uncus is located first by following the tentorial edge anteromedially. When removal of the uncus is completed, its posterior segment makes the anterior wall of the temporal horn, and removal of this part of the uncus will expose the tip of the temporal horn spontaneously. Lastly, the use of a neuronavigation system to assist the localization of the temporal horn is an option. However, neuronavigation may not always be reliable because of brain shift at this stage. The second cortical incision line starts from the most posterior end of the first incision, and it is directed perpendicularly toward the floor of temporal fossa (Fig. 36.2). The posterior line of the neocortical resection extends inferiorly traversing the superior, middle, inferior temporal, and fusiform gyri, respectively, and ends at the collateral sulcus. The temporal horn is located generally just dorsal to the base of the collateral sulcus and can be found by following the collateral sulcus pia as described previously. The average distance from the depth of the collateral sulcus to the temporal horn is 3 to 6 mm.35 Thus, the posterior end of the first incision and superior end of the second incision lines intersect at the temporal horn. A third incision is directed to the collateral sulcus by cutting across the temporal stem and the white matter of the basal temporal lobe. This third incision disconnects the temporal neocortex from parahippocampus–hippocampus complex and completes the lateral neocortical temporal resection by dividing the collateral sulcus from its posterior end to the tip of the temporal horn at rhinal sulcus level. Then, the entire lateral neocortex is removed as an en bloc specimen (Fig. 36.3).

Mesial Temporal Resection For the next step, it is important to locate several anatomical landmarks and structures before proceeding to resect the mesial temporal structures. Hippocampus, fimbria, lateral ventricular sulcus, collateral eminence, choroid plexus, choroidal fissure, inferior choroidal point, and amygdala need to be fully exposed and can be distinctly recognized at this stage. The hippocampus sits over the subiculum of parahippocampal gyrus. It has a short, wide head that continues with a gradually narrowing body and tail. The tail makes a backward–upward turn at the trigone level around the posterior cerebral peduncle. The anterior portion of the hippocampal head blends into the posterior uncus and amygdala (Fig. 36.4). The hippocampus can be easily recognized between the collateral eminence and choroidal fissure. The lateral ventricular sulcus lies between the hippocampus proper and the collateral eminence and extends anteriorly toward the amygdala–hippocampal junction. The body of the hippocampus is lined by choroid plexus, and the choroidal point is seen at the most anterior part of the choroidal fissure. If the choroid plexus is lifted gently upward and medially, the choroidal fissure and fimbria would be fully exposed (Fig. 36.4). If we retract the choroid plexus laterally over the hippocampus, then stria terminalis is exposed fully. When the anterior end of the choroid plexus is pulled backward, the velum terminale and the choroidal point at the tip of the ­posterior uncus can be visualized (Fig. 36.5).

The anterior choroidal artery (AChA) runs across the ambient and crural cisterns near the choroid plexus. It pierces the arachnoid plane to supply the choroid plexus at the inferior choroidal point by giving rise to numerous branches. The anterior fimbria and stria terminalis join to form the velum terminale and create the anterior border of the choroidal fissure where the inferior choroidal point is also located (Fig. 36.5). The fimbria is a narrow, flat band covering the mesial border of the ­hippocampus. It is located just above the dentate gyrus and continues as fimbria fornix posteriorly. The temporal horn is fully unroofed to expose the most anterior part of the temporal horn that includes bulging amygdala, posterior uncus, amygdala–hippocampal junction, and posteriorly head and body of the hippocampus. The uncal recess is a distinct landmark that separates the head of the hippocampus from the amygdala. Better exposure of the hippocampal tail can be provided with the help of a tapering ribbon retractor (Fig. 36.6a). The ribbon is placed on the most posterior end of the unroofed part of the temporal horn, and the remaining part of the roof is gently elevated laterally for this purpose. The hippocampal tail can be exposed with this maneuver back to the point where it makes a medial and upward turn. Obtaining this exposure is very critical for a satisfactory resection of mesial temporal structures. After locating the intraventricular landmarks, resection of mesial temporal structures starts with an incision on the lateral ventricular sulcus that is the demarcation line between the collateral eminence and hippocampus. The ependyma of the lateral ventricular sulcus is coagulated posteroanteriorly as an entry point to the parahippocampal gyrus. The medial pial bank of the collateral sulcus is exposed by suctioning parahippocampus intragyrally. Intragyral removal of the parahippocampus is completed along the collateral eminence, starting from hippocampus proper to the amygdala–hippocampal junction. Then, the lateral wall of the parahippocampus–hippocampus complex

Fig. 36.4  Choroidal point (asterisk) and anterior part of the choroidal fissure is exposed by peeling the fimbria. Note surrounding structures including choroidal plexus (a), fimbria (b), hippocampus (c), and posteromedial part of uncus (d).

345

346

IVc  Temporal Lobe Epilepsy and Surgical Approaches

Fig. 36.5  (a) Choroidal point (asterisk) is seen surrounded by anterior tip of choroidal plexus (a), fimbria (b), velum terminale (c), stria terminalis (d), head of hippocampus (e). (b) The anterior part of the fimbria and stria terminalis joins to form velum terminale (asterisk).

Fig. 36.6  (a) Head and body of the hippocampus are exposed, and a retractor (asterisk) is placed to elevate the temporal roof for further exposure of the hippocampal tail. Note choroidal sulcus (a) and surgical resection line on collateral eminence (b). (b) Entire hippocampus is subpially dissected as an en bloc specimen between collateral eminence (a) and fimbria (b).

is subpially dissected by peeling it from the collateral sulcus pia using the Penfield dissectors (Fig. 36.6b). Then, the dissection continues mesially toward the tentorial edge until the pia along the mesial border of the parahippocampus and hippocampal sulcus is encountered. At this stage, the subiculum of the hippocampus is peeled off toward the hippocampal sulcus. The parahippocampal gyrus lateral to this line is emptied further by

suctioning it anteriorly toward the entorhinal area and uncus. At this stage, the hippocampus proper can easily be retracted laterally into the cavity created by intragyral aspiration of the parahippocampus. This maneuver provides an excellent view of the anterior end of the hippocampal sulcus. The hippocampal sulcus fans out at this junction between pes hippocampi, uncus and anterior end of the parahippocampus (Fig. 36.5a). This anatomy provides the surgeon with an excellent starting point for the dissection of hippocampal sulcus between fimbria, inferior choroidal point, and choroidal fissure. The most anterior end of the fimbria can be easily opened and peeled away from the pia of the choroidal fissure just lateral to tela choroidea. Then, the fimbria can be lifted with Rhoton microdissectors (Codman, MA) at the inferior choroidal point level. Then, underlying pia and vasculature can be exposed (Fig. 36.4). The fimbria is further opened with R ­ hoton dissectors along its length all the way to the hippocampal tail. At this stage, the hippocampus is further retracted laterally with the suction, and the hippocampal sulcus is exposed as a two-layered pial folding with several tiny arteries running between pial layers. The hippocampal sulcus is a very critical landmark in this procedure and should be fully visualized. It separates the hippocampus proper and the subiculum. The subiculum constitutes the most medial part of the parahippocampus bulging into the middle incisural space. The hippocampal arteries and arising arterioles (Uchimura’s arteries) are located within the hippocampal sulcus (Fig. 36.7). These thin hippocampal arteries mostly form a group of two to six thin vessels from the AChA and medial P2 segment of posterior cerebral artery (PCA) close to the free edge of the tentorium. After having a satisfactory exposure of the hippocampal sulcus, hippocampal arterioles are coagulated with fine-tipped bipolar forceps and cut with microscissors one by one (Fig. 36.7c). Again, it should be noted that these arteries in young pediatric patients are extremely thin and can rupture easily with manipulation. Further, the distances between the hippocampal arteries and the AChA and P2 segment are very short in pediatric patients. Therefore, coagulation of ­hippocampal arteries should be performed carefully using very fine-tipped bipolars by staying close to hippocampus

36  Anteromesial Temporal Lobectomy

Fig. 36.7  (a) Fimbria is lifted with a dissector to expose the hippocampal sulcus and hippocampal arteries. (b) Further dissection and elevation of fimbria (a) expose subiculum (b) and hippocampal arteries extending into hippocampal sulcus. (c) Subiculum and hippocampal sulcus are fully exposed, and hippocampal arteries have been coagulated.

proper. Then, the head of the hippocampus is fully dissected subpially from the underlying pia and lifted upward and posteriorly. This maneuver provides a very nice subpial plane at the base of the whole hippocampus–parahippocampus complex. Then, the hippocampal head is mobilized and lifted upward and posteriorly. The remaining parahippocampal attachments are dissected subpially using a Penfield number 4 dissector. This way the whole hippocampus and underlying part of the parahippocampus are dissected and mobilized all the way back to the hippocampal tail. Then, the tail is resected with bipolar coagulation at its upward turn behind the quadrigeminal plate, and the hippocampus is removed en bloc (Fig. 36.8). The final step of the procedure is the resection of the amygdala while emptying the content of the anterior uncus. During this stage of the procedure, using strictly subpial dissection and showing the utmost respect to pial barriers are critical to protect the underlying vasculature, third nerve, and cerebral peduncle. The anterior amygdala blends into the uncus, and we use a microsuction with the suction regulator at the low-suction setting and Penfield’s dissectors to peel the uncal content from the pia below the incisura (Fig. 36.9a). Ultrasonic aspirator in a low setting is also a very useful tool to empty the uncal content. After completing the resection of uncus and anterior basal amygdala, the cerebral peduncle and third nerve can be seen under the intact pia (Fig. 36.9b). Although

Fig. 36.8  The hippocampus proper is removed en bloc for histological examination. Further resection of the hippocampal tail is performed with the ultrasonic aspirator.

the anterior and basal borders of the amygdala are very well defined, there are no dorsomedial anatomical boundaries of the amygdala. Amygdala merges with striatum superiorly. Therefore, it is more challenging to define the dorsomedial resection borders of the amygdala.

347

348

IVc  Temporal Lobe Epilepsy and Surgical Approaches

Fig. 36.9  (a) Uncus and amygdala are peeled from the pia are with dissectors or low-power suction. (b) Uncus is emptied subpially. Third nerve (a) under the pia is visible along with some residual part of amygdala (b). (c) After subpial removal of amygdala and uncus, edge of tentorium (a), third nerve (b), and posterior cerebral artery (c) are visible under intact pia.

The M1 segment of the MCA can be seen subpially. A line extending from inferior choroidal point at the anterior tip of the temporal horn to the angle of the MCA at the limen insula makes the anterior–superior border of the resection line of amygdala. Inferior choroidal point is located just behind the uncus and at the inferior end of the choroidal plexus and fissure. It corresponds to the entry point of anterior choroidal artery to the temporal horn.36 The dissection at the a ­ nterior– superior border should be done very carefully because of the presence of the small MCA branches supplying the basal ganglia. After completing the amygdala resection, the s­ urgical c­ avity is

reexplored and all devascularized residual cortical tissues are removed with the ultrasonic aspirator without violating the pia. At this stage, the tentorial edge, third nerve, internal cerebral artery, PCA, lateral edge of the midbrain between cerebral peduncle, and tectum can be seen under the pia in the ambient and crural cisterns. After hemostasis, the surgical cavity is filled with warm saline irrigation, and the dura is closed in a watertight fashion with 4–0 Nurolon sutures. The bone flap is replaced with microplates, and the temporal muscle, fascia, and galea are closed as two separate layers using 3–0 and 4–0 Vicryl sutures. The skin is closed with 4–0 sutures or subcuticular fashion.

36  Anteromesial Temporal Lobectomy

„„ Complications The complication rates in temporal lobe surgery series in children have been reported to be between 2 and 8%.2,7,9 Mortality is a rare occurrence and is reported as lower than 0.5%.37 Postoperative complications, although rare, may be devastating, and using appropriate surgical techniques and extreme caution at critical stages of the surgery is essential to avoid complications. The most commonly reported complications are visual field defects, infection, stroke, manipulation or retraction hemiparesis, third nerve palsy, and language disturbances. Lopez-Gonzalez et al reported overall complication rate as 7%.38 The most common type of visual field defect is superior quadrantanopsia, with a reported incidence of 9% in a Benifla et al’s series of 126 children with TLE.7 Benifla et al also reported a 4% incidence of homonymous hemianopia. The overall complication rate was 14.9% in Clusmann and ­colleagues’ series, but only 2.2% had permanent deficits.8 Incomplete or complete quadrantanopsia incidences were 28.2 and 3.8%, respectively. In a separate study by Kim et al, the rate of postoperative visual field defect was 22%.3 Another common complication is dysphasia, which is mostly transient. Transient dysphasia can be seen in approximately half of the dominant site temporal resections and frequently resolves within a few weeks.39 Possible reason for this finding is the disconnection of the mesial and neocortical temporal lobe and retraction related to physiological disruption. Although rare, third and fourth nerve palsies can also be seen after AMTL. Using strict subpial technique and avoiding cautery around the tentorium or high-power suction application during the uncus resection may help to avoid these complications. Partial seventh nerve palsy is another well-known complication and occurs secondary to injury of the facial nerve branches located within the temporalis fascia. This injury can be easily avoided with the technique we described here by avoiding dissection of the temporalis fascia. However, traction and monopolar cauterization in close proximity to the facial nerve may also cause facial palsy and should be taken into consideration during the craniotomy. One of the most devastating, although rare, complications in temporal lobe resection is hemiplegia. It is a well-recognized but rare complication.5 It has been termed manipulation hemiplegia and frequently is related to injuries of the AChA and PCA during the resection of the mesial temporal structures. Maintaining the utmost respect to subpial

technique, meticulous protection of the pia throughout the surgery, coagulation and cutting of the hippocampal arterioles strictly in hippocampal sulcus, and staying away from the main arteries (AChA or PCA) decrease the risk of injury to these vessels. MCA-related hemiplegia secondary to compression with retractors may cause this problem as well.

„„ Outcome The seizure control rate of temporal resections in children is different than adults; the main reason is the heterogeneity of underlying pathologies. The most common neuropathological substrates in children are cortical dysplasia and neoplasms followed by gliosis and MTS.1,​7,​40 Temporal lobe resection is a safe and effective surgical technique in the management of TLE with reported seizure control rates between 60 and 80%.1,​2,​7,​8,​9,​10,​11,​37,​41,​ 42,​43 Seizure-free outcome rate was reported as 78% by Sinclair et al9 in their series of 42 patients. Benifla et al7 reported 74%, and Clusmann et al8 reported 87% good seizure control rates (Engel classes 1 and 2) with temporal lobe resection in 126 and 89 children, respectively. The best outcome was seen in the patients with temporal lobe neoplasms (88–92%) followed by the patients with gliosis (86%) and MTS (70%) in Benifla et al’s series.7 The lowest seizure control rate was seen in the patients with cortical dysplasia. Mittal et al reviewed their experience with 109 children at the MNI and reported Engel classes 1 and 2 outcomes in 86.3% of patients at more than 5 years of follow-up.44 Jarrar et al found that the seizure-free rate in their series was 82% 5 years after surgery but decreased to 53% after 10 years.45 Maton et al reported their experience with temporal lobe resection during early life in 20 children younger than 5 years old.41 Sixty-five percent of the children were seizure free, and an additional 15% had more than 90% seizure reduction at a mean follow-up of 5.5 years. Smyth et al46 reported 63.3% overall good seizure control rate (Engel classes 1 and 2) in the preadolescent age group. MTS patients had a 76.9% seizure control rate, which compared favorably to cortical dysplasia and gliosis groups in this study. The Great Ormond Street Hospital series reported seizure-free rates as 73% in lesional, 58% in MTS, and 33% in dual pathology groups.5 Kim et al reported 88% seizure-free outcome in the temporal resection group of their epilepsy surgery series in children.3 Cleveland Clinic series reports seizure freedom rate as 76% in 1st year, 54% in 5th year, and 41% in 12th year.38 In a meta-analysis that included 36 studies covering 1,318 pediatric patients, seizure freedom rate was found to be 76%.47

References 1. Wyllie E, Comair YG, Kotagal P, Bulacio J, Bingaman W, Ruggieri P. Seizure outcome after epilepsy surgery in children and adolescents. Ann Neurol 1998;44(5):740–748

3. Kim SK, Wang KC, Hwang YS, et al. Epilepsy surgery in children: outcomes and complications. J Neurosurg Pediatr 2008;1(4):277–283

2. Adelson PD, Peacock WJ, Chugani HT, et al. Temporal and extended temporal resections for the treatment of intractable seizures in early childhood. Pediatr Neurosurg 1992;18(4): 169–178

4. Cossu M, Lo Russo G, Francione S, et al. Epilepsy surgery in children: results and predictors of outcome on seizures. Epilepsia 2008;49(1):65–72 5. Harkness W. Temporal lobe resections. Childs Nerv Syst 2006;22(8):936–944

349

350

IVc  Temporal Lobe Epilepsy and Surgical Approaches 6. Miserocchi A, Cascardo B, Piroddi C, et al. Surgery for temporal lobe epilepsy in children: relevance of presurgical evaluation and analysis of outcome. J Neurosurg Pediatr 2013;11(3):256–267 7. Benifla M, Otsubo H, Ochi A, et al. Temporal lobe surgery for intractable epilepsy in children: an analysis of outcomes in 126 children. Neurosurgery 2006;59(6):1203–1213, discussion 1213–1214 8. Clusmann H, Kral T, Gleissner U, et al. Analysis of different types of resection for pediatric patients with temporal lobe epilepsy. Neurosurgery 2004;54(4):847–859, discussion 859–860 9. Sinclair DB, Aronyk K, Snyder T, et al. Pediatric temporal lobectomy for epilepsy. Pediatr Neurosurg 2003;38(4):195–205 10. Duchowny M, Levin B, Jayakar P, et al. Temporal lobectomy in early childhood. Epilepsia 1992;33(2):298–303 11. Mohamed A, Wyllie E, Ruggieri P, et al. Temporal lobe epilepsy due to hippocampal sclerosis in pediatric candidates for epilepsy surgery. Neurology 2001;56(12):1643–1649 12. Schramm J. Temporal lobe epilepsy surgery and the quest for optimal extent of resection: a review. Epilepsia 2008;49(8):1296–1307 13. de Almeida AN, Teixeira MJ, Feindel WH. From lateral to mesial: the quest for a surgical cure for temporal lobe epilepsy. Epilepsia 2008;49(1):98–107 14. Feindel W, Leblanc R, de Almeida AN. Epilepsy surgery: historical highlights 1909–2009. Epilepsia 2009;50(Suppl 3):131–151 15. Penfield W, Flanigin H. Surgical therapy of temporal lobe seizures. AMA Arch Neurol Psychiatry 1950;64(4):491–500 16. Penfield W, Baldwin M. Temporal lobe seizures and the technic of subtotal temporal lobectomy. Ann Surg 1952;136(4):625–634 17. Penfield W, Jasper H. Epilepsy and Functional Anatomy of the Human Brain. Boston, MA: Little Brown; 1954:815–816 18. Penfield W, Milner B. Memory deficit produced by bilateral lesions in the hippocampal zone. AMA Arch Neurol Psychiatry 1958;79(5):475–497 19. Scoville WB, Milner B. Loss of recent memory after bilateral hippocampal lesions. J Neurol Neurosurg Psychiatry 1957;20(1):11–21 20. Niemeyer P. The transventricular amygdalo-hippocampectomy in temporal lobe epilepsy. In: Baldwin M, Bailey P, eds. Temporal Lobe Epilepsy. Springfield, IL: CC Thomas; 1958:461–482 21. Yaşargil MG, Teddy PJ, Roth P. Selective amygdalohippocampectomy: operative anatomy and surgical technique. In: Symon L, et al., eds. Advances and Technical Standards in Neurosurgery. Vol. 12. New York, NY: Springler-Wien; 1985:93–123 22. Rasmussen T, Jasper H. Temporal lobe epilepsy: indication for operation and surgical technique. In: Baldwin M, Bailey P, eds. Temporal Lobe Epilepsy. Springfield, IL: CC Thomas; 1958:440–460 23. Rasmussen T, Branch C. Temporal lobe epilepsy; indications for and results of surgical therapy. Postgrad Med 1962;31:9–14 24. Rasmussen T. Surgical treatment of patients with complex partial seizures. In: Penry JK, Daly DD, eds. Advances in Neurology. Vol. 11. Complex Partial Seizures and Their Treatment. New York, NY: Raven Press; 1975:415–449 25. Rasmussen T. Surgical aspects of temporal lobe epilepsy: results and problems. In: Gillingham J, Gybels J, Hitchcock ER, Szikla G, eds. Advances in Stereotactic and Functional Neurosurgery Acta Neurochirurgica, Supp 30. Vienna: Springer-Verlag; 1980:13–24 26. Rasmussen TB. Surgical treatment of complex partial seizures: results, lessons, and problems. Epilepsia 1983;24 (Suppl 1):S65–S76 27. Feindel W, Penfield W, Jasper H. Localization of epileptic discharges in temporal lobe automatism. In: Transactions of the American Neurological Association. New York, NY: Springer; 1952:14–17

28. Feindel W, Penfield W. Localization of discharge in temporal lobe automatism. AMA Arch Neurol Psychiatry 1954;72(5):603–630 29. Feindel W, Rasmussen T. Temporal lobectomy with amygdalectomy and minimal hippocampal resection: review of 100 cases. Can J Neurol Sci 1991;18(4, Suppl):603–605 30. Goldring S, Edwards I, Harding GW, Bernardo KL. Results of anterior temporal lobectomy that spares the amygdala in patients with complex partial seizures. J Neurosurg 1992;77(2):185–193 31. Olivier A. Transcortical selective amygdalohippocampectomy in temporal lobe epilepsy. Can J Neurol Sci 2000;27(Suppl 1): S68–S76, discussion S92–S96 32. Spencer DD, Spencer SS, Mattson RH, Williamson PD, Novelly RA. Access to the posterior medial temporal lobe structures in the surgical treatment of temporal lobe epilepsy. Neurosurgery 1984;15(5):667–671 33. James JS, Radhakrishnan A, Thomas B, et al. Diffusion tensor imaging tractography of Meyer’s loop in planning resective surgery for drug-resistant temporal lobe epilepsy. Epilepsy Res 2015;110:95–104 34. Wen HT, Rhoton AL Jr, Marino R Jr. Gray matter overlying anterior basal temporal sulci as an intraoperative landmark for locating the temporal horn in amygdalohippocampectomies. Neurosurgery 2006;59(4, Suppl 2):ONS221–ONS227, discussion ONS227 35. Campero A, Tróccoli G, Martins C, Fernandez-Miranda JC, Yasuda A, Rhoton AL Jr. Microsurgical approaches to the medial temporal region: an anatomical study. Neurosurgery 2006;59(4, Suppl 2):ONS279–ONS307, discussion ONS307–ONS308 36. Tubbs RS, Miller JH, Cohen-Gadol AA, Spencer DD. Intraoperative anatomic landmarks for resection of the amygdala during medial temporal lobe surgery. Neurosurgery 2010;66(5):974–977 37. Adelson PD. Temporal lobectomy in children with intractable seizures. Pediatr Neurosurg 2001;34(5):268–277 38. Lopez-Gonzalez MA, Gonzalez-Martinez JA, Jehi L, Kotagal P, Warbel A, Bingaman W. Epilepsy surgery of the temporal lobe in pediatric population: a retrospective analysis. Neurosurgery 2012;70(3):684–692 39. Kraemer DL, Spencer DD. Temporal lobectomy under general anesthesia. Tech Neurosurg 1995;1:32–39 40. Spencer S, Huh L. Outcomes of epilepsy surgery in adults and children. Lancet Neurol 2008;7(6):525–537 41. Maton B, Jayakar P, Resnick T, Morrison G, Ragheb J, Duchowny M. Surgery for medically intractable temporal lobe epilepsy during early life. Epilepsia 2008;49(1):80–87 42. Arruda F, Cendes F, Andermann F, et al. Mesial atrophy and outcome after amygdalohippocampectomy or temporal lobe removal. Ann Neurol 1996;40(3):446–450 43. Clusmann H, Schramm J, Kral T, et al. Prognostic factors and outcome after different types of resection for temporal lobe epilepsy. J Neurosurg 2002;97(5):1131–1141 44. Mittal S, Montes JL, Farmer JP, et al. Long-term outcome after surgical treatment of temporal lobe epilepsy in children. J Neurosurg 2005;103(5, Suppl):401–412 45. Jarrar RG, Buchhalter JR, Meyer FB, Sharbrough FW, Laws E. Long-term follow-up of temporal lobectomy in children. Neurology 2002;59(10):1635–1637 46. Smyth MD, Limbrick DD Jr, Ojemann JG, et al. Outcome following surgery for temporal lobe epilepsy with hippo­ campal involvement in preadolescent children: emphasis on mesial temporal sclerosis. J Neurosurg 2007;106(3, Suppl): 205–210 47. Englot DJ, Rolston JD, Wang DD, Sun PP, Chang EF, Auguste KI. Seizure outcomes after temporal lobectomy in pediatric patients. J Neurosurg Pediatr 2013;12(2):134–141

37

  Selective Amygdalohippocampectomy Uğur Türe, Ahmet Hilmi Kaya, Berrin Aktekin, and Canan Aykut Bingöl

Summary Hippocampal sclerosis is one of the leading causes of intractable epilepsy and especially affects younger population. Selective amygdalohippocampectomy (SAH) results with a high ratio of seizure control in these patients. The aim is to resect the amygdala, hippocampus, and accompanying parahippocampal gyrus with least disturbance to neighboring structures such as temporal neocortex and optic radiation. Various SAH techniques have been described previously. Previously, our preferred technique was the pterional transsylvian–transamygdalar SAH technique in our epilepsy unit. Recently, the senior author (UT) developed the paramedian supracerebellar transtentorial (PST) SAH, which offers a more complete resection of the hippocampus and amygdala and is now the most frequently used technique in our clinic. In this chapter, we discuss the two techniques of SAH in detail. Keywords:  amygdalohippocampectomy, epilepsy, hippocampus, mediobasal temporal region, paramedian supracerebellar– transtentorial approach, pterional transsylvian–­transamygdalar approach

„„ Introduction The aim of SAH is the selective resection of the amygdala, the hippocampus, and the parahippocampal gyrus. A profound knowledge of the vascular supply of this area and its possible variations and a full understanding of the surgical anatomy of the limbic system are the “sine qua non” of this surgical concept. Recent developments in neuroimaging, especially MRI with high resolution, have enabled clear visualization of abnormalities in the mediobasal temporal region (MTR).1 In turn, this development has facilitated surgical decision making. The availability of fiber tractography, especially with 3-T MRI, and an interest in the white matter anatomy, specifically information gained from fiber dissection, have also contributed to develop more efficient surgical techniques.1,​2,​3 The efficacy of surgery in the management of drug-resistant temporal lobe epilepsy has been demonstrated in a prospective randomized trial.4 However, controversy remains regarding which resection method yields the best results for seizure control and better neuropsychological outcome. The temporal neocortical resection leaving the hippocampus or amygdala behind can result in seizure-free rates of about 50%.5 On the

other hand, anteromesial temporal lobectomy (AMTL) provides better seizure control rates. The seizure control rates with SAH is also similar to AMTL, while showing considerable evidence of the neuropsychological outcome being better in patients undergoing SAH.4,​5,​6,​7,​8 Although class I evidence for seizure outcome based on type and extent of resection of MTR structures is rare, SAH appears to provide a similar seizure control rate and a better cognitive outcome than temporal lobe resection based on available data.5,​8 Still, it remains unclear whether larger mediobasal resection leads to a better seizure outcome. In children, seizure outcome and functional recovery are better.6 Various SAH techniques were described previously.9​–​18 However, we preferred using the pterional transsylvian SAH, described by Yaşargil, in our epilepsy unit.17,​18 Recently, the senior author (UT) developed the paramedian supracerebellar transtentorial selective amygdalohippocampectomy (PST-SAH),9,​16 which is now our most preferred technique. In this chapter, we will describe our experience with these two techniques.

„„ Patient Selection and Preoperative Evaluation Selecting the right candidate for SAH is important in terms of obtaining expected goals for cognitive outcome and alleviation of seizures. Patients suffering from mediobasal temporal epilepsy (MTE), with hippocampal sclerosis and intractable seizures need to be clearly identified by defined underlying hippocampal pathology shown on MRI, ictal and interictal electroencephalography (EEG) findings on long-term video-EEG monitoring as well as positron emission tomography (PET) scans, neuropsychological evaluation, and clinical semiology. Initial precipitating incidents, including febrile seizures, trauma, hypoxia, and intracranial infections before the age of 5 and prior to the onset of habitual nonfebrile seizures are very common in patients with MTE.19 Habitual seizures begin earlier in MTE with hippocampal sclerosis, the majority occurring in patients between the ages of 4 and 16 years; however, these seizures can begin earlier or much later with the patient still showing the same pathological changes and excellent response to surgery. Focal seizures occur in more than 90% of patients, but secondary generalized seizures are rare and may correlate with the extent of the lesion. Auras and automotor seizures, sometimes with impaired consciousness, are characteristics of MTE with hippocampal sclerosis. Auras are mainly charac-

352

IVc  Temporal Lobe Epilepsy and Surgical Approaches terized by ascending epigastric sensation, whereas gradual impairment of consciousness is typically associated with oroalimentary automatism in about 70% of patients.19 Dystonic posturing occurs in 20 to 30% of patients and is contralateral to the side of seizure onset. Specific baseline and follow-up neuropsychological testing are important. MRI is the most important investigational tool. Improvements in MRI techniques, especially 3-T MRI scanners, contribute to diagnosis of hippocampal sclerosis significantly. MRI volumetric investigations and spectroscopy give more information on T1, T2, and fluid-attenuation inversion recovery (FLAIR) findings. The [18F] fluorodeoxyglucose (FDG)-PET frequently demonstrates ipsilateral hypometabolism interictally. Ictal and postictal single photon emission computed tomography (SPECT) are preferred over interictal SPECT. In one-third of patients, interictal epileptiform anomalies are lateralized and localized to the lesion. In the other two-thirds, bilateral d ­ ependent or independent (or both) epileptic activity

Fig. 37.1  Medial surface of the left temporal operculum and mediobasal temporal region (MTR) in cadaver brain, superomedial view. Dotted line indicates the incision through the collateral eminence to the collateral sulcus, and the incision through the posterior limit of the hippocampus and parahippocampal gyrus via pterional transsylvian–transamygdalar approach. Arrow indicates the angle of the surgical approach. ahg, anterior Heschl’s gyrus; fi, fimbria; h, hippocampus; ips, inferior peri-insular sulcus; li, limen insula; pc, piriform cortex; phg, posterior Heschl’s gyrus; ppl, polar planum; s, subiculum; scc, splenium of corpus callosum; tpl, temporal planum; u, uncus. The white letters denote sulci and fissures.

is detected. Sphenoid electrode recordings also disclose more information about lateralization. The ictal onset is not always detected by scalp video-EEG recording, and seizures are lateralized in 80% of patients.19 Invasive EEG recordings with depth and s­ ubdural electrodes are needed in patients with discordant findings on MRI, semiology, functional imaging examinations, and ­electrophysiology.

„„ Surgical Techniques Pterional Transsylvian–Transamygdalar Selective Amygdalohippocampectomy Exact knowledge of the topographical, white matter, and vascular anatomy of the region is crucial for a successful outcome after SAH (Fig. 37.1, Fig. 37.2, and Fig. 37.3).16,​20​–​34 A pterional craniotomy is performed in usual way as described previously.18,​33,​35 The dura is opened in a semicircular fashion

Fig. 37.2  Coronal section of the left-sided cadaver cerebral hemisphere through the amygdala, anterior view. The hippocampus (h) demonstrated at the tip of the temporal horn the amygdala (a) located at the anterosuperomedial side of the hippocampus. Arrow indicates the angle of the surgical approach. ac, anterior commissure; cc, corpus callosum; cn, caudate nucleus; cp, cerebral peduncle; cs, collateral sulcus; fg, fusiform gyrus; gp, globus pallidus; I, insula; ic, internal capsule; ips, inferior peri-insular sulcus; ot, optic tract; p, putamen; pg, parahippocampal gyrus; T1, superior temporal gyrus; T2, middle temporal gyrus; T3, inferior temporal gyrus; t1, superior temporal sulcus; t2, inferior temporal sulcus; ts, temporal stem. The white letters denote sulci and fissures.

37  Selective Amygdalohippocampectomy

Fig. 37.3  (a) Superolateral view of the left MTR and internal capsule following using fiber dissection technique in cadaver brain. Arrow indicates the angle of the surgical approach for transsylvian–transamygdalar subarachnoid hemorrhage (SAH). (b) Superior view of the left MTR with arterial vascularization in cadaver brain. Asterisk indicates the anterior choroidal artery. A1, first segment of anterior cerebral artery; a, amygdala; ac, anterior commissure; af, anterior fossa; alic, anterior limb of internal capsule; alv, atrial portion of lateral ventricle; ap, ansa peduncularis; ce, collateral eminence; cc, corpus callosum; cp, cerebral peduncle; cs, collateral sulcus; fi, fimbria; h, hippocampus; ha, hippocampal arteries; ica, internal carotid artery; M1, first segment of middle cerebral artery; on, optic nerve; ot, optic tract; P2, second segment of posterior cerebral artery; pc, piriform cortex; pg, parahippocampal gyrus; plic, posterior limb of internal capsule; s, subiculum; slic, sublentiform portion of internal capsule; tp, temporal pole; u, uncus.

above the sylvian fissure, with the incision arching toward the sphenoid ridge and orbit. Chiasmatic and carotid cisterns may be explored through the fronto-orbital aspect of the frontal lobe before opening the sylvian fissure. The arachnoid between the optic nerve and the internal carotid artery (ICA) must then be opened. These procedures release large amounts of cerebrospinal fluid (CSF), which relaxes the brain and facilitates further dissection. Next, the proximal part of the sylvian fissure is opened frontal or temporal side to the superficial sylvian veins, depending on the variations in the venous anatomy. A simple spreading action with fine bipolar forceps is usually adequate to dissect the fissure. As the sylvian fissure is dissected more deeply, longer fine-tipped forceps are needed. The thickened arachnoid bands must be divided with microscissors where necessary. Dissection continues to expose middle cerebral artery (MCA) bifurcation, then following M1 segment of MCA, leading to the ICA bifurcation, where the arachnoid fibers between the temporal and fronto-orbital areas are well separated, as are the vessels along the proximal sylvian fissure down to the ICA and its branches. These structures, as well as the position of the oculomotor nerve, the tentorial edge, and uncus may be inspected. The lateral branches of the ICA (posterior communicating artery [PCoA], anterior choroidal artery [AChA], and striatocapsular arteries) and the cortical branches of the M1 segment (temporopolar, anterior, and middle temporal arteries) and its variations are identified, as are the position, variation, and courses of the lenticulostriate arteries.33 The limen insula and the inferior trunk of the M2 segment are observed. The M2 segment curves slightly laterally in the inferior peri-insular sulcus and lies just over the inferior insular vein. The surgeon must find sufficient space to make an incision in the piriform cortex between the temporal arteries, by mobilizing them if needed. Significant variations exist among patients regarding the major vascular supply to the amygdala, uncus, hippocampus, and parahippocampal gyrus. The resection of the piriform cortex just anterolateral to the M1 segment and anteroinferior to the limen insula enables the surgeon to reach the amygdala. The superior part of the

­ mygdala is identified a few millimeters under the incision a line by its hazelnut color in the white matter. The amygdala must first be removed piecemeal with both a tumor forceps (to gain histological specimens) and gentle suction. During the removal of amygdala, the temporal horn of the lateral ventricle is entered, allowing clearer orientation of the hippocampus and the extent of the superior, posterior, and lateral aspects of the amygdala. While approaching the ventricular wall, however, particularly in the medial plane, the surgeon must keep in mind the presence of subependymal veins returning from the amygdala. These vessels run subependymally to the atrial vein of the temporal horn, which runs through the choroidal fissure to the basal vein of Rosenthal. The variations in the venous drainage of the insula, MTR, cerebral peduncle, optic tract, and thalamus to the basal vein have been described comprehensively ­elsewhere.23,​32 At this stage of the operation, it is of utmost importance that the optic tract is clearly identified. In this surgical technique, full resection of the amygdala bares certain risks of damaging the optic tract and midbrain. Therefore, we prefer leaving the medial portion of the amygdala intact. After subtotal removal of the amygdala, the rest of the piriform cortex and the anterior part of the parahippocampal gyrus are removed subpially. The transparent curtain of pial and arachnoidal membranes near the lateral part of the carotid cistern and the anterior part of the crural and ambient cisterns may be identified readily anteroinferiorly, after subpial resection. After the pia is opened, important anatomical details can be identified, such as the entrance of the AChA to the choroidal fissure along the crural cistern and the optic tract and the basal vein of Rosenthal, which lie medial to the AChA, the cerebral peduncle, the P2 segment of the posterior cerebral artery, and the oculomotor nerve. The tela choroidea, the transparent membrane from which the choroid plexus arises, can be isolated by displacing the choroid plexus medially over the choroidal fissure. Through the tela choroidea, important structures such as the AChA, the hippocampal vein, and ventricular tributaries of the basal vein of Rosenthal can be identified. Subsequently, fine forceps are used to reflect the choroid plexus medially and open the tela

353

354

IVc  Temporal Lobe Epilepsy and Surgical Approaches

Fig. 37.4  Coronal sections of the fluidattenuated inversion recovery (FLAIR) MRIs demonstrate left-sided hippocampal sclerosis in a 12-year-old child with mediobasal temporal epilep­sy (MTE), and preoperative FDG-PET scans demonstrate left-sided temporal hypoactivity.

c­ horoidea between the choroid plexus and the tenia fimbria. At this point, the hippocampal and uncal branches of the AChA must be coagulated and divided. However, great care must be taken not to injure the main stem of the AChA and its medial branches to the peduncle, optic tract, pallidum, internal capsule, thalamus, lateral geniculate body, and choroid plexus. As the choroidal fissure along the tenia fimbria is opened, the medial part of the parahippocampal gyrus (subiculum) within the lateral wing of the transverse fissure can be identified. Exiting the hippocampal sulcus and entering into the basal vein of Rosenthal are the hippocampal and parahippocampal veins, which run over the subiculum in a lateromedial direction. They should be isolated from the arachnoidal membranes and meticulously preserved. The hippocampus is supplied by the hippocampal arteries, which lie beneath the veins described above and enter the hippocampus most often by penetrating the hippocampal sulcus. They usually originate from the P2 segment just proximal to the P2–P3 junction, or from the P3 segment itself, or from branches of the P3 segment, and occasionally from the AChA. At this stage of the operation, the hippocampal arteries are coagulated and divided. The head of the hippocampus–parahippocampal gyrus is transected at the level of the proximal portion of the fimbria, and en bloc resection is done. The middle portion of the hippocampus and parahippocampal gyrus are removed with suction. We prefer this technique instead of en bloc resection of the whole

hippocampus–parahippocampal gyrus, because it preserves the anterior temporal stem (Fig. 37.4, Fig. 37.5, and Fig. 37.6). The posterior limit of the resection of the hippocampal tail is just at the level of the posterior rim of the cerebral peduncle, some 10 to 15 mm before and inferior to the isthmus cinguli. The resection is performed inferolaterally through the posterior part of the hippocampus–parahippocampal gyrus, in the direction of the collateral sulcus and tentorial edge. Resection continues with forceps and suction along the sulcus, in a semicircular fashion within the temporal horn lateral to the hippocampus, and then enters the collateral and rhinal sulci. This semicircular resection, 4- to 5-cm long and 5- to 10-mm deep, extends down to the free edge of the tentorium, leaving the fusiform gyrus untouched laterally. The loops and branches of the temporo-occipital trunk arising from the P2–P3 junction can be identified within the collateral sulcus. The branches that supply the parahippocampal gyrus are coagulated and divided. As the limbic areas are resected, the hippocampal veins are again exposed, coagulated, and divided at a proper distance from the basal vein of Rosenthal. Occasionally, bleeding may occur from the pial bed of the resected limbic structures in the cavity; these areas require coagulation with bipolar forceps. Generally, opening the extension of the collateral sulcus provides access to the tentorium some 2.5 cm from its free edge and in its anterior half. In patients with herniation of the mediobasal temporal

37  Selective Amygdalohippocampectomy

Fig. 37.5  Postoperative coronal (a, b) and sagittal (c) sections of the fluid-attenuated inversion recovery (FLAIR), and T1-weighted MRIs (d) demonstrate left-sided pterional transsylvian–transamygdalar subarachnoid hemorrhage (SAH) in a 12-year-old child with mediobasal temporal epilep­sy (MTE). Note the preserved temporal stem.

structures, the mediobasal dissection must be done meticulously because of the potential for damage to the underlying structures. In such cases, staying in the subpial plane allows removal of the parahippocampal gyrus, while the P2 segment with its branches, the superior cerebellar artery, third nerve, and fourth nerve (lying below the tentorial edge) are protected by the pia and a double layer of arachnoid. During this procedure, self-retaining retractors should not be placed. Instead, the tip of the suction tube covered with a moist cottonoid sponge to be used as a gentle temporary retractor. After careful hemostasis is achieved in the resection cavity and around the MCA, the ICA, the AChA, the PCoA, and their branches, the dura is closed with a running suture, and the bone flap is replaced in the normal fashion. After the amygdala and uncus–hippocampus–parahippocampal gyrus are removed, the neighboring structures, including the superior, middle, and inferior temporal gyri, and the fusiform (lateral temporo-occipital) and lingual (medial temporo-occipital) gyri remain undamaged. Furthermore, removal of the anterior one-third of the hippocampus–parahippocampal gyrus enables the pathologist to carry out scientific studies on resected structures. As it is less complex and allows the surgeon to preserve the anterior temporal stem, gradual removal or subpial suction of the rest of the hippocampus–parahippocampal gyrus is a preferable approach compared to en bloc resection.

Paramedian Supracerebellar–Transtentorial Selective Amygdalohippocampectomy Following extensive cadaveric studies, we found the PST approach to be suitable in accessing the entire length of the MTR including the amygdala and piriform cortex. We first used this approach to remove a tumor in the entire length of the MTR. The PST approach provides an excellent panoramic view of the whole MTR without disturbing the neighboring structures. We then applied this approach to patients with MTE caused by hippocampal sclerosis, especially in posterior part.9,​16,​36,​37 Hence, we now choose this approach for SAH (Fig. 37.7). We prefer to use this approach in semisitting position. Detailed information on the semisitting positioning was given previously, and only some important aspects are discussed here.16 After endotracheal intubation, the patient undergoes placement of a transesophageal echocardiography (TEE) probe to monitor for possible air embolism. Before the positioning, a “bubble test” with the Valsalva maneuver is done to further evaluate for any possible shunting from the right-toleft atrium. This bubble test is crucial because transthoracic echocardiography, performed preoperatively, may miss atrial or ventricular septal defects, and it is imperative to confirm that there are no such defects. TEE provides detailed realtime information about the heart, great vessels, and every air

355

356

IVc  Temporal Lobe Epilepsy and Surgical Approaches

Fig. 37.6  Postoperative fiber tractography demonstrate preserved uncinate fasciculus (uf), occipitofrontal fasciculus (of), and posterior thalamic peduncle (pt), which includes the optic radiation, and postoperative FDG-PET scans demonstrate a metabolic activity in the left mediobasal temporal region (MTR).

Fig. 37.7  The left paramedian sagittal section through the hippocampal formation of a formalin-fixed cadaveric head is shown. The figure shows the PST approach for SAH. The arrow indicates the route of the surgical approach. Abbreviations with white letters denote sulci and fissures. ab, amygdaloid body; ac, anterior commissure; cer, cerebellum; chp, choroid plexus; cos, collateral sulcus; cr, corona radiata; hi, hippocampus; ocp, occipital pole; pu, putamen; slic, sublentiform portion of the internal capsule; te, tentorium; trs, transverse sinus.

­bubble entering the blood circulation during surgery. Thus, it is possible to detect air inflow at an early stage and p revent any complications with appropriate maneuvers (Fig. 37.8). With the patient in the supine position, the skull clamp of the Mayfield–Kees three-point fixation device is applied to the head. One skull pin is fixed to the mastoid of the contralateral side and two pins to the ipsilateral frontal and parietal regions. While the surgeon holds the patient’s head securely, custom-made leg holder and a pillow are placed to position the patient’s legs approximately 90 degrees to the torso with the knees flexed 90 degrees. The seat section of the operating table should be parallel to the ground and the back section tilted up approximately 25 degrees. This smaller degree of tilt helps prevent complications associated with air embolism and preserves the surgeon’s comfort when the operating table is elevated and the surgeon’s armrest is employed. After the patient is in the desired position, the Mayfield crossbar adaptor is fitted to the accessory rail of the back section of the operating table to allow changes in the degree of elevation of the head without having to go under the drapes to disconnect the Mayfield system during surgery. The patient’s head is then fixed in a neutral-straight position with moderate flexion. Extreme flexion of the head must be avoided, especially in elderly patients. To preserve the

37  Selective Amygdalohippocampectomy

Fig. 37.8  The lateral view of the patient in the semi-sitting position is shown. The seat section (ss) of the operating table is parallel to the ground, and the back section (bs) is tilted up approximately 25 degrees. The head is flexed in the neutral position without any rotation or tilt. The back section’s accessory rail is used for the Mayfield crossbar adaptor (cba). With this modification, the position of the back section can be changed easily without having to go under the drapes to disconnect the Mayfield system from the operating table adapter during surgery. scl, skull clamp; swa, swivel adaptor; te, tentorium; tee-p, probe of the TEE.

intravascular volume of the patient in this semisitting position, we avoid the use of diuretics. A vertical, linear paramedian incision is made. It passes through the midpoint of an imaginary line between the external occipital protuberance and the mastoid process. ­ The incision extends one-third above and two-thirds below the superior nuchal line. Depending on the patient’s skin thickness and muscle mass, the incision may be extended superiorly and inferiorly, but a 12-cm incision is generally enough for exposure. The occipital artery is usually encountered twice, initially superficial to the occipital belly of the occipitofrontalis muscle, and later while dissecting the suboccipital muscles. Next, we split the occipital belly of the fronto-occipital muscle, and the suboccipital muscles are separated with plasma blade in line with the incision. The occipital artery beneath the splenius capitis muscle should be coagulated and divided. The muscle mass is retracted with two Weitlaner’s retractors, one curved superiorly and one straight inferiorly. The external occipital protuberance should be exposed medially and the asterion laterally to allow sufficient exposure for subperiosteal dissection. There is no need to expose the foramen magnum region. Three burr holes are then made. The first is placed in line with the skin incision and 2 cm above the superior nuchal line, as this is the most superior aspect of the exposure. The second hole is placed just lateral to the external occipital protuberance, right over the transverse sinus and just lateral to the torcular, and the third hole is placed on the asterion (the junction of the transverse and sigmoid sinuses). A dural dissector is then passed to separate the dura and the transverse sinus from the bone, and the bone flap is turned with the craniotome, which extends one-third above the transverse sinus and two-thirds below the transverse sinus. Placing the craniotomy well above the transverse sinus helps in two ways: the exact location of the ­transverse sinus varies, and the transverse sinus sometimes

needs to be pulled up depending on the tentorial angle. We try not to expose the mastoid air cells and, irrespective of opening air cells, we apply bone wax to the bone edges of the mastoid side to prevent CSF leak. With this exposure, we usually have approximately 6 cm of bone uncovered in the mediolateral plane and 5 cm in the superoinferior plane. This space is necessary to achieve a wide mediolateral corridor, which allows excellent visualization and enough room for maneuvering during the operation. At the stage of dural opening, the operating microscope is introduced. It is critical to mention that we use a surgical microscope with a 300-mm objective lens without autofocus and zoom functions (OPMI 1 FC, Zeiss, Oberkochen, Germany). The counterweight-balanced microscope stand is controlled with a pistol grip hand switch and mouth switch (Contraves, Zurich, Switzerland). We make two openings in the dura. The first is a transverse opening about 15-mm long placed right over the lower aspect of the craniotomy to access the cisterna magna. CSF is then released to relax the posterior fossa and allow the supracerebellar infratentorial corridor to open. A cottonoid is left here, allowing access to the cisterna magna to release more CSF as necessary. The transverse and occipital sinuses should not be opened. Before the second incision, the micro-Doppler (Mizuho America, Inc., Beverly, MA) is used to confirm the exact location and extension of the transverse sinus. A curvilinear opening in the dura is made about 15 mm below the transverse sinus. This opening is parallel to the transverse sinus in the center of the exposure. At the ends of the exposure, the dural incision curves up closer to the transverse sinus. It is important to open the dura wide from the midline to the lateral aspect of the opening to create a wide mediolateral corridor. This corridor allows visualization of the supracerebellar space to work between the tentorial draining veins without sacrificing them. This wide corridor also allows visualization of the medial temporal structures beyond the “napkin ring” formed by the midbrain medially and the petrous ridge laterally. When the surgeon approaches the transverse sinus during the second dural opening, a small, inadvertent opening may appear in the transverse sinus, which might lead to an air inflow to the blood circulation. Thus, we alert our anesthesia team to watch for that phenomenon. After dural opening, adhesions of arachnoid villi are commonly found from the most posterior– superior aspect of the cerebellum to the dura. We release these arachnoid villi adhesions, a maneuver that opens the supracerebellar operative corridor. After it is opened, the dura is reflected over the transverse sinus. At this point, the patency of the transverse sinus is checked again with a micro-Doppler to make sure that tenting of the dura did not compromise flow. The superior (tentorial) surface of the cerebellum is dissected from the tentorium (Fig. 37.9). The supracerebellar space is then explored and special attention is paid to understand any venous variations. Generally, one small vein in the supracerebellar paramedian region drains to the tentorium. In our first cases, we sacrificed these veins to get sufficient working space. Although there were no clinical or radiological problems with this sacrifice, later on we try to preserve these veins. We were able to mobilize the veins and modify the tentorial incision according to the venous variations to open enough space in most of the cases. The surgeon must

357

358

IVc  Temporal Lobe Epilepsy and Surgical Approaches

Fig. 37.9  (a) The anatomical landmarks necessary to localize the right-sided paramedian supracerebellar transtentorial (PST) craniotomy and the operative field after the dural opening are shown in this illustration. The vertical dashed line indicates the skin incision. Please note the small dural opening just above the inferior border of the craniotomy, which is made to reach cisterna magna to release CSF. The ideal spots for the three burr holes are demarcated. The first burr hole is placed 2 cm above the superior nuchal line; the second burr hole is placed just lateral to the external occipital protuberance right over the transverse sinus, and the third burr hole is placed on the asterion just medial to the junction of the transverse and sigmoid sinuses. Care must be taken not to open the mastoid air cells to prevent a CSF leak. Asterix indicates the tentorial incision. (b) Illustration shows the inferior view of the middle portion of the right MTR (M) and the midbrain after the tentorial opening. The exposure is wide enough without retraction of the cerebellum. The anterior leaflet of the tentorium is retracted downward with three holding sutures and Dandy’s forceps, and the posterior leaflet is retracted upward. (c) The illustration of the surgical field after the selective removal of the right amygdala, hippocampus, and parahippocampal gyrus is shown. The asterisk indicates the bed of the amygdala. A1, precommunicating (first) segment of the anterior cerebral artery; bu, base unit; bvr, basal vein of Rosenthal; cer, cerebellum; chp, choroid plexus; cos, collateral sulcus; cp, cerebral peduncle; eoc, external occipital crest; fg, fusiform gyrus; ICA, internal carotid artery; lgb, lateral geniculate body; M, middle portion of the mediobasal temporal region (MTR) ; M1, sphenoidal (first) segment of the middle cerebral artery; P2, second segment of the posterior cerebral artery ; P3, third segment of posterior cerebral artery; pt, pulvinar of the thalamus; rth, roof of the temporal horn; trs, transverse sinus; zpr, zygomatic process; 3, inferior posterior temporal artery; 4, inferior temporo-occipital artery; II, optic nerve; III, oculomotor nerve; IV, trochlear nerve.

not sacrifice the lateral petrosal vein or the tentorial veins around the midline, which are the main veins draining the posterior fossa.

Dissection is carried forward to open the posterior aspect of the quadrigeminal cistern and the ambient cistern and to identify the superior cerebellar artery, the fourth nerve, the pineal

37  Selective Amygdalohippocampectomy gland, the vein of Galen, and the basal vein of Rosenthal. Inferior colliculus, the most important landmark during surgery, is identified. To take advantage of the mediolateral corridor, the arachnoid should be opened medially from the quadrigeminal cistern all the way to the lateral aspect of the ambient cistern. During the dissection of the arachnoid to uncover ambient and quadrigeminal cisterns, care must be taken to identify the fourth nerve. The fourth nerve advances between the posterior cerebral artery and superior cerebellar artery after emerging from the dorsal aspect of the crus cerebri. Before the tentorial opening, the fourth nerve should be followed until it pierces the tentorium. Although the fourth nerve remains below of trajectory of the PST approach, its presence should always be kept in mind during the rest of the surgery. The tentorium is coagulated and incised. An incision is made about 2 cm posterolateral from the posterior portion of the tentorial incisura, right in the center of the exposure. A cottonoid is placed above this opening to protect the supratentorial structures. The incision is then gently carried toward the tentorial hiatus to the lateral and posterior aspect of the quadrigeminal cistern. The incision is then extended from the starting point anteriorly to the midpoint of the petrous ridge. After the tentorium has been cut in this fashion, the anterior leaflet of the tentorium is retracted downward with three sutures, with Dandy’s forceps hanging from the tails of the sutures. The cerebellum hangs by gravity and the self-retraining retractor is never used in any part of the surgery. Next, the posterior leaflet is pulled up in a similar fashion. Tack-up sutures are then placed to help elevate the upper leaflet of the tentorium to displace the transverse sinus. This maneuver opens more space for visualization and manipulation. In some patients, we modify this tentorial incision because of venous variations. We leave a cuff of dura around the tentorial draining vein; hence, we are able to preserve the vein in most patients (Fig. 37.9b). After this maneuver, various anatomical structures can be identified as landmarks to help with orientation during the rest of the surgery. The inferior colliculus, the trochlear nerve, P3 segment of the posterior cerebral artery and superior cerebellar artery, the galenic venous system, and the superior petrosal vein are helpful landmarks throughout the surgery. Large portions of the MTR and fusiform gyrus become visible. The collateral sulcus, which separates the parahippocampal gyrus and fusiform (medial temporo-occipital) gyrus, should be identified. In patients with hippocampal sclerosis, because of atrophic mediobasal structures, the uncus and third nerve may be disclosed at this step. In patients with tumors, however, it may be dangerous to try to expose these structures because of the voluminous tumor tissue in this step. These various landmarks guide the surgeon during the rest of surgery; this is the great advantage of the PST approach. To remove the parahippocampal gyrus, hippocampus, and amygdala, intraoperative surface EEG is done first, and then a depth electrode is passed into the hippocampus and the parahippocampal gyrus. Recordings are made in all cases. The inferior colliculus is the main landmark for the starting point of the parahippocampal gyrus resection. In patients with MTE, the inferior colliculus is also the posterior limit of the parahippocampal gyrus resection, as the anterior extension of the visual cortex needs to be preserved. In patients with tumors extending behind this point, the tumor is dissected anterior to this point initially. We then reach and resect the posterior

portion at the end of the surgery. The collateral sulcus, which marks the lateral limit of resection, is identified next. The choroidal fissure and midbrain form the medial limit of resection. A perpendicular posterior parahippocampal incision is made with bipolar forceps at the level of the inferior colliculus, which is limited by the collateral sulcus laterally. Subpial dissection is done, preserving the major inferior temporal arteries from the P3 to P2 segments of the posterior cerebral artery. Next, subpial resection of the posterior aspect of the subiculum is done and is carried forward to the uncus medially. The hippocampal arteries from the P2 to P3 segments of the posterior cerebral artery are identified, coagulated, and cut. The uncal arteries are then identified, coagulated, and cut. Subpial dissection is continued laterally to identify the collateral sulcus and, with gentle dissection of the inferior aspect of the parahippocampal gyrus just medial to the collateral sulcus, the dentate gyrus is identified by its distinct grayish color. The alveus is then dissected, and the posterior aspect of the temporal horn, near the atrium, is entered. Unique to this exposure is that the roof of the temporal horn is in line of sight. The choroidal fissure is identified in the temporal horn of the lateral fissure, and dissection is carried forward to the tip of the temporal horn. Dissection is performed between the fimbria and the choroid plexus. At this point, the body and most of the head of the hippocampus are ready to be delivered without damage, allowing for detailed ­pathological examination of the hippocampus in patients with MTE. In patients with tumors, the hippocampus and parahippocampal gyrus are removed gradually with bipolar forceps, suction tubes, and ultrasonic surgical aspiration. The remainder of the head of the hippocampus, the uncus, and the amygdala are then resected, taking care to stop at the point where the lateral wall meets the roof of the temporal horn near the collateral eminence. Resection of the amygdala commences at its junction with the tail of the caudate nucleus; the amygdala is identified clearly by its hazelnut color under the ependymal layer. This technique offers best view of the amygdala and complete resection when compared to other surgical methods. This technique prevents injury to the optic radiation within Meyer’s loop. Caution must be exercised during subpial resection of the medial aspect of the amygdala as both the optic tract and the AChA are at risk. We prefer to use only gentle suction at this stage of the procedure, when we can identify the crural cistern subpially and the point where the AChA departs the ICA. At this point, the high-definition (HD) neuroendoscope (Aesculap, Tuttlingen, Germany) is introduced to visualize the inferior and anterior aspects of the parahippocampal gyrus. This area is not well seen through the microscope, as it is hidden from view by the petrous ridge. This phenomenon is truer in patients with tumors than in those with MTE, in whom we try SAH. The technique of placing a cotton ball below the parahippocampal gyrus brings it into view; therefore, resecting the most anterior and inferior aspects of the parahippocampal gyrus becomes possible. With the HD neuroendoscope and a suction tube with a curved tip, we continue to resect the remaining tissue in this region, attaining hemostasis. The ICA with its perforators, the AChA, the PCoA, the third and fourth nerves, the roof of the temporal horn with the choroid plexus and the tail of the caudate nucleus, the P2 and P3 segments of the posterior cerebral artery with their main branches, the superior

359

360

IVc  Temporal Lobe Epilepsy and Surgical Approaches c­ erebellar artery, and the basal vein of Rosenthal should be seen (Fig. 37.9c). After hemostasis is achieved, the tentorial sutures are released and reapproximated to reconstruct the anatomical position of the tentorium. The small inferior and large superior incisions of the dura of the posterior fossa are closed watertight and the bone flap is replaced with skull-fixation devices. The occipital belly of the occipitofrontal muscle is reapproximated and the suboccipital muscles are brought together with 2–0 Vicryl. The wound is then closed. Patients are extubated in the operating room at the end of the procedure, then transferred to the intensive care unit until the following morning. In the intensive care unit, the patient is kept with the head elevated 30 degrees. Postoperative MRI is done the morning after surgery. A detailed MRI with fiber tractography is repeated routinely 2 to 3 months after surgery (Fig. 37.10 and Fig. 37.11). The PST discloses excellent landmarks during the entire surgery. Therefore, the PST is not a difficult approach compared to others. Of course, it does require a totally different perspective of the MTR region, but this perspective can be studied with cadaver dissections and provides excellent orientation with precise landmarks. The brain stem and major vascular structures near the MTR are located medially, and the vascular course with its branches is visualized intracisternally, preventing major injury to these structures during resection of the MTR. Neither the visual cortex nor the optic radiation is related to the surgical trajectory, and therefore the risk of injury to these structures is minimal. Nonetheless, the surgeon should be aware of the location of the lateral geniculate body when opening the choroidal fissure. For patients with MTE, this technique facilitates real SAH with complete resection of the amygdala and more posterior resection of the hippocampus. This approach may be an advantage for more successful seizure control and better

neuropsychological and ­psychosocial outcome in patients with posteriorly located hippocampal sclerosis. However, this difference should be verified through further studies. Long-standing discussions and controversies regarding the visual field defect after epilepsy surgery38,​39,​40,​41 can be avoided by using the PST approach. With precise knowledge of surgical anatomy and microsurgical technique, many surgeons have achieved excellent results with the pterional–transsylvian approach without visual field defects.3,​17,​42 The senior author (UT) has had the same experience when using that approach for SAH. After using the PST approach, however, we feel more comfortable using it for MTE cases as well, especially if hippocampal sclerosis is evident in the posterior portion. The PST approach may be preferable, as it allows more posterior resection compared to the pterional–­transsylvian approach. We must mention, however, that removing a sclerotic hippocampus in MTE cases is much easier than removing MTR tumors, as a small and atrophied hippocampus and an evident temporal horn make the surgery less complicated, since even removing the hippocampus in one piece becomes possible for the sake of research studies. We recommend the PST approach as an alternative for those who have difficulty with SAH through the pterional transsylvian–transamygdalar approach. One possible criticism of the PST approach for MTE patients is the difficulty of temporal pole resection, if it is necessary. As we gained more experience using this approach in tumor cases, however, we noticed that it is possible to remove the temporal pole. Recurrent seizures after anterior SAH may be due to iatrogenic damage to the temporal neocortex, sylvian region, or temporal pole. However, none of these areas is damaged during the PST approach. This difference also needs further evaluation and studies in large series. A critical factor in the success of the PST approach is to have a small optical head for the surgical microscope. For the surgeon to work in the depth of a small space with both hands c­ ontinuously,

Fig. 37.10  A 12-year-old patient with an 11-year history of intractable seizures. Preoperative axial T2- and coronal T1-weighted inversion recovery MRIs disclose right-sided hippocampal sclerosis.

37  Selective Amygdalohippocampectomy

Fig. 37.11  Postoperative MRIs (T2 axial, T2 coronal, and T1 sagittal) of the same patient disclose resection after the Selective amygdalohippocampectomy-paramedian supracerebellar transtentorial-subarachnoid hemorrhage (PST-SAH). The patient is seizure free without any medication for 6 years following surgery.

without stretching the arms, the use of a surgical microscope without autofocus and zoom functions that can be controlled with a mouth switch is important. The hydraulic armrest is also important equipment for the surgeon’s comfort. Like many other approaches, the PST approach has potential problems and difficulties. For example, the semisitting position needs special attention from the neurosurgeon, anesthesiologist, nurses, technicians, and the entire operating team.35,​43,​44 However, modern neuroanesthesiology with TEE and modern monitoring methods have made this position safe. It is important to mention that the absence of any septal defect must be confirmed with TEE through the bubble test, as transthoracic echocardiography may miss small d efects. If the atrial septal defect was detected using TEE, we can use the pterional transsylvian–transamygdalar approach for SAH.

Surgical Considerations The mediobasal temporal structures, such as the amygdala, uncus, hippocampus, and parahippocampal gyrus, can vary considerably in size and form.20,​28 Advances in neuroradiology and particularly in MRI technology have provided considerable help in studying the extension, variations, and exact preoperative and postoperative morphology of the MTR.1,​45 The MRI findings suggesting hippocampal sclerosis include the loss of h ­ ippocampal volume on coronal thin-slice FLAIR and T2-weighted images with associated ex vacuo enlargement of the temporal horn of the lateral ventricle and the loss of gray or white matter differentiation within the temporal tip.45 The presence of small vascular malformations or gliomas is also readily identified. It is imperative that the surgeon should review the

361

362

IVc  Temporal Lobe Epilepsy and Surgical Approaches MRI images with details to appreciate the MTR anatomy of the patient before proceeding for the surgical intervention. Pterional transsylvian–transamygdalar SAH can be regarded as safe, as it does not damage the neocortical part of the temporal lobe. After SAH, the remaining structures of the temporal lobe, namely the superior, middle, and inferior temporal gyri, and the fusiform and lingual gyri remain surgically untouched. Our experience suggests that the pterional transsylvian–transamygdalar approach is superior to prevent injury to the temporal stem when compared to a standard temporal ­lobectomy. This is demonstrated by postoperative fiber tractography (Fig. 37.6). The available data imply that preserving more functional temporal lobe tissue is ­critical for improved preservation of a patient’s neurocognitive function postoperatively, particularly in high-functioning ­individuals with dominant temporal lobe epilepsy. In transsylvian SAH, about 10 to 20% of the most medial part of the amygdala, where it abuts the basal ganglia, anterior commissure, and tail of the caudate nucleus, remains intact. Thus, the amygdala is not totally removed. Furthermore, posterior transection of the hippocampus–parahippocampal gyrus is generally done at the level of the posterior margin of the cerebral peduncle, where the P2–P3 segment junction is located, at the level of the ascending tail of the hippocampus and 10 to 15 mm before the isthmus cinguli. So, further dissection far posteromedially for more posterior resection poses risk of ­damaging the Meyer’s loop.3 PST-SAH due to its anatomical characteristic exposes the MTR to the direct view of surgical trajectory with its whole extent. ­Initial resection of parahippocampal gyrus directly discloses whole extent of hippocampus and faces surgical trajectory toward amygdala via its ­inferolateral axis. Such trajectory enables clear visualization of most superomedial border of amygdala as resected. The posterior portions of the hippocampus and parahippocampal gyrus can be removed through this approach, possibly allowing better seizure control, especially in patients with posterior hippocampal sclerosis. Additionally, PST-SAH is a good alternative to other amygdalohippocampectomy techniques, since there is no danger of injuring the major white matter structures, such as the uncinate fasciculus, anterior commissure, inferior and posterior thalamic peduncles (including the optic radiation), and the frontopontine, temporopontine, and occipitopontine fasciculi. Furthermore, there is no danger of damage to the superficial and deep sylvian veins, the middle cerebral artery and

its branches, and especially the temporal pole and temporal ­neocortex. A comprehensive surgical video for the PST-SAH was published recently.9

„„ Conclusion The comparison of the most commonly used surgical techniques in the management of MTE is still an unresolved issue.5 According to a recently published review, SAH appears to have similar seizure outcome and possibly a better cognitive outcome than temporal lobectomy. If the patient data that are obtained using appropriate preoperative tests strictly show the origin of the seizures to be unilateral MTR, and again, if detailed MRI studies show that clear hippocampal sclerosis but lateral cortical organizations are normal, then SAH provides an ­opportunity to reach and to remove the pathological region without disturbing overlying healthy tissue. In addition, patients with dominant temporal lobe epilepsy who are highly functional, or those whose preoperative Wada results ­suggest a risk of significant verbal memory loss with temporal neocortical resection, the selective procedure is the preferred operation. If, however, preoperative evaluation data show discordant findings, then the epilepsy team should carry out additional tests, such as invasive EEG monitoring. There is an increasing tendency for recommending surgical interventions to pediatric epilepsy patients in earlier ages.46 This recommendation is based on previous observations that chronic epilepsy in children inhibits the development of personality and is frequently associated with debilitating behavioral and psychiatric problems, including temper tantrums, aggressiveness, attention deficit disorders, and hyperactivity.47–​​50 Although the appropriate candidates for SAH in pediatric age group is much limited than adults, this surgical approach is a viable approach in some carefully selected pediatric epilepsy patients. To perform SAH safely and effectively, an exact knowledge of the vascular supply and the surgical anatomy is essential along with microsurgical techniques. SAH is a safe microneurosurgical procedure with a favorable chance of success in relieving medically refractory seizures originating in the MTR. Postoperative neuropsychological performance also seems to be better with selective resection of MTR. In our experience both SAH techniques, pterional transsylvian–transamygdalar approach and PST approach, are safe and effective. Especially for posterior hippocampal sclerosis, we prefer the PST approach over the other.

References ­1. Widjaja E, Raybaud C. Advances in neuroimaging in patients with epilepsy. Neurosurg Focus 2008;25(3):E3 2. Türe U, Yaşargil MG, Friedman AH, Al-Mefty O. Fiber dissection technique: lateral aspect of the brain. Neurosurgery 2000;47(2):417–426, discussion 426–427 3. Yaşargil MG, Türe U, Yaşargil DCH. Impact of temporal lobe surgery. J Neurosurg 2004;101(5):725–738 4. Tanriverdi T, Olivier A, Poulin N, Andermann F, Dubeau F. Long-term seizure outcome after mesial temporal lobe epilepsy surgery: corticalamygdalohippocampectomy versus selective amygdalohippocampectomy. J Neurosurg 2008;108(3):517–524 5. Schramm J. Temporal lobe epilepsy surgery and the quest for optimal extent of resection: a review. Epilepsia 2008;49(8):1296–1307

6. Gleissner U, Sassen R, Schramm J, Elger CE, Helmstaedter C. Greater functional recovery after temporal lobe epilepsy surgery in children. Brain 2005;128(Pt 12):2822–2829 7. Khan N, Wieser HG. Psychosocial outcome of patients with amygdalohippocampectomy. J Epilepsy 1992;5(2):128–134 8. Paglioli E, Palmini A, Portuguez M, et al. Seizure and memory outcome following temporal lobe surgery: selective compared with nonselective approaches for hippocampal sclerosis. J Neurosurg 2006;104(1):70–78 9. Harput MV, Türe U. In reply: the paramedian supracerebellar-­ transtentorial selective amygdalohippocampectomy for mediobasal temporal epilepsy. Oper Neurosurg (Hagerstown) 2018;15 (3):E34–E35

37  Selective Amygdalohippocampectomy 10. Hori T, Yamane F, Ochiai T, et al. Selective subtemporal amygdalohippocampectomy for refractory temporal lobe epilepsy: operative and neuropsychological outcomes. J Neurosurg 2007;106(1):134–141 11. Miyagi Y, Shima F, Ishido K, et al. Inferior temporal sulcus approach for amygdalohippocampectomy guided by a laser beam of ­stereotactic navigator. Neurosurgery 2003;52(5):1117–1123, discussion 1123–1124 12. Niemeyer P. The transventricular amygdala-hippocampectomy in temporal lobe epilepsy. In: Baldwin M, Bailey P, eds. The Temporal Lobe Epilepsy. Springfield, IL: Charles C Thomas; 1958:461–482 13. Olivier A. Transcortical selective amygdalohippocampectomy in temporal lobe epilepsy. Can J Neurol Sci 2000;27(Suppl 1):S68– S76, discussion S92–S96 14. Park TS, Bourgeois BF, Silbergeld DL, Dodson WE. Subtemporal transparahippocampal amygdalohippocampectomy for surgical treatment of mesial temporal lobe epilepsy. Technical note. J Neurosurg 1996;85(6):1172–1176 15. Shimizu H, Suzuki I, Ishijima B. Zygomatic approach for resection of mesial temporal epileptic focus. Neurosurgery 1989;25(5):798–801 16. Türe U, Harput MV, Kaya AH, et al. The paramedian supracerebellar-transtentorial approach to the entire length of the mediobasal temporal region: an anatomical and clinical study. Laboratory investigation. J Neurosurg 2012;116(4):773–791 17. Wieser HG, Yaşargil MG. Selective amygdalohippocampectomy as a surgical treatment of mesiobasal limbic epilepsy. Surg Neurol 1982;17(6):445–457 18. Yaşargil MG, Teddy PJ, Roth P. Selective amygdalohippocampectomy: operative anatomy and surgical technique. In: Symon L, ­Brihaye J, Guidetti B, et al., eds. Advances and Technical Standards in Neurosurgery. Vienna: Springer; 1985:93–123 19. Wieser HG; ILAE Commission on Neurosurgery of Epilepsy. ILAE Commission Report. Mesial temporal lobe epilepsy with hippocampal sclerosis. Epilepsia 2004;45(6):695–714 20. Duvernoy H. The Human Hippocampus. 4th ed. Berlin: Springer-Verlag; 2013 21. Erdem A, Yaşargil G, Roth P. Microsurgical anatomy of the hippocampal arteries. J Neurosurg 1993;79(2):256–265 22. Gloor P. The Temporal Lobe and Limbic System. New York, NY: Oxford University Press; 1997 23. Huang YP, Wolf BS. The basal cerebral vein and its tributaries. In: Newton TH, Potts DG, eds. Radiology of the Skull and Brain. St Louis, MO: CV Mosby; 1974:2111–2154 24. Klingler J, Gloor P. The connections of the amygdala and of the anterior temporal cortex in the human brain. J Comp Neurol 1960;115:333–369 25. Marinković S, Milisavljević M, Puskas L. Microvascular anatomy of the hippocampal formation. Surg Neurol 1992;37(5):339–349 26. Marinković SV, Milisavljević MM, Vucković VD. Microvascular anatomy of the uncus and the parahippocampal gyrus. Neurosurgery 1991;29(6):805–814 27. Mega MS, Cummings JL, Salloway S, Malloy P. The limbic system: an anatomic, phylogenetic, and clinical perspective. J Neuropsychiatry Clin Neurosci 1997;9(3):315–330 28. Nieuwenhuys R, Voogd J, Huijzen C van. The Human Central ­Nervous System. 4th ed. Berlin: Springer; 2008 29. Türe U, Pamir MN. Small petrosal approach to the middle portion of the mediobasal temporal region: technical case report. Surg Neurol 2004;61(1):60–67, discussion 67 30. Türe U, Yaşargil DCH, Al-Mefty O, Yaşargil MG. Topographic anatomy of the insular region. J Neurosurg 1999;90(4):720–733

31. Wen HT, Rhoton AL Jr, de Oliveira E, et al. Microsurgical anatomy of the temporal lobe: part 1: mesial temporal lobe anatomy and its vascular relationships as applied to amygdalohippocampectomy. Neurosurgery 1999;45(3):549–591, discussion 591–592 32. Wolf BS, Huang YP. The insula and deep middle cerebral venous drainage system: Normal anatomy and angiography. Am J Roentgenol Radium Ther Nucl Med 1963;90:472–489 33. Yaşargil MG. Thieme; 1984

Microneurosurgery,

Vol

I.

Stuttgart:

Georg

34. Yaşargil MG. Microneurosurgery IVA. Stuttgart: Georg Thieme Verlag; 1994 35. Yaşargil MG. Microneurosurgery IVB. Stuttgart: Georg Thieme Verlag; 1996 36. Harput MV, Türe U. The paramedian supracerebellar-transtentorial approach to remove a posterior fusiform gyrus arteriovenous malformation. Neurosurg Focus 2017;43(VideoSuppl1, Suppl1):V7 37. Manilha R, Harput VM, Türe U. The paramedian supracerebellar-transtentorial approach for a tentorial incisura meningioma: 3-dimensional operative video. Oper Neurosurg (Hagerstown) 2018;15(1):102 38. Cushing H. The field defects produced by temporal lobe lesions. Brain 1922;44:341–396 39. Falconer MA, Wilson JL. Visual field changes following anterior temporal lobectomy: their significance in relation to Meyer’s loop of the optic radiation. Brain 1958;81(1):1–14 40. Marino R Jr, Rasmussen T. Visual field changes after temporal lobectomy in man. Neurology 1968;18(9):825–835 41. Polyak SL. The Vertebrate Visual System: Its Origin, Structure, and Function and Its Manifestations in Disease with an A ­ nalysis of Its Role in the Life of Animals and in the Origin of Man, Preceded by a Historical Review of Investigations of the Eye, and of the Visual Pathways and Centers of the Brain. Chicago, IL: University of Chicago Press; 1957 42. Yaşargil MG, Krayenbühl N, Roth P, Hsu SP, Yaşargil DC. The selective amygdalohippocampectomy for intractable temporal limbic seizures. J Neurosurg 2010;112(1):168–185 43. Jadik S, Wissing H, Friedrich K, Beck J, Seifert V, Raabe A. A standardized protocol for the prevention of clinically relevant venous air embolism during neurosurgical interventions in the semisitting position. Neurosurgery 2009;64(3):533–538, discussion 538–539 44. Matjasko J, Petrozza P, Cohen M, Steinberg P. Anesthesia and surgery in the seated position: analysis of 554 cases. Neurosurgery 1985;17(5):695–702 45. Jack CR Jr, Rydberg CH, Krecke KN, et al. Mesial temporal sclerosis: diagnosis with fluid-attenuated inversion-recovery versus spinecho MR imaging. Radiology 1996;199(2):367–373 46. Mittal S, Montes JL, Farmer JP, et al. Long-term outcome after surgical treatment of temporal lobe epilepsy in children. J Neurosurg 2005;103(5, Suppl):401–412 47. Harbord MG, Manson JI. Temporal lobe epilepsy in childhood: reappraisal of etiology and outcome. Pediatr Neurol 1987;3(5):263–268 48. Kotagal P, Rothner AD, Erenberg G, Cruse RP, Wyllie E. Complex partial seizures of childhood onset. A five-year follow-up study. Arch Neurol 1987;44(11):1177–1180 49. Lindsay J, Ounsted C, Richards P. Long-term outcome in children with temporal lobe seizures. I: Social outcome and childhood factors. Dev Med Child Neurol 1979;21(3):285–298 50. Wyllie E. Surgical treatment of epilepsy in pediatric patients. Can J Neurol Sci 2000;27(2):106–110

363

38

  Surgical Management of Lesional Temporal Lobe Epilepsy Oğuz Çataltepe

Summary Temporal lobe lesions constitute significant amount of surgical specimens obtained from children with intractable temporal lobe epilepsy (TLE). The most common pathological substrates are related to mesial temporal sclerosis (MTS), tumors, vascular abnormalities, gliosis, and developmental disorders. Among neoplasms, developmental tumors, such as dysembryoplastic neuroepithelial tumor (DNET), or slowly growing low-grade glial tumors and oligodendrogliomas are the most common ones. Surgical interventions for lesional TLE in children have dual therapeutic goals: stopping the seizures and removing the lesion while preserving cortical function. However, epileptogenic zone may frequently stretch beyond the anatomical boundaries of the lesion. Therefore, determining the extent of resection in lesional epilepsy patients is critical to optimize surgical outcome, and it may not always be feasible and certain compromises may be required in some cases. The factors effecting good outcome in epilepsy patients are not only dependent on the type of the lesion but also related to the location of the lesion as well as the extent of the epileptogenic zone and the area of resection. Therefore, the surgical strategy in lesional epilepsy patients is a multifaceted topic. Although there is consensus on critical significance of complete lesion resection, the importance of additional resection of adjacent cortex or the mesial temporal structures remains controversial. Keywords:  lesional epilepsy, developmental neoplasms, vascular lesions, dual pathology

„„ Introduction Temporal lobe lesions constitute 30 to 70% of surgical specimens obtained from children with intractable TLE.1,​2,​3,​4,​5,​6,​7,​8

Developmental brain tumors and low-grade neoplasms are the most common causes of TLE in children. Although hippocampal sclerosis is very common lesion in adults with TLE, it is less frequently seen in the pediatric patient population. In this chapter, we will discuss the surgical strategies in children with lesional temporal epilepsy, mainly focusing on tumors and vascular malformations. Two other common neuropathological substrates, cortical dysplasia and MTS, are subjects of other chapters in this book. The goal in neuro-oncological surgery is always total resection of the tumor when it is feasible. This is also relevant for tumor-related epilepsy cases with one significant difference; an equally important additional surgical target: the epileptogenic zone. Therefore, surgical interventions for lesional TLE in children have dual therapeutic goals: stopping the seizures and removing the lesion while preserving cortical function. These dual therapeutic goals can be achieved by determining both the causative relationship between lesion and seizures as well as the spatial relationship between lesion and epileptogenic zone. Although epileptogenic zone frequently corresponds to the cortex immediately adjacent to the lesion, it may also stretch beyond the anatomical boundaries of the lesion. Therefore, determining the extent of resection in lesional epilepsy patients is critical to optimize surgical outcome, and it might be quite challenging at times. Furthermore, reaching these goals may not always be feasible, and certain compromises may be required in some cases. In this chapter, we will discuss the surgical management of lesional TLE patients based on published data and our own clinical experience. There are limited published data specifically obtained from children with TLE, and there are even smaller numbers of reports addressing lesional TLE in children. The real frequency of lesional TLE cases is not clear, and reported rates show wide variations in large surgical series (Table 38.1).1,​2,​3,​4,​5,​6,​7,​8

Table 38.1  Pathological substrates in pediatric lesional TLE series

Number

Tumor

MTS

CD

Vasc.

Dual pathology

42

33%

19%

9%

—

11%

Clusmann et al

89

46%

31%

1%

—

—

Mittal et al3

109

35%

45%

35%

5%

25%

Benifla et al

126

52%

13%

7%

3%

8%

5

Kan et al

33

28%

28%

22%

9%

5%

Kim et al6

59

54%

23%

18%

—

—

20

40%

20%

30%

—

—

Sinclair et al

1 2

4

Maton et al

7

Abbreviations: CD, cortical dysplasia; MTS, mesial temporal sclerosis; TLE, temporal lobe epilepsy; Vasc., vascular lesion.

38  Surgical Management of Lesional Temporal Lobe Epilepsy

„„ Pathological Substrate The most common pathological substrates in medically refractory epilepsy patients are MTS, tumors, vascular abnormalities, gliosis, and developmental disorders.9 Here, we will review lesional TLE in childhood by focusing on neoplasms and vascular abnormalities and refer the reader to related chapters in this book for details regarding the surgical management of MTS and cortical dysplasia. Published series describing pathological substrates in TLE patients mostly include mixed patient populations, both adults and children. Only a few reports exclusively cover the pediatric age group.1,​2,​3,​4,​5,​6,​7,​8,​10 As seen in Table 38.1, published data are heterogeneous and most likely heavily biased by referral patterns of epilepsy centers. However, it is still fair to state that neoplasms are the most common pathological substrates seen in children with lesional TLE along cortical dysplasia and vascular lesions. MTS constitutes, at least in some series, a substantial percentage of the cases, but the true frequency of MTS in children with TLE is still not clear.

Tumors The exact frequency of tumors in children with intractable TLE is not well established yet because of the limited number of studies focusing exclusively on the pediatric population. Çataltepe et al11 reported that temporal tumor-related epilepsy patients constituted 40% of pediatric epilepsy surgery cases in their series. Published data imply that the most common neoplasms in TLE patients are developmental tumors, such as DNET, or slowly growing low-grade glial tumors and ­oligodendrogliomas.1,​3,​4,​5,​10,​11,​12,​13 However, the frequencies of the neoplasms seen in epilepsy patients are quite different among published series. Although ganglioglioma is the most common neoplasm in some series, low-grade astrocytoma or DNET is more common in some other studies (Table 38.2). Among other commonly seen glioneuronal tumors are desmoplastic infantile astrocytoma, gangliocytoma or ganglioglioma, papillary glioneuronal tumor, and central neurocytoma.10 The vast majority of patients with temporal lobe tumor–related epilepsy have some common distinctive characteristics, and some authors even define them as a distinct clinicopathological group, so-called “long-term epilepsy-associated tumors (LEATs).”14 These characteristics include a well-­ differentiated histological pattern, frequently both

neuronal and glial differentiation, cortical localization of the tumors (69–91%), frequent involvement of mesial s­ tructures (48%), frequent association with cortical dysplasia (40–80%), ­indolent biological nature, young age, seizures f­requently as the only symptom, long-standing history of seizure disorder, normal neurological examination, and favorable outcome after surgery.11,​13,​14,​15 All these characteristics make temporomesial glioneuronal tumors a distinct anatomo-clinico-pathological group with complex epileptogenic mechanisms with a more widespread epileptic network.16

Vascular Malformations Most common vascular malformations causing epilepsy are cavernous hemangiomas and arteriovenous malformations (AVMs). The most common presenting symptoms of AVM are hemorrhage and seizures. Hemorrhage is by far the commonest initial manifestation of AVM in children, whereas only 10 to 25% of the patients present with seizures.17,​18,​19,​20 Seizures in AVM patients most likely originate from gliotic, nonfunctional brain parenchyma interspersed in and around the AVM nidus. These gliotic changes develop secondary to focal ischemia induced by “steal” phenomena and possibly constitute the main reason for seizures in AVM patients. In Yaşargil's series of 414 operated cerebral AVM patients that included both children and adults, initial manifestations were hemorrhage and seizures in 77.8 and 14.7% of the patients, respectively.21 Yaşargil's series includes 74 children (17.8%) younger than 15 years old, and approximately 11% of them had temporal lobe AVMs (6.8% extramesial and 4.1% hippocampal AVMs). In the same series, seizure as an initial manifestation was found in 40% of the patients, when the AVM was located in the temporal lobe. Cavernous hemangiomas are relatively common congenital lesions that occur in 0.1 to 0.5% of the general population and constitute 5 to 13% of all intracerebral vascular malformations.22,​23,​24,​25,​26,​27 Cavernous hemangiomas occur mostly in the supratentorial region, 15 to 20% of them being in temporal lobe.23,​24,​25,​26,​27,​28,​29 In a large series, temporal lobe location in cavernous hemangioma patients was found to be ­significantly higher (48%), and 40% were located in mesiobasal structures.30 Epilepsy incidence in symptomatic patients with cavernous hemangioma has been reported between 30 and 70%.22,​24,​25,​26,​27,​28,​31,​32,​33 Cavernous hemangiomas that are located in the temporal lobe have a much higher tendency to be

Table 38.2  Tumor types in pediatric TLE series

Sinclair et al1 Clusmann et al

2

Mittal et al3 Çataltepe et al Benifla et al

11

4

Kim et al6 Maton et al

7

Uliel-Sibony et al

10

Number

DNET

GG

LGA

Oligo

14

14%

57%

21%

—

41

17%

56%

7%

7%

38

13%

55%

13%

7%

29

52%

13%

21%

10%

65

15%

24%

24%

—

32

37%

40%

—

9%

8

25%

25%

50%

—

48

6%

25%

41%

12.5%

Abbreviations: DNET, dysembryoplastic neuroepithelial tumors; GG, ganglioglioma; LGA, low-grade astrocytoma; Oligo, oligodendroglioma; TLE, temporal lobe epilepsy.

365

366

IVc  Temporal Lobe Epilepsy and Surgical Approaches ­ ssociated with intractable epilepsy, and they are far more a likely than AVM to be medically refractory.22,​33,​34 Lobar localization, multiple lesions, larger size of the nidus, and/or hemosiderin fringe are other seizure predictors.27 In children, seizures are the most common manifestation of cavernous malformations (45–54%).29,​31,​35,​36 The estimated risk for seizure development was reported as 1.5% per year in patients with single lesions and 2.5%/lesion peryear in patients with multiple lesions.30,​35 Approximately 40% of cases are medically intractable seizures.25,​26,​27

„„ Mechanism of Seizures The epileptogenic mechanisms involved in lesion-related epilepsy are not clear. Several mechanisms, including direct pressure and irritation of the cortical tissue, gliotic changes and disruption of vascular structures of the surrounding cortex, morphological alterations at the cellular level, changes in inhibitory and excitatory neurotransmitter levels, and denervation hypersensitivity have been proposed to play a role.37,​38,​39,​40 Chronic changes in surrounding brain tissue either by mechanical or vascular mechanisms can also be responsible for seizures induced by slowly growing low-grade tumors. Developmental tumors may even have intrinsic epileptogenicity, because they are frequently associated with cortical dysplasia and contain cells with a rich array of neurochemical properties, including altered inhibitory and excitatory local circuits.37,​38 The location of the lesion is a critical factor as well. Brain tumors associated with epilepsy are often located in the cortex or in gray–white matter junction. When the lesion ­ is located in the temporal lobe, its direct or indirect effects on hippocampus may cause seizures. The lesion location may interfere with cortical afferents and efferents and lead to relative deafferentation of a certain cortical area that has intrinsic epileptogenicity. Small hemorrhages in and around the tumors also cause hemosiderin deposits, which are highly epileptogenic.37 Secondary epileptogenesis might also be responsible for seizures in some patients. It has been shown with intracranial electroencephalography (EEG) recordings that approximately one-half of the patients with neocortical temporal tumors have independent epileptogenic areas in ipsilateral mesial structures.41 The proposed mechanisms for seizures in patients with AVM include vascular steal phenomenon, focal ischemia of the adjacent cortex secondary to AV shunting, progressive intralesional and perilesional gliosis, demyelination, hemosiderin lining in AVM bed, and secondary epileptogenesis in the temporal lobe.17,​19,​34 It has also been suggested that mass effect on the surrounding brain, cortical irritation, presence of calcification, gliosis in the surrounding brain tissue, and accumulation of iron-containing substances in hemosiderin fringe are responsible for the seizures in cavernous hemangioma patients.22,​23,​24,​25,​26,​27,​28

„„ Surgical Strategy Although the histological subtype of the lesion is always the major factor influencing clinical outcome in any given patient,

the determinants of the seizure-related outcome in lesional epilepsy are more complicated. The factors effecting good outcome in epilepsy patients are not only dependent on the type of the lesion but also related to the location of the lesion as well as the extent of the epileptogenic zone and the area of resection.39,​41,​42 Therefore, the surgical strategy in lesional epilepsy patients is a multifaceted topic. It should be defined based on the location, extent of the lesion, and epileptogenic zone, as well as the histopathological diagnosis. Determining the optimal surgical strategy in these cases is a challenging task and involves some controversy. Lesionectomy alone, lesionectomy with resection of the epileptogenic zone, or lesionectomy with resection of the ipsilateral mesial structures all have their advocates in discussions about the appropriate surgical approaches in these cases. However, there is limited clinical evidence to support specific resective strategies in lesional TLE cases.39 Therefore, until more data are available, the surgical strategy for each patient should be determined on an individual basis by considering histological type and location of the lesion, extent of the epileptogenic zone, and the spatial relation between the lesion and the epileptogenic zone.

Extent of Resection The spatial relationship between the epileptogenic zone and the lesion is the most critical factor to determine the extent of surgical resection in lesional epilepsy patients. There are several conditions for optimizing seizure control in children with lesional epilepsy. First, the lesion should be completely identified and resected. Second, the epileptogenic zone should be contained within the resected area, and finally the remaining cortical and subcortical areas should not develop independent seizures after surgery. Unsuccessful results in lesional epilepsy surgery are frequently related to incomplete resection of the lesion or epileptogenic zone or the presence of additional or secondary epileptogenic foci.43 Another reason may be having an extensive epileptogenic zone beyond the boundaries of the lesion. Some well-known examples of this include the presence of surrounding gliosis in AVM, hemosiderosis rim associated with cavernous hemangioma, dysplastic areas associated with developmental tumors, and dual pathology.

Lesion Lesionectomy alone is probably the most commonly used surgical approach in lesional epilepsy cases. Although there is a wide consensus regarding the significance of total tumor resection for good seizure control, the results of this surgical approach in the published epilepsy series are quite different. Khajavi et al44 reported that seizure-free outcome was only correlated with the extent of tumor resection but not with additional resection of the surrounding cortex. Conversely, Jooma et al45 reported that epilepsy patients who underwent lesionectomy procedure alone had a significantly lower seizure-free outcome rate compared with patients who had additional cortical resections of the adjacent epileptogenic zone. Sugano et al46 reported that after complete resection of mass lesions, they still found residual spikes in the mesial structures in up to 86% of the patients and recommended additional resection in these areas for better seizure control.

38  Surgical Management of Lesional Temporal Lobe Epilepsy

Epileptogenic Zone The first step in the planning lesional epilepsy surgery is to define the relationship between the localization of the seizures and the location of the lesion. If the clinical and electrophysiological characteristics of the seizures are fully correlated with the location of the lesion, then the next step would be to determine whether the epileptogenic zone exceeds the anatomical boundaries of the lesion. The surgeon needs to work closely with the epilepsy or neurophysiology team to map the epileptogenic zone and to determine its spatial relationships with the structural lesion (Fig. 38.1). The surgical resection strategy is then designed based on lesion location, the extent of the epileptogenic zone, and the relative position of adjacent eloquent cortex. Anterior temporal lobectomy (ATL) including the lesion (Fig. 38.2) or tailored lesionectomy including the surrounding epileptogenic cortex are two commonly used surgical approaches in lesional TLE. Clusmann et al2 describe their “preoperative tailoring” technique as aiming for complete resection of the lesion demonstrated on MRI and extending the resection further whenever clinical or electrophysiological data suggest a seizure onset in the respective distant areas, such as the hippocampus. Imaging techniques such as functional imaging studies, magnetoencephalography, diffusion tensor imaging, intraoperative electrocorticography (ECoG), and

invasive monitoring data obtained with depth or subdural electrodes are among the recommended techniques to determine the extent of the neocortical and hippocampal resections.

Mesial Temporal Structures It has been generally assumed that the primary epileptogenic activity arises in the vicinity of the lesion. However, the situation may be more complicated if the lesion is located in the temporal lobe because secondary involvement of the mesial temporal structures in seizure generation is not rare with temporal lobe lesions. It is well known that the hippocampus is a very epileptogenic structure and may even generate a separate focus for secondary epileptogenesis in these cases. Therefore, much attention is given to the mesial structures in TLE patients even if the lesion is located in temporal neocortex. This is a legitimate concern because it is possible that epileptogenic zone may not be limited to cortex adjacent to the lesion itself but may also include the mesial temporal structures. If that is the case, then resection of the mesial temporal structures along with the lesion may be needed to obtain a seizure-free outcome. However, indications for resection of the adjacent mesial temporal structures in lesional TLE cases are highly controversial. If the tumor directly involves mesial temporal structures, then the surgical decision is relatively

Fig. 38.1  Intraoperative pictures from a 14-year-old patient with a cortical lesion (dysembryoplastic neuroepithelial tumor) associated with focal cortical dysplasia in posterior temporal region. (a) Surgical cavity after resection of the tumor by completely emptying the involved gyrus. (b) Intraoperative electrocorticography after resection of the lesion to map the adjacent dysplastic epileptogenic cortex. (c) Further resection of the adjacent dysplastic cortex based on electrocorticographic mapping.

Fig. 38.2  Preoperative (a) and postoperative (b) axial magnetic resonance images of a 10-year-old patient with ganglioglioma located in the right parahippocampus. The tumor was resected along with anterior temporal lobe and mesiotemporal structures.

367

368

IVc  Temporal Lobe Epilepsy and Surgical Approaches straightforward. Conversely, if mesial temporal structures are not directly in the lesional zone, then the risks and benefits of the resection of mesial temporal structures should be assessed carefully. Well-defined histological and electrophysiological changes in the hippocampus are very important to determine the most appropriate surgical approach in these cases. Unfortunately, imaging findings of hippocampal neuronal losses are quite ­subtle in lesional TLE patients, unlike MTS patients. Therefore, it may be very difficult to appreciate pathological changes, such as neuronal loss, in the hippocampus of the patients with temporal neocortical lesions using available imaging techniques. Even if the hippocampus demonstrates epileptogenic properties by intracranial monitoring techniques, imaging studies may not be grossly abnormal.41 Usui et al41 reviewed 15 TLE patients who had structural lesions in temporal lobe but normal-­ appearing mesial temporal structures on MRI. They reviewed intracranial EEG data obtained through bilateral temporal subdural and depth electrodes and documented independent ictal discharges arising from ipsilateral mesial structures in 47% of the patients. They concluded that these findings support the presence of independent epileptogenicity of mesial structures in lesional TLE. In this series, clinical semiology, scalp EEG, and MRI images were not found predictive for mesial onset of the seizures. Furthermore, the presence of mesial temporal onset of seizures was not associated with hippocampal neuronal loss in histological examination. In another study, Mathern et al47 showed, using depth electrodes, that ictal onset EEG activity either started from mesial temporal structures or first propagated in the mesial contacts of depth electrodes in 94% of the patients with temporal lobe lesions. Mihara and Baba48 identified independent ictal discharges arising from ipsilateral mesial structures in 47% of the patients with neocortical temporal lesions and concluded that there is independent epileptogenicity of both neocortical lesions and ipsilateral mesial structures in this subgroup of patients. The potential complexity of accurately localizing the epileptogenic zone in lesional TLE patients causes difficulties in determining the correct surgical approach in many cases. Clusmann et al2 classified their surgical approach toward mesial structures of lesional TLE cases in five groups as complete amygdalohippocampectomy, partial (anterior) hippocampectomy, resection of temporal pole plus amygdalohippocampectomy, basal temporal resection plus amygdalohippocampectomy, and removal of a mesial lesion that extends into the lateral temporal lobe. They reported comparable rates of seizure relief with all these approaches. Similarly, Fried et al15 published their results of 41 temporal tumors and stated that seizure-free outcome rate was 87% without any correlation with the extent of hippocampal resection. Another report showed similar (81%) seizure-free outcome with hippocampal-sparing resections in lesional TLE cases.49 Morris et al also published a high rate of seizure-free outcome in their temporal lobe tumor series using a surgical approach that spared hippocampus.50 Therefore, they recommended that hippocampus should not be resected unless it is structurally abnormal. Two other series showed similar seizure control rates without resecting hippocampus of the epilepsy patients with occult vascular malformations in temporal lobe.51,​52 In summary, it is still not clear whether

resection of mesial structures along the lesion is needed for good seizure control in lesional TLE cases. Currently, there is no clear-cut evidence in relevant literature to devise an optimal strategy regarding resection of the mesial structures along the temporal lobe lesions. The presence of radiologically normal-appearing m ­ esial structures along with a temporal lobe lesion presents an additional challenge because even if the mesial structures are radiologically normal appearing, they may still be histologically abnormal and capable of generating independent seizures. This may be related to abnormal synaptic reorganization of hippocampus that may be induced by seizures propagating from temporal neocortical lesions. The critical question is whether the epileptogenic capacity of the ipsilateral radiologically normal-appearing hippocampus would still continue after resection of the temporal tumor. Unfortunately, little data exist to answer this question. Another avenue to assess the ipsilateral radiologically normal-appearing hippocampus is its functional status. Impaired performance on the Wada test correlates well with the extent of hippocampal neuronal loss according to Morioka et al.53 Therefore, if positron emission tomography (PET), neuropsychological, and the Wada test results are suggestive of dysfunction of the ipsilateral mesial temporal structures, then this information should be taken into account when determining the extent of the mesial temporal resection. Conversely, if these test results show well-functioning mesial structures, then mesial structures might be preserved.41

Dual Pathology Dual pathology in TLE cases can be described as coexistence of hippocampal sclerosis with another epileptogenic lesion, such as a tumor or cavernoma. This is not a rare occurrence in epilepsy patients. Although its incidence in the literature varies, Drake et al54 reported a 56% dual pathology rate in their series of pediatric TLE secondary to pediatric temporal lobe tumors. Otsubo et al55 reported dual pathology in 54% of patients with epileptogenic ganglioglioma. Conversely, we found a dual pathology rate of 8% in our pediatric temporal lobe tumor series.11 The presence of dual pathology creates a challenge in surgical strategy of lesional TLE patients similar to the one that we discussed previously. Few studies address this issue. Li et al56 demonstrated that seizure-free outcome rate was significantly low if only one of the lesions was removed in patients with dual pathology, but it was quite high (73%) when both sclerotic hippocampus and the lesion were removed. Morris et al57 also recommend resection of sclerotic hippocampus along with the temporal tumor if imaging studies are correlated with hippocampal sclerosis. In another study, Cascino et al58 reported that lesionectomy alone yielded unsatisfactory results in TLE with a dual pathological entity and recommended resection of hippocampus along with the tumors in these cases. These reports all emphasize the significance of careful evaluation of lesional TLE patients with dual pathology to determine the necessity of amygdalohippocampectomy along with lesion removal. This decision needs to be tailored case by case based on the electrophysiological findings and functional status of the mesial temporal structures.

38  Surgical Management of Lesional Temporal Lobe Epilepsy

Fig. 38.3  Dysembryoplastic neuroepithelial tumors located in (a) mesial temporal and (b) lateral temporal regions.

Location The surgical approaches in lesional TLE cases are closely related to the location of the lesion within temporal lobe and its proximity to mesial temporal structures (Fig. 38.3). The accessibility of the lesion, its relationship to eloquent cortex, and mesial structures are all critical components in determining the surgical approach and extent of the resection. There is no well-­defined, localization-based classification of temporal lobe lesions to describe and to discuss surgical approaches in TLE patients. We prefer to classify temporal lobe lesions, mainly tumors, in three groups based on their location within the temporal lobe as we described previously.11 Here, we will summarize our surgical approaches in these cases based on this classification: yy Mesial temporal lobe tumors: the parahippocampus, amygdala, and hippocampus. yy Basal temporal lobe tumors: basal part of inferior temporal gyrus and fusiform gyrus. yy Lateral temporal lobe tumors: the superior and middle temporal gyri. It is our observation that mesial temporal tumors generally respect pial boundaries of the collateral sulcus and very commonly remain restricted to the mesial temporal region (Fig. 38.4 and Fig. 38.5). The only exception to this is the highgrade, malignant tumor. Therefore, low-grade mesial temporal lobe tumors can be classified separately from temporal neocortical tumors. Conversely, separation between basal and lateral temporal tumors is less precise, and tumors extending to both sides are not rare. However, surgical access to basal temporal tumors has its unique challenges, and it is still helpful to use this classification to define the surgical approach and its rationale more clearly. We define our approach based not only on the tumor location in the temporal lobe but also on being in the dominant or nondominant hemisphere to provide a framework to discussion. It should be emphasized that the location-based description of our surgical approach is defined simply to provide a general perspective to our discussion and certainly should not be considered as a strict guideline. In practicality, we define our surgical approach on a case-by-case

Fig. 38.4  Magnetic resonance image of a mesiotemporal tumor that is entirely confined in the parahippocampus. Note that collateral sulcus (arrow) is well preserved without any tumor involvement in the fusiform gyrus.

basis by considering many factors, including age, availability of detailed electrophysiological data, and feasibility of the Wada and neuropsychological tests. The critical part of the decision-making process in these cases is defining the spatial relationship between the epileptogenic zone and the lesion itself. Invasive monitoring techniques may be needed in some cases to define the epileptogenic zone more clearly. This way, the surgical resection area can be carefully mapped using invasive monitoring, stimulation studies, and functional imaging.

Mesial Temporal Tumors Total removal of the tumor along with the mesial temporal structures is the goal (Fig. 38.5). If the tumor does not involve adjacent mesial structures, such as an amygdala tumor

369

370

IVc  Temporal Lobe Epilepsy and Surgical Approaches

Fig. 38.5  Preoperative (a) and postoperative (b) MRIs of a dysembryoplastic neuroepithelial tumor located in mesiotemporal region, invading both the parahippocampus and hippocampus. The tumor was resected along with mesial temporal structures. Note that the preoperative MRI shows no tumor involvement in the fusiform gyrus despite the large size of tumor and significant enlargement of the parahippocampus. Although displaced, the collateral sulcus (arrow) remains intact and well preserved.

with radiologically normal-appearing hippocampus, we still consider hippocampectomy unless neuropsychological tests suggest a high risk for memory dysfunction. Total removal of the tumor is the goal if the adjacent mesial structures are radiologically and electrophysiologically normal and the neuropsychological tests are normal. If the adjacent mesial structures are not radiologically and electrophysiologically normal, then the Wada and neuropsychological test results become critical in determining the extent of the resection. If the potential risks are acceptable, then resection of the tumor along with the mesial temporal structures is performed.

Basal Temporal Tumors If the epileptogenic zone does not extend to the mesial temporal region, then we perform lesionectomy only. If the epileptogenic zone extends to the mesial temporal region as well, then we resect mesial structures if the potential risks are acceptable based on the Wada and neuropsychological test results.

Lateral Temporal Tumors If the epileptogenic zone does not extend beyond the anatomical boundaries of the lesion, then we perform lesionectomy only. If the epileptogenic zone extends to the surrounding cortical tissue, then tailored resection is performed by including both lesion and epileptogenic area. There are certainly lesions that cannot be classified strictly within these three anatomical subgroups, such as dual pathologies, AVMs, and lesions covering very large areas in the temporal lobe. These are challenging cases, and the surgical approach should be tailored carefully in these patients by considering the histological, radiological, functional, and electrophysiological characteristics of each case. When the neurophysiological and clinical semiology findings are not colocalized with the lesion, further investigation including invasive monitoring should be considered.

„„ Special Considerations in Vascular Malformations Cavernous Hemangioma Adequate seizure control rate with medical management is approximately 60% in cavernous hemangioma-related ­epilepsy, and the surgical indications and approaches in TLE patients with cavernous hemangioma remain controversial.22–​28

If the patient has rare seizures, medical management can be c­onsidered as a first-line treatment. However, location of the lesion and risk of bleeding should also be considered to decide whether surgical management is preferable. If seizures are medically intractable or associated with recurrent bleeding and functional impairment, then surgery should be strongly ­considered.33 Extent of resection and timing of the surgical intervention are main controversial topics in these patients. ­ Some authors discuss that early operation in patients with shorter duration of seizures provides better seizure control and helps to avoid creating secondary epileptogenic foci in h ­ ippocampus.33 However, limited data exist to support this hypothesis. Surgical approaches include lesionectomy limited to the cavernous hemangioma itself, extended lesionectomy (cavernoma and additional removal of perilesional hemosiderin fringe), and tailored temporal lobectomy to remove epileptogenic zone in addition to0 c­ avernoma itself using either intraoperative ECoG or extraoperative monitoring with intracranial electrodes.22,​27 Simple lesionectomy provides seizure-free outcome in 70 to 80% of the patients.22 Complete resection of the cavernoma along with surrounding hemosiderin fringe is the most commonly recommended technique. It is suggested that hemosiderin fringe is responsible for the epileptogenic activity because of the iron content of hemosiderin and induced gliotic changes in surrounding tissues. Therefore, resection of perilesional hemosiderin fringe is critical for good seizure control, according to some authors.30,​34 However, this approach remains controversial, and some study results could not verify this assumption. Although it is not well established, there are data suggesting that the extended resection based on the ECoG provides better seizure control in this patient group.33 Postsurgical failure on seizure control rates were reported between 30 and 70%.25 Englot et al59,​60 analyzed the results of 31 studies that include 1,226 patients with epilepsy secondary to cavernous angiomas; 75% of these patients had Engel class 1 outcome. According to Englot et al, poor seizure outcome predictors are cavernoma size greater than 1.5 cm, presence of multiple lesions, drug resistance, presence of generalized seizures, epilepsy duration more than 1 year, and incomplete lesionectomy. Another topic of discussion is the timing of the surgery. Early surgical intervention for seizure control in cavernous hemangioma patients was correlated with a better postoperative o ­ utcome according to some authors. Stefan and Hammen33 reported 91.7% seizure-free outcome in patients who were operated on within 2 years of seizure onset. Cappabianca et al28 stated that lesionectomy was treatment of choice for patients with a ­ short history of seizures (< 1 year). For longer seizure histories,

38  Surgical Management of Lesional Temporal Lobe Epilepsy ­however, they recommend tailored surgery with neurophysiological investigations to detect possible secondary seizure foci. Another critical issue in surgical strategy is determining whether the patient has dual pathology (i.e., cavernous hemangioma with hippocampal sclerosis). Hammen et al31 found that it is critical to remove both cavernoma and sclerotic hippocampus in dual pathology cases. Stefan and Hammen33 also emphasized the importance of assessment for dual pathology. If there is no discordance between location of lesion and clinical and electrophysiological features of seizures, then simple lesionectomy might be a reasonable approach in cavernomas located in the temporal lobe. We prefer lesionectomy with removal of the surrounding hemosiderin fringe as the first step in cavernous hemangioma patients (Fig. 38.6). If the seizures persist, then more rigorous investigations including invasive monitoring can be performed as second step. Cavernous angiomas located in temporal lobe should be assessed carefully from anatomo-clinico-physiological standpoint, and invasive monitoring before removal of cavernoma should be considered if there is poor concordance between the location of the ­cavernoma and the electrophysiological and clinical semiology data. We also believe that removal of the ipsilateral mesial structures is the correct approach if the cavernoma is located in the mesial temporal lobe, especially if there are abnormal imaging findings at mesial temporal structures on MRI and/or seizures are long-standing and medically refractory.

Arteriovenous Malformations AVM-related seizures are frequently well controlled with medical management.34 Radiosurgery is also very effective for ­seizure control in many AVM patients.20,​61 Therefore, the surgi-

cal treatment decision for AVMs is usually related to intracerebral bleeding or risk of future hemorrhage. Conversely, data in the literature are limited regarding the natural course of seizures in AVM patients because the main focus of AVM literature is mostly vascular events and their surgical management. The main goal in cerebral AVM surgery is removal of the AVM nidus without disruption or damage to the surrounding brain tissue. There are some data in the literature suggesting that AVM resection not only provides protection from potentially catastrophic hemorrhage but also helps to control the seizures.19 Yeh et al62,​63 have published two studies on the surgical management of seizures in AVM patients. They recommend intraoperative ECoG covering temporal neocortex and mesial structures if the seizures involve temporal lobe in AVM patients. Using this technique, they performed additional cortical excisions, based on the ECoG data, including resection of mesial temporal structures in 9 of 17 patients with temporal lobe AVMs with good seizure control rates. In another study, Turjman et al identified several characteristics of AVMs that correlate with a high risk of epilepsy as the initial presentation.19 These features are presence of a superficial nidus close to the cortex, temporoparietal location, and feeders from the middle cerebral artery. The main questions regarding AVM-related epilepsy are whether the surrounding brain tissue and gliotic areas around the AVM have independent epileptogenic properties, rate of secondary epileptogenesis of mesial temporal structures in temporal lobe AVMs, appropriate diagnostic techniques to determine independent epileptogenic activities, and right approach to improve postoperative seizure control in AVM patients. Unfortunately, little data exist in the literature to answer these questions. There are not even satisfactory data to determine whether the extent of resection

Fig. 38.6  (a) Preoperative coronal MRI of a 12-year-old patient shows a cavernous hemangioma located in superior temporal gyrus. (b–d) Intraoperative pictures of the same patient: surgical exposure, dissection, and removal of cavernous hemangioma and surgical field after resection of the hemosiderin fringe on the cavity walls. (e) Postoperative coronal MRI shows residual surgical cavity.

371

372

IVc  Temporal Lobe Epilepsy and Surgical Approaches beyond the gliotic margins of an AVM has any influence over postoperative seizure control.42 These critical questions on AVM-related seizures deserve more detailed investigation.

„„ Outcome Tumors Surgical outcome in tumor-related pediatric TLE cases is much better than some other lesional epilepsy cases, such as cortical dysplasia. In a large pediatric series with TLE patients secondary to neoplasm, Engel classes 1 and 2 outcome rates were reported as high as 88%. This rate was 70% in the MTS group and 50% in the patients with cortical dysplasia in the same series.4 In another pediatric series with 120 patients, Engel class 1 outcome was reported as 93.3%.14 This study also reported that shorter duration of epilepsy and temporal lobe localization were positive predictors for good outcome.14 Choi et al64 reported class 1 seizure-free outcome in 77% of the patients with tumor-­related TLE. They noted that complete resection of the tumor was the most significant factor in obtaining a seizure-free outcome. Other authors have reported seizure-free outcomes in 75 to 95% of patients with tumor-related epilepsy.3,​5,​65 The main predictor of seizure-free outcome is complete resection of the tumor according to some studies.12,​44 Another predictor is histological type of the tumor. Luyken et al66 noted that seizure-free outcome was highest in patients with ganglioglioma and oligodendroglioma (> 90%) compared to pilocytic and grade 2 astrocytomas (61 and 66%, respectively). The same study revealed clinical and radiological predictors for poorer seizure control as having a longer history of seizures, additional EEG focus, dual pathology, and incomplete tumor resection. Clusmann et al.2 published their results in 89 children with TLE and 46.6% of these patients had neoplastic lesions. They reported good outcome (Engel classes 1 and 2) in 83.3% of ganglioglioma and DNET patients. The location of the tumor within the temporal lobe also appears to have a significant effect on the outcome. Giulioni et al67 reported the results of 21 tumor-related (ganglioglioma) TLE patients who all underwent lesionectomy only. In this series, 66.6 and 33.3% of the patients, respectively, had Engel classes 1 and 2 outcomes, respectively. Although the seizure-free outcome rate was 100% in temporal neocortical tumors, this rate was only 60% in temporomesial tumors in this series. This was despite gross total tumor resection in 80% of patients with temporomesial tumor. Therefore, Giulioni et al67 suggested that further neurophysiological assessment should be performed to define and resect the additional epileptogenic zone in patients with temporomesial ganglioglioma. Again, Giulioni et al68 reported their experience with glioneuronal tumors associated with epilepsy in children, and seizure-free outcome rate was 86.6%. Interestingly, they did not find any difference between lateral temporal and mesial temporal tumors in terms of the postoperative outcomes. They did not observe an effect of the duration of seizure history on the postoperative seizure outcomes. They also reported a high rate (40%) of associated dysplastic cortex among these patients. Surgical approach is another factor that might potentially affect clinical outcome. Unfortunately, the data comparing the results of different surgical techniques are very limited.

Chan et al69 compared the results of two different surgical approaches in patients with DNET. They performed tumor resection with temporal lobectomy in 12 patients and only lesionectomy in 6 patients. Although all patients with temporal lobectomy had seizure-free outcomes, this rate was only 33% in patients who underwent lesionectomy alone. Chan et al69 explained this difference with the presence of dysplastic cortex around DNET lesions. In another study, Minkin et al70 reported 24 children with DNETs. A total of 15 lesions were located in the temporal lobe; 10 of these were in the lateral temporal, and 5 were in the mesial temporal lobe. Overall, 83.3% of the patients had seizure-free outcome. They performed lesionectomy alone in 73% of the patients and lesionectomy plus amygdalohippocampectomy in the remaining patients. All of the patients who had lesionectomy plus amygdalohippocampectomy had seizure-free outcome. Therefore, Minkin et al stressed the significance of amygdalohippocampectomy in the treatment of temporal DNETs and recommended extensive presurgical evaluation in this patient group. However, all these studies are retrospective assessments, most include a relatively small number of patients, and none is specifically designed to compare the effect of different surgical approaches on outcome. In another study, it was found that tailored resections provided better results than gross total or subtotal resections.59,​60 The rate of seizure-free outcome was 43% after subtotal resection, 79% after gross total resection, and 87% after gross total resection plus hippocampectomy plus corticectomy in this study. They stated that cortical dysplasia, gliosis, and hippocampal sclerosis associated with neoplasms play significant role on continuing seizures even after gross total excision of the tumor, therefore they see tailored resection of temporal lobe tumors as a key factor in achieving seizure control.59 “Lesionectomy plus” approach should be considered especially in cases with abnormal imaging findings and/or electrical activity in mesial temporal structures and the cases with dual pathology.71

Cavernous Hemangioma Postoperative seizure control rates are greater than 70% in most of the surgical cavernous hemangioma series.29 Baumann et al30,​72 reported Engel class 1 outcome in 70% of cavernous hemangioma patients in a series that included both adults and children. In contrast to other studies, Baumann et al30,​72 did not find that longer seizure duration is a predictor for poorer outcome in cavernous hemangioma patients. They reported favorable outcome predictors as mesiotemporal location, removing hemosiderin-stained brain tissue along the cavernous hemangioma, being older than 30 years of age, having no additional seizure foci, and no secondarily generalized seizures. In another study, Fortuna et al73 reviewed 56 pediatric cavernous hemangioma patients and reported seizure-free outcome in 73.2% of them and significant improvement in an additional 19.6% of the patients. Two other studies reported, respectively, 82.9 and 84% seizure-free outcome postoperatively in cavernous hemangioma patients.24,​28

Arteriovenous Malformation As mentioned previously, the goal in the surgical management of AVMs is the complete removal of the AVM nidus. Whether

38  Surgical Management of Lesional Temporal Lobe Epilepsy this is sufficient to have good postoperative seizure control is not clear. Again, the benefit of resecting the surrounding brain parenchyma and gliotic tissue or ipsilateral mesial temporal structures is not clear from a seizure-control standpoint. Data to answer these questions clearly are very limited. In one of the studies addressing this question, Yeh et al62,​63 documented remote seizure foci in mesial temporal region using depth electrodes in temporal AVM patients. They then resected mesial structures in addition to AVM nidus in 67% of the patients to control seizures. They reported excellent seizure control in 78% of the patients. Another large series was published by Hoh et al.74 They reported 141 patients with brain AVMs who had multimodality treatment. In this series, 33% of the patients had seizures before the treatment. They found that large size (> 3 cm) of the AVM and temporal location were associated with seizures more frequently. The seizure-free (Engel class 1) outcome rate was 66%, and class 2 outcome was 10% after surgical treatment. Yasargil21 reported a postoperative seizure-free outcome without any medication in all patients (n = 12) with hippocampal AVMs. This rate was 18% in extramesial temporal AVM patients, with an additional 56% with good seizure control with medications. Gerszten et al75 reported 72 children with AVM who were treated with Gamma Knife radiosurgery. ­Twenty-one percent had seizures, and 85% became seizure-free after the Gamma Knife treatment.

„„ Conclusion Low-grade tumors are one of the most common pathological substrates in TLE during childhood. No consensus in the surgical management of lesion-related TLE has been established to date. Although there is consensus on critical significance of complete lesion resection, the importance of additional resection of adjacent cortex or the mesial temporal structures remains controversial. There are strong implications in the published series on dual pathology lesions that lesionectomy alone without removal of the sclerotic hippocampus may not yield good seizure control rates. Further investigation is needed in these patients to determine whether the resection of both the lesion and sclerotic hippocampus is indicated for better outcomes. Data are very limited regarding AVM-related TLE and the most effective surgical approach in its management. Lesionectomy is the most commonly used technique in the surgical management of intractable TLE in cavernous hemangioma patients, and a significant number of studies suggest that removal of surrounding hemosiderin fringe in addition to lesionectomy would further increase seizure control rates in these patients.

References 1. Sinclair DB, Wheatley M, Aronyk K, et al. Pathology and neuroimaging in pediatric temporal lobectomy for intractable epilepsy. Pediatr Neurosurg 2001;35(5):239–246 2. Clusmann H, Kral T, Fackeldey E, et al. Lesional mesial temporal lobe epilepsy and limited resections: prognostic factors and outcome. J Neurol Neurosurg Psychiatry 2004;75(11):1589–1596 3. Mittal S, Montes JL, Farmer JP, et al. Long-term outcome after surgical treatment of temporal lobe epilepsy in children. J Neurosurg 2005;103(5, Suppl):401–412 4. Benifla M, Otsubo H, Ochi A, et al. Temporal lobe surgery for intractable epilepsy in children: an analysis of outcomes in 126 children. Neurosurgery 2006;59(6):1203–1213, discussion 1213–1214 5. Kan P, Van Orman C, Kestle JRW. Outcomes after surgery for focal epilepsy in children. Childs Nerv Syst 2008;24(5):587–591 6. Kim SK, Wang KC, Hwang YS, et al. Epilepsy surgery in children: outcomes and complications. J Neurosurg Pediatr 2008;1(4):277–283 7. Maton B, Jayakar P, Resnick T, Morrison G, Ragheb J, Duchowny M. Surgery for medically intractable temporal lobe epilepsy during early life. Epilepsia 2008;49(1):80–87 8. Hennessy MJ, Elwes RDC, Honavar M, Rabe-Hesketh S, Binnie CD, Polkey CE. Predictors of outcome and pathological considerations in the surgical treatment of intractable epilepsy associated with temporal lobe lesions. J Neurol Neurosurg Psychiatry 2001;70(4):450–458 9. Bronen RA, Fulbright RK, Spencer DD, Spencer SS, Kim JH, Lange RC. MR characteristics of neoplasms and vascular malformations associated with epilepsy. Magn Reson Imaging 1995;13(8):1153–1162 10. Uliel-Sibony S, Kramer U, Fried I, Fattal-Valevski A, Constantini S. Pediatric temporal low-grade glial tumors: epilepsy outcome following resection in 48 children. Childs Nerv Syst 2011;27(9):1413–1418 11. Çataltepe O, Turanli G, Yalnizoglu D, Topçu M, Akalan N. Surgical management of temporal lobe tumor-related epilepsy in children. J Neurosurg 2005;102(3, Suppl):280–287

12. Englot DJ, Chang EF, Vecht CJ. Epilepsy and brain tumors. Handb Clin Neurol 2016;134:267–285 13. Zaatreh MM, Firlik KS, Spencer DD, Spencer SS. Temporal lobe tumoral epilepsy: characteristics and predictors of surgical outcome. Neurology 2003;61(5):636–641 14. Pelliccia V, Deleo F, Gozzo F, et al. Early and late epilepsy surgery in focal epilepsies associated with long-term epilepsy-associated tumors. J Neurosurg 2017;127(5):1147–1152 (published online) 15. Fried I, Kim JH, Spencer DD. Limbic and neocortical gliomas associated with intractable seizures: a distinct clinicopathological group. Neurosurgery 1994;34(5):815–823, discussion 823–824 16. Giulioni M, Rubboli G, Marucci G, et al. Seizure outcome of epilepsy surgery in focal epilepsies associated with temporomesial glioneuronal tumors: lesionectomy compared with tailored resection. J Neurosurg 2009;111(6):1275–1282 17. Horgan MA, Florman J, Spetzler RF. Surgical management of AVM in children. In: Alexander MJ, Spetzler RF, eds. Pediatric Neurovascular Disease. New York, NY: Thieme; 2006:104–115 18. Menovsky T, van Overbeeke JJ. Cerebral arteriovenous malformations in childhood: state of the art with special reference to treatment. Eur J Pediatr 1997;156(10):741–746 19. Turjman F, Massoud TF, Sayre JW, Viñuela F, Guglielmi G, Duckwiler G. Epilepsy associated with cerebral arteriovenous malformations: a multivariate analysis of angioarchitectural ­characteristics. AJNR Am J Neuroradiol 1995;16(2):345–350 20. Schäuble B, Cascino GD, Pollock BE, et al. Seizure outcomes after stereotactic radiosurgery for cerebral arteriovenous malformations. Neurology 2004;63(4):683–687 21. Yasargil MG. Microneurosurgery. New York, NY: Thieme; 1988: 13–23, 11–136, 393–396 22. Cosgrove GR. Occult vascular malformations and seizures. Neurosurg Clin N Am 1999;10(3):527–535 23. Mottolese C, Hermier M, Stan H, et al. Central nervous system cavernomas in the pediatric age group. Neurosurg Rev 2001;24 (2–3):55–71, discussion 72–73

373

374

IVc  Temporal Lobe Epilepsy and Surgical Approaches 24. Moran NF, Fish DR, Kitchen N, Shorvon S, Kendall BE, Stevens JM. Supratentorial cavernous haemangiomas and epilepsy: a review of the literature and case series. J Neurol Neurosurg Psychiatry 1999;66(5):561–568 25. Sevy A, Gavaret M, Trebuchon A, et al. Beyond the lesion: the epileptogenic networks around cavernous angiomas. Epilepsy Res 2014;108(4):701–708 26. Jehi LE, Palmini A, Aryal U, Coras R, Paglioli E. Cerebral cavernous malformations in the setting of focal epilepsies: pathological findings, clinical characteristics, and surgical treatment principles. Acta Neuropathol 2014;128(1):55–65 27. Cossu M, Raneri F, Casaceli G, Gozzo F, Pelliccia V, Lo Russo G. Surgical treatment of cavernoma-related epilepsy. J Neurosurg Sci 2015;59(3):237–253 28. Cappabianca P, Alfieri A, Maiuri F, Mariniello G, Cirillo S, de Divitiis E. Supratentorial cavernous malformations and epilepsy: seizure outcome after lesionectomy on a series of 35 patients. Clin Neurol Neurosurg 1997;99(3):179–183 29. Yeh D, Crone KR. Cavernous malformations in children. In: ­Alexander MJ, Spetzler RF, eds. Pediatric Neurovascular Disease. New York, NY: Thieme; 2006:65–71 30. Baumann CR, Acciarri N, Bertalanffy H, et al. Seizure outcome after resection of supratentorial cavernous malformations: a study of 168 patients. Epilepsia 2007;48(3):559–563 31. Hammen T, Romstöck J, Dörfler A, Kerling F, Buchfelder M, Stefan H. Prediction of postoperative outcome with special respect to removal of hemosiderin fringe: a study in patients with cavernous haemangiomas associated with symptomatic epilepsy. Seizure 2007;16(3):248–253 32. Requena I, Arias M, López-Ibor L, et al. Cavernomas of the central nervous system: clinical and neuroimaging manifestations in 47 patients. J Neurol Neurosurg Psychiatry 1991;54(7):590–594 33. Stefan H, Hammen T. Cavernous haemangiomas, epilepsy and treatment strategies. Acta Neurol Scand 2004;110(6):393–397 34. Kraemer DL, Awad IA. Vascular malformations and epilepsy: clinical considerations and basic mechanisms. Epilepsia 1994;35(Suppl 6):S30–S43 35. Giulioni M, Acciarri N, Padovani R, Galassi E. Results of surgery in children with cerebral cavernous angiomas causing epilepsy. Br J Neurosurg 1995;9(2):135–141 36. Di Rocco C, Iannelli A, Tamburrini G. Cavernomas of the CNS in children. A report of 22 cases. Acta Neurochir (Wien) 1996;138:1267–1274, discussion 1273–1274 37. Fish DR. How do tumors cause epilepsy? In: Kotagal P, Luders HO, eds. The Epilepsies: Etiologies and Prevention. San Diego, CA: ­Academic Press; 1999:301–314 38. Bartolomei JC, Christopher S, Vives K, Spencer DD, Piepmeier JM. Low-grade gliomas of chronic epilepsy: a distinct clinical and pathological entity. J Neurooncol 1997;34(1):79–84 39. Çataltepe O, Comair YG. Strategies in operating on patients with tumor-related epilepsy. In: Kotagal P, Luders HO, eds. The Epilepsies: Etiologies and Prevention. San Diego, CA: Academic Press; 1999:365–370 40. Staley K. Molecular mechanisms of epilepsy. Nat Neurosci 2015;18(3):367–372 41. Usui N, Mihara T, Baba K, et al. Intracranial EEG findings in patients with lesional lateral temporal lobe epilepsy. Epilepsy Res 2008;78(1):82–91 42. Awad IA, Rosenfeld J, Ahl J, Hahn JF, Lüders H. Intractable epilepsy and structural lesions of the brain: mapping, resection strategies, and seizure outcome. Epilepsia 1991;32(2):179–186 43. Van Ness PC. Pros and cons of lesionectomy as treatment for partial epilepsy. In: Kotagal P, Luders HO, eds. The Epilepsies: Etiologies and Prevention. San Diego, CA: Academic Press; 1999:391–397 44. Khajavi K, Comair YG, Wyllie E, Palmer J, Morris HH, Hahn JF. Surgical management of pediatric tumor-associated epilepsy. J Child Neurol 1999;14(1):15–25 45. Jooma R, Yeh HS, Privitera MD, Gartner M. Lesionectomy versus electrophysiologically guided resection for temporal lobe

tumors manifesting with complex partial seizures. J Neurosurg 1995;83(2):231–236 46. Sugano H, Shimizu H, Sunaga S. Efficacy of intraoperative electrocorticography for assessing seizure outcomes in intractable epilepsy patients with temporal-lobe-mass lesions. Seizure 2007;16(2):120–127 47. Mathern GW, Babb TL, Pretorius JK, Melendez M, Lévesque MF. The pathophysiologic relationships between lesion pathology, intracranial ictal EEG onsets, and hippocampal neuron losses in temporal lobe epilepsy. Epilepsy Res 1995;21(2):133–147 48. Mihara T, Baba MK. Combined use of subdural and depth electrodes. In: Luders H, Comair Y, eds. Epilepsy Surgery. 2nd ed. ­Philadelphia, PA: Lippincott Williams & Wilkins; 2001 49. Lee SK, Lee SY, Kim KK, Hong KS, Lee DS, Chung CK. Surgical outcome and prognostic factors of cryptogenic neocortical epilepsy. Ann Neurol 2005;58(4):525–532 50. Morris HH, Estes ML, Gilmore R, Van Ness PC, Barnett GH, Turnbull J. Chronic intractable epilepsy as the only symptom of primary brain tumor. Epilepsia 1993;34(6):1038–1043 51. Cohen DS, Zubay GP, Goodman RR. Seizure outcome after lesionectomy for cavernous malformations. J Neurosurg 1995;83(2): 237–242 52. Kraemer DL, Griebel ML, Lee N, Friedman AH, Radtke RA. Surgical outcome in patients with epilepsy with occult vascular malformations treated with lesionectomy. Epilepsia 1998;39(6):600–607 53. Morioka T, Hashiguchi K, Nagata S, et al. Additional hippocampectomy in the surgical management of intractable temporal lobe epilepsy associated with glioneuronal tumor. Neurol Res 2007;29(8):807–815 54. Drake J, Hoffman HJ, Kobayashi J, Hwang P, Becker LE. Surgical management of children with temporal lobe epilepsy and mass lesions. Neurosurgery 1987;21(6):792–797 55. Otsubo H, Hoffman HJ, Humphreys RP, et al. Detection and management of gangliogliomas in children. Surg Neurol 1992;38(5):371–378 56. Li LM, Cendes F, Andermann F, et al. Surgical outcome in patients with epilepsy and dual pathology. Brain 1999;122(Pt 5):799–805 57. Morris HH, Matkovic Z, Estes ML, et al. Ganglioglioma and intractable epilepsy: clinical and neurophysiologic features and predictors of outcome after surgery. Epilepsia 1998;39(3):307–313 58. Cascino GD, Jack CR Jr, Parisi JE, et al. Operative strategy in patients with MRI-identified dual pathology and temporal lobe epilepsy. Epilepsy Res 1993;14(2):175–182 59. Englot DJ, Berger MS, Barbaro NM, Chang EF. Factors associated with seizure freedom in the surgical resection of glioneuronal tumors. Epilepsia 2012;53(1):51–57 60. Englot DJ, Han SJ, Berger MS, Barbaro NM, Chang EF. Extent of surgical resection predicts seizure freedom in low-grade temporal lobe brain tumors. Neurosurgery 2012;70(4):921–928, discussion 928 61. Trussart V, Berry I, Manelfe C, Arrue P, Castan P. Epileptogenic cerebral vascular malformations and MRI. J Neuroradiol 1989;16(4):273–284 62. Yeh HS, Kashiwagi S, Tew JM Jr, Berger TS. Surgical management of epilepsy associated with cerebral arteriovenous malformations. J Neurosurg 1990;72(2):216–223 63. Yeh HS, Tew JM Jr, Gartner M. Seizure control after surgery on cerebral arteriovenous malformations. J Neurosurg 1993;78(1):12–18 64. Choi JY, Chang JW, Park YG, Kim TS, Lee BI, Chung SS. A retrospective study of the clinical outcomes and significant variables in the surgical treatment of temporal lobe tumor associated with intractable seizures. Stereotact Funct Neurosurg 2004;82(1):35–42 65. Kim SK, Wang KC, Hwang YS, Kim KJ, Cho BK. Intractable epilepsy associated with brain tumors in children: surgical modality and outcome. Childs Nerv Syst 2001;17(8):445–452 66. Luyken C, Blümcke I, Fimmers R, et al. The spectrum of long-term epilepsy-associated tumors: long-term seizure and tumor outcome and neurosurgical aspects. Epilepsia 2003;44(6):822–830

38  Surgical Management of Lesional Temporal Lobe Epilepsy 67. Giulioni M, Gardella E, Rubboli G, et al. Lesionectomy in epileptogenic gangliogliomas: seizure outcome and surgical results. J Clin Neurosci 2006;13(5):529–535 68. Giulioni M, Galassi E, Zucchelli M, Volpi L. Seizure outcome of lesionectomy in glioneuronal tumors associated with epilepsy in children. J Neurosurg 2005;102(3, Suppl):288–293 69. Chan CH, Bittar RG, Davis GA, Kalnins RM, Fabinyi GCA. Longterm seizure outcome following surgery for dysembryoplastic neuroepithelial tumor. J Neurosurg 2006;104(1):62–69 70. Minkin K, Klein O, Mancini J, Lena G. Surgical strategies and seizure control in pediatric patients with dysembryoplastic neuroepithelial tumors: a single-institution experience. J Neurosurg Pediatr 2008;1(3):206–210 71. Tandon N, Esquenazi Y. Resection strategies in tumoral epilepsy: is a lesionectomy enough? Epilepsia 2013;54(Suppl 9):72–78

72. Baumann CR, Schuknecht B, Lo Russo G, et al. Seizure outcome after resection of cavernous malformations is better when surrounding hemosiderin-stained brain also is removed. Epilepsia 2006;47(3):563–566 73. Fortuna A, Ferrante L, Mastronardi L, Acqui M, d’Addetta R. Cerebral cavernous angioma in children. Childs Nerv Syst 1989;5(4):201–207 74. Hoh BL, Chapman PH, Loeffler JS, Carter BS, Ogilvy CS. Results of multimodality treatment for 141 patients with brain arteriovenous malformations and seizures: factors associated with seizure incidence and seizure outcomes. Neurosurgery 2002;51(2):303– 309, discussion 309–311 75. Gerszten PC, Adelson PD, Kondziolka D, Flickinger JC, Lunsford LD. Seizure outcome in children treated for arteriovenous malformations using gamma knife radiosurgery. Pediatr Neurosurg 1996;24(3):139–144

375

39

  Surgical Management of MRI-Negative Temporal Lobe Epilepsy Francesco Cardinale, Piergiorgio d’Orio, and Michele Rizzi

Summary MRI-negative patients with epilepsy onset in pediatric age is rare especially for the temporal lobe. Stereoelectroencephalography (SSEG) monitoring is recommended to help localize the seizure focus. When surgery is performed, nearly 50% of children are free of disabling seizures (Engel class 1). Histological examination typically is not helpful in these cases although there are reports of cortical dysplasia and occasional hamartomas. Keywords:  pediatric epilepsy surgery, MRI-negative, temporal lobe epilepsy, stereoelectroencephalography, intracerebral electrodes, invasive recording, surgical planning, multimodal imaging, outcome predictors, seizure outcome

„„ Introduction Surgical intervention aiming to remove the epileptogenic zone (EZ) is an effective treatment option in managing drug-resistant epilepsy.1 EZ can be defined as “the site of the beginning and of primary organization of the epileptic seizures.”2 The presence of a lesion visible on MRI is a positive predictor for postoperative seizure freedom in both adult and pediatric populations.3,​4 However, there are many children with refractory epilepsy and no MRI-visible lesion. These are difficult cases, and key point for good outcome in these cases is determining the a ­ ppropriate presurgical strategy. Once the noninvasive workup has been completed, invasive recordings are often necessary for the delineation of the EZ in such challenging cases. At our center, we adopt the SEEG methodology since 1996 according to the basic principles of Talairach and Bancaud,5,​6 even though the technical workflow has been largely updated throughout the years. When long-term video-SEEG monitoring leads to the EZ definition, then surgical resection is indicated. In this chapter, we will illustrate the typical workflow adopted at the “Claudio Munari” Center for Epilepsy Surgery, Milan, by mainly focusing on the surgical population of MRI-negative patients with temporal lobe epilepsy (TLE) onset in pediatric or adolescent age. We will also include the description of an illustrative case.

„„ Patients and Methods General Workflow at the “Claudio Munari” Center The general workflow at the “Claudio Munari” center includes a noninvasive presurgical workup, an optional invasive investigation by means of SEEG electrodes and a final surgical intervention for resection or disconnection of the EZ.

Noninvasive Presurgical Workup Our comprehensive presurgical workup for all patients includes obtaining the medical history of the family and patient with particular attention to subjective and objective ictal phenomena, neurological examination, neuropsychological testing, interictal scalp electroencephalography (EEG), and brain MRI. All patients are scanned using an Achieva 1.5-T magnet with an eight-channel coil and SENSE technology (Philips Healthcare, Best, The Netherlands). The diagnostic protocol includes a three-dimensional (3D) volume fast-field-echo (FFE) T1-weighted (T1W) sequence (contiguous axial slices with 560 × 560 reconstruction matrix; 0.46 × 0.46 × 0.9 mm voxel; no interslice gap; time of repetition [TR], 7.3 ms; time of echo [TE], 3.3 ms); axial, coronal, and sagittal turbo-spin-echo (TSE) fluid attenuated inversion recovery (FLAIR) sequences (axial, coronal, and sagittal slices with 288 × 288 reconstruction matrix; 0.87 × 0.87 × 3 mm voxel; 0.6-mm interslice gap; TR, 11 ms; TE, 140 ms); axial and coronal TSE T2-weighted (T2W) sequences (axial and coronal slices with 512 × 512 reconstruction matrix; 0.45 × 0.45 × 3 mm voxel; 1-mm interslice gap; TR, 6.5 ms; TE, 100 ms); a coronal TSE inversion recovery (IR) T1W sequence (coronal slices with 512 × 512 reconstruction matrix; 0.45 × 0.45 × 3 mm voxel; 0.3-mm interslice gap; TR, 3.4 ms; TE, 15 ms). Only when a tumor is suspected, a 3D FFE T1W scan is obtained with gadolinium enhancement. For patients who are not undergoing SEEG investigation, major brain veins are imaged by means of an axial venous phase contrast angiography (PCA) sequence (contiguous axial slices with 256 × 256 reconstruction matrix; 0.9 × 0.9 × 1.6 mm voxel; 0.8-mm interslice overlapping; TR, 0,0015

39  Surgical Management of MRI-Negative Temporal Lobe Epilepsy ms; TE, 6.7 ms). Recently, we also added a 3D-volume TSE FLAIR sequence (contiguous axial slices with 320 × 320 reconstruction matrix; 0.78 × 0.78 × 1 mm voxel; no interslice gap; TR, 8 ms; TE, 346 ms).7 When a temporal seizure origin is suspected, images are oriented parallel (axial) or perpendicular (coronal) to the longest hippocampal axis. Functional MRI (fMRI) and diffusion-weighted images (DWI) are obtained in order to map cortical functional activations and reconstruct the main white matter bundles, respectively. Special techniques for image processing can be performed to help identify a lesion, such as SUrface PRojected–FLAIR (SUPR-FLAIR) analysis.8 It is a surface-based analysis aimed at comparing the normalized cortical intensity values of the patient’s FLAIR scan against comparable images from a number of healthy controls. The ability to detect a lesion is increased with the processing analysis, even when no obvious abnormalities are visible on conventional MRI investigation. The dynamic evaluation of mapped P values on a template hemisphere indicates where FLAIR intensity values are higher than in controls (the peak of the first positive lowest P-value cluster [PLPC]), also providing information about other cortical areas that could be considered physiologically or pathologically (epileptogenic network) connected. Brain [18F] fluorodeoxyglucose positron emission tomography (FDG-PET) is available at our center since 2010, unfortunately it is feasible only for cooperative patients. Therefore, younger children who would need a general anesthesia are excluded from this diagnostic modality. Interictal PET scan is obtained with an integrated PET or CT camera, Biograph TruePoint (Siemens Healthcare, Erlangen, Germany). The dataset is made of 148 axial slices with 336 × 336 reconstruction matrix, 1 × 1 × 3 mm voxel. FDG-PET is acquired, when possible, in most MRI-negative cases. When indicated, video-EEG long-term monitoring is performed with the goal of recording at least one habitual electroclinical seizure. In order to facilitate the ictal recording in c­ hildren who do not have any seizures, sleep deprivation or tapering of antiepileptic agents prior to admission is r­ ecommended.

Stereoelectroencephalography As previously suggested by Cossu et al in 2008, SEEG is indicated when noninvasive investigations fail to satisfactorily localize the EZ: (1) with a negative MRI, ictal or interictal scalp EEG findings are incongruous with ictal clinical semiology; (2) with a focal MRI abnormality, electroclinical findings suggest a wide involvement also of extralesional areas; (3) ictal clinical semiology is discordant with an apparently localizing ictal scalp EEG pattern (irrespective of MRI evidence); (4) with large/diffuse/ hemispheric/multifocal/bilateral MRI abnormalities, electroclinical evidences suggest a localized or lateralized ictal onset; (5) anatomical and/or electroclinical involvement of highly eloquent areas. In the latter instance, functional mapping ­conducted by intracerebral electrical stimulations (IES) allows identification of both eloquent cortex and subcortical critical bundles. Depending on the child’s cooperation, IES are employed to identify primary motor, somatosensory, visual, and speech areas.9 Our policy for the implantation of SEEG electrodes has been detailed in previous publications.10,​11,​12 Briefly, the electrode trajectories are planned on multimodal images integrating all ­relevant above-mentioned image datasets, plus images aimed

at visualizing and reconstructing the brain vasculature three-­ dimensionally. 3D cone-beam CT digital subtraction angiography (3D CBCT DSA) is obtained processing the images acquired by means of the O-arm (Medtronic; Minneapolis, Minnesota, MN), a mobile CBCT scanner. Schematically, two image datasets are obtained with a reconstructed 3D volume of a 200 × 150 mm cylinder, described by a series of 12-bit Digital Imaging and COmmunications in Medicine (DICOM) files (192 axial slices, 512 × 512 matrix, 0.415 × 0.415 × 0.833 mm anisotropic ­voxels). The first dataset (bone mask) is registered to and subtracted from the subsequent one, obtained while injecting iodinated contrast medium during an otherwise standard catheter angiography procedure. All the processed datasets are co-registered and processed mainly by means of one of the software programs provided by FSL,13 Freesurfer,14 and 3D Slicer,15 three of the most popular open-source software packages. Segmented vessels, ­Freesurfer-derived segmented hemispheres and all other useful data are imported in Voxim (IVS, Chemnitz, Germany), the stereotactic software package supplied with the robotic system. Entry points (EPs) and target points (TPs) are manually defined for every trajectory by looking at multiplanar reconstructions, brain and vascular surface rendering, and images reformatted according to the planned vector of the trajectory. The electrodes (Microdeep Intracerebral Electrodes-D08; Dixi ­Medical, Besançon, France; or Depth Electrodes Range 2069; Alcis, Besançon, France) are implanted under general anesthesia. Once the patient is registered, the robotic system (­ Neuromate, Renishaw-Mayfield SA, Morges, Switzerland) aligns the tool holder along the vectors of each planned trajectory to assist the anchor bolt fixation to the skull. Subsequently, the frame is removed, and the electrodes are advanced into the brain under radioscopic control (Fig. 39.1). Finally, a postimplantation 3D CT is obtained with the O-arm for assessing the final ­positioning of the devices. Video-SEEG monitoring is started the day following the electrode implantation, with the aim of recording at least one habitual seizure. The definition of the EZ is also made based on the results of intracerebral stimulations, which can induce seizures and allow to map the eloquent structures in both the gray and white matter.16,​17 At the end of the video-SEEG monitoring period, radiofrequency thermocoagulation (rf-THC) can be also performed, when indicated.18,​19 This technique has been available at our center since 2008.

Surgical Planning The results of the presurgical work-up are discussed in a multidisciplinary conference with epileptologists, neurosurgeons, neuroradiologists, and neuropsychologists. The results of ­ noninvasive investigations, along with the SEEG findings (if available) are reviewed and discussed in order to plan the topography of the EZ resection or disconnection. The multimodal integration of all available data in a scene assembled with 3D Slicer is of great help for such discussions and for the final planning of the surgery (Fig. 39.2). The surgical plan is therefore the result of a combination of factors, including both epileptological and anatomical information. Once the surgical plan has been completed, the patient is operated on with the assistance of the S7 neuronavigation

377

378

IVc  Temporal Lobe Epilepsy and Surgical Approaches

Fig. 39.1  Implantation of SEEG electrodes. (a) Robotic-assisted twist drilling aimed at fixing the guiding screws. (b) The electrodes are advanced into the brain under X-ray control.

and/or disconnecting the EZ. We searched our institutional database and clinical charts for patients fulfilling the following criteria: • yy yy yy

Age younger than 18 years at epilepsy onset. Absence of structural lesions at visual inspection of MRI. Brain resection limited to the temporal lobe. Postoperative follow-up of at least 12 months.

Statistical Analysis

Fig. 39.2  A multimodal scene built with 3D Slicer, including the Freesurfer-derived 3D reconstruction of the pial surface, as well as the models of the brain vasculature and SEEG electrodes obtained with the 3D CBCT DSA modality.

s­ ystem (Medtronic; Minneapolis, Minnesota, MN). Resections are performed mainly with the use of the subpial technique and/or ultrasound aspiration. Great care is reserved to the study and preservation of the regional vascular pattern (Fig. 39.3).

MRI-Negative Temporal Lobe Epilepsy Case Series at “Claudio Munari” Epilepsy Surgery Center Here, we will discuss MRI-negative patients who had TLE with an onset during childhood or adolescence years and have been treated at “Claudio Munari” Epilepsy Surgery Center within the last 20 years. We analyzed the demographics, clinical characteristics, presurgical assessment data, surgical treat­ ment, and outcome of this patient group.

Patient Selection Between May 1996 and May 2017, 1,719 patients with drug-­ resistant epilepsy underwent neurosurgical interventions at the “Claudio Munari” Epilepsy Surgery Center aimed at resecting

Outcome on seizures was assessed according to Engel’s classification.20 It was considered a good outcome when class 1 (free of disabling seizures) was achieved, and a poor outcome in all other cases. We examined a number of possible predictors of the outcome on seizures. The categorical variables examined were gender, handedness, a positive history for a seizure-free period since epilepsy onset, seizure rate, sleep-related seizures, adoption of video-EEG, video-SEEG and FDG-PET, site and type of resection, and pathological substrate. Numerical variables examined were age at the seizure onset, duration of epilepsy, and age at surgery. The Welch two-sample t-test was used to analyze numerical variables. Fisher’s two-tailed exact test was performed to analyze categorical variables. p < 0.05 was considered as evidence of findings not attributable to chance. Statistical analysis was performed using R 3.4.0.21

„„ Results Twenty-six patients fulfilled the above-reported inclusion criteria; 16 were male and 10 were female. The main demographic and clinical data are shown in Table 39.1. The mean age at epilepsy onset (± standard deviation [SD]) was 10.5 years ± 4.45 (range: 1–18). Fifteen patients were drug resistant from the beginning, while the remaining 11 were drug responsive for an initial period. The length of this free-of-seizure interval (± SD) was 7.45 years ± 6 (range: 1–20). The mean duration (± SD) of the epilepsy before surgery was 19.2 years ± 7.9 (range 5–35) for all the 26 patients. The mean age at surgery (± SD) was 30 years ± 8.2 (range: 16–44). Only two patients were younger than 18 at surgery time, being 16 and 17 years old. At surgery time, seizures occurred rarely, monthly, weekly, or daily in 1, 6,

39  Surgical Management of MRI-Negative Temporal Lobe Epilepsy

Fig. 39.3  The main steps of an anteromesial temporal lobectomy performed on the right side. (a) After a frontotemporal skin incision, the craniotomy is performed. (b) Typical pterional craniotomy should be avoided. The edges of the bony operculum must be positioned in order to obtain a complete exposure of the temporal lobe. (c) After the dural flap is reflected, the sylvian fissure, as well as the superior, middle, and inferior temporal gyri are exposed. (d) The brain resection is performed by means of the ultrasonic aspirator and a simple dissector, which is very helpful for the subpial dissection. The dissection of the superior temporal gyrus from the insular cortex is depicted. (e) After the resection of the amygdala and uncus, cisternal neurovascular structures are visible in transparency through the arachnoid membrane. (f) Great care must be taken when dissecting the hippocampus and the parahippocampal gyrus.

13, and 6 patients, respectively. Only two patients suffered from predominantly nocturnal seizures. All patients but one underwent noninvasive long-term ­video-EEG monitoring. Seventeen patients underwent ­videoSEEG monitoring; five of them underwent rf-THC, but none of them became seizure free. Seventeen patients were operated on the right side, the remaining nine on the left side. A complete anteromesial temporal lobectomy was performed in 23 of the 26 patients, while in the remaining three subjects the mesial structures were spared. The amount of the lateral resection was mainly based on the anatomo-electro-clinical correlations, especially when SEEG findings were available. The resection of the superior temporal gyrus was limited to the projection of the precentral sulcus in case of language-dominant hemisphere. The topography of the resection on the other temporal gyri was not different between the two sides. Focal cortical dysplasia (FCD) was found in three patients, and hamartoma in one. No pathological diagnosis was made in the remaining 22 subjects. Fourteen of 26 patients had a good outcome, while the remaining 12 had a poor outcome. Twelve of the 14 patients were Engel class 1a or 1c, thus were free from any type of seizure.

The mean follow-up (± SD) was 91.5 months ± 69. Ten of the 14 patients with a good outcome have either already stopped their medications, or are beginning to taper their medication. No major complications occurred with SEEG procedure or surgical cortical resections. After open surgery, only one minor intracranial bleeding was observed and treated conservatively, without any sequelae. An expected quadrantanopia occurred in 12 patients. None of the explanatory variables was significantly associated with the outcome on seizures (Table 39.2 and Table 39.3).

Illustrative Case The patient was a left-handed, 32-year-old man at surgery time (Fig. 39.4). Weekly seizures started at the age of 9, and these were mostly during sleep. They were characterized by a slow right deviation of the head, rapidly followed by dystonic posturing of the left upper limb. Late gestural and oroalimentary automatisms frequently occurred, along with a short language deficit. Scalp video-EEG monitoring allowed recording interictal and ictal epileptiform abnormalities in the right temporofrontal region. MRI was normal, while i­nterictal

379

Sex

Male

Male

Male

Female

Male

Female

Female

Female

Female

Male

Female

Female

Male

Male

Male

Female

ID

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

Right

Right

Right

Right

Right

Right

Right

Right

Right

Right

Right

Right

Right

Left

Left

Left

Handedness

7

11

4

9

11

8

15

15

11

5

16

8

18

8

14

11

Age at epilepsy onset (years)

0

0

15

7

2

5

0

0

1

6

0

20

0

8

0

0

Free interval (years)

29

5

24

22

22

11

8

15

16

17

14

31

20

24

30

14

36

16

28

38

33

19

23

30

Daily

Daily

Monthly

Weekly

Weekly

Weekly

Weekly

Daily

Weekly

Daily

22 27

Weekly

Monthly

Rare

Monthly

Weekly

Weekly

Seizure rate

30

39

38

32

44

25

Epilepsy Age at duration surgery (years) (years)

Table 39.1  Main demographic and clinical characteristics

No

No

No

No

No No

Yes

No

No

Yes

No

No

No

Yes

Yes

No

Yes

No

Yes

FDGPET

No

No

No

No

No

No

Yes

No

Yes

No

No

No

No

Sleep-­ related seizures

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

VEEG

Yes

No

Yes

No

No

No

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

v-SEEG

Left

Right

Left

Right

Right

Right

Right

Left

Left

Right

Right

Right

Right

Right

Right

Right

Side

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Only lateral

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Resection

113

36

140

Cryptogenic 171

Cryptogenic 186

Cryptogenic 73

Cryptogenic 12

Cryptogenic 232

FCD type 1

Cryptogenic 88

Cryptogenic 158

Cryptogenic 169

Cryptogenic 236

Cryptogenic 16

Cryptogenic 43

Cryptogenic 144

FCD type 2a

Poor

Good

Good

Good

Poor

Poor

Good

Poor

Poor

Poor

Good

Poor

Good

Good

Good

Good

Outcome

(Continued)

Follow-up (months)

Cryptogenic 49

FCD type 1

Pathology

380 IVc  Temporal Lobe Epilepsy and Surgical Approaches

Male

Female

Male

Female

Male

Male

Male

Male

Male

Male

17

18

19

20

21

22

23

24

25

26

Right

Left

Right

Right

Right

Right

Right

Right

Right

Right

Handedness

16

2

12

15

10

14

12

13

7

1

Age at epilepsy onset (years)

9

13

1.5

0

0

0

0

0

0

0

Free interval (years)

24

35

7

23

16

9

22

26

22

13

40

37

19

38

26

23

34

39

29

14

Epilepsy Age at duration surgery (years) (years)

Weekly

Weekly

Daily

Weekly

Monthly

Monthly

Weekly

Monthly

Weekly

Daily

Seizure rate

No

No

No

No

No

No

No

No

No

No

Sleep-related seizures

No

No

No

No

Yes

Yes

Yes

Yes

No

No

FDGPET

Yes

Yes

Yes

Yes

Yes

No

Yes

Yes

Yes

Yes

VEEG

No

No

Yes

Yes

Yes

No

Yes

No

No

Yes

v-SEEG

Left

Right

Left

Right

Left

Left

Right

Left

Right

Right

Side

Lateromesial

Lateromesial

Only lateral

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Lateromesial

Only lateral

Resection

Follow-up (months)

140 Cryptogenic 31

Hamartoma

Cryptogenic 44

Cryptogenic 39

Cryptogenic 35

Cryptogenic 49

Cryptogenic 16

Cryptogenic 25

Cryptogenic 57

Cryptogenic 77

Pathology

Abbreviations: FCD, focal cortical dysplasia; FDG-PET, [18F]-fluorodeoxyglucose positron emission tomography; VEEG, video electroencephalography; v-SEEG, video stereoelectroencephalography.

Sex

ID

Table 39.1 (Contiuned) Main demographic and clinical characteristics

Poor

Poor

Good

Good

Poor

Good

Good

Poor

Good

Poor

Outcome

39  Surgical Management of MRI-Negative Temporal Lobe Epilepsy

381

382

IVc  Temporal Lobe Epilepsy and Surgical Approaches Table 39.2  Categorical explanatory variables and outcome on seizures (two-tailed Fischer’s exact test)

Variable

Type

Categories

Frequencies (no. of cases)

Outcome (%)

P value

Good

Poor

Gender

Binomial

Female Male

10 16

40 62.5

60 37.5

0.42

Handedness

Binomial

Left Right

4 22

50 75.5

50 25.5

0.6

Seizure-free presurgical period

Binomial

No Yes

15 11

66.6 36.3

33.3 63.6

0.23

Seizure rate

Multinomial

Rare Monthly Weekly Daily

1 6 13 6

100 50 61.5 33.3

0 50 38.5 66.6

0.66

Sleep-related seizures

Binomial

No Yes

24 2

58.3 0

41.7 100

0.2

Long-term ­video-EEG ­monitoring

Binomial

No Yes

1 25

100 52

0 48

1

Long-term ­video-SEEG ­monitoring

Binomial

No Yes

9 17

44.4 58.8

55.5 41.2

0.68

FDG-PET

Binomial

No Yes

16 10

43.8 70

56.2 30

0.25

Resection side

Binomial

Left Right

9 17

33.3 64.7

66.6 35.3

0.22

Type of resection

Binomial

Mesiolateral Only lateral

23 3

56.5 33.3

43.5 66.6

0.58

Pathology

Multinomial

Cryptogenic Hamartoma Type 1 FCD Type 2a FCD

22 1 2 1

54.5 0 50 50

45.5 100 50 50

0.85

Abbreviation: EEG, electroencephalography; FCD, focal cortical dysplasia; FDG-PET, [18F]-fluorodeoxyglucose positron emission tomography; SEEG, stereoelectroencephalography.

Table 39.3  Numerical explanatory variables and outcome on seizures

All patients mean (± SD)

Good mean (± SD)

Poor mean (± SD)

P value

Age at seizure onset (years)

10.5 (± 4.45)

11.9 (± 3.86)

8.9 (± 4.71)

0.1

Duration of epilepsy (years)

19.2 (± 7.86)

17.4 (± 7.8)

21.3 (± 7.75)

0.22

Age at surgery (years)

30 (± 8.19)

29.8 (± 8.1)

30.17 (± 8.64)

0.91

Abbreviation: SD, standard deviation. Note: The Welch two-sample t-test was computed to compare the means grouped by results.

FDG-PET showed a slight hypometabolism in the right temporal and orbitofrontal regions. fMRI demonstrated a left dominance for language. A right temporal and anterior frontal SEEG exploration was performed. Repetitive bursts of low-­ amplitude fast discharges, spikes, and polyspikes were mostly located in the anterior and superior temporal neocortex, often involving the superior temporal sulcus and the superior temporal gyrus more posteriorly. Ictal discharges originated from the same region, followed by late spreading to hippocampus and orbitofrontal cortex. In detail, the seizure started from the anterior part of the superior temporal gyrus (T1–T2 leads), located at the level of the peak of the first PLPC. Based on

these recordings, a right temporal resection was performed, including the pole and the superior temporal gyrus. The temporomesial structures were spared. The patient is seizure free 37 months after surgery, and medical therapy is under progressive reduction. Histological examination revealed a type 2a FCD in the superior temporal gyrus. The first PLPC fits well with the EZ, being included in the resection zone. Continuing to lower the P-value offset, the result of statistical analysis progressively spreads posteriorly and anteriorly to the lateral temporal cortex and the orbitofrontal region, respectively. This progressive enlargement mimics the spreading of ictal electrical activity recorded by SEEG electrodes.

39  Surgical Management of MRI-Negative Temporal Lobe Epilepsy

Fig. 39.4  An illustrated case. (a) The multimodal scene illustrates the position of the temporofrontal SEEG electrodes. (b) A coronal FLAIR image cut at the level of the FCD found at histological examination on the superior temporal gyrus, where the most internal leads of the electrode T record from. No obvious lesions are visible. (c) The SUPR-FLAIR analysis shows the peak (green cross) of the first PLPC fitting well with the position of the “hottest” recording leads. Presurgical (d) and postoperative (e) multiplanar and 3D reconstructions.

„„ Discussion It is very difficult to find any literature about the particular population discussed in this section.22,​23,​24 This patient group also constitutes a minor cohort of the total population of the epilepsy patients operated on at the “Claudio Munari” center between 1996 and 2017. In fact, only two subjects with MRI-negative TLE were operated on in “pediatric” (< 18 years) age group at our center. Although the remaining patients had seizure onset in childhood (< 18 year of age), their surgeries were performed at greater than 18 years of age. A first question arises: is it correct to consider only patients who underwent surgery during the considered range of age? Or is it more appropriate to include also the patients whose epilepsy started in pediatric age, regardless the age at surgery? We chose to adopt the second criteria, because age at epilepsy onset is a fully patient-related characteristic, while age at surgery is, at least partially, an operator-related one. Thus, our study included 26 patients with an epilepsy onset before the age of 18. A second question arises: why most of the patients who could be operated on in pediatric age underwent surgery after waiting for a so long time? This is an important topic, because many literature evidences reported a shorter illness duration and a younger age at surgery as positive predictors of a good outcome on seizures.3 We therefore analyzed the reason patients included in our study waited a mean of 19 years

before being operated on, despite their epilepsy onset occurred in mean at the age of 10.5. Eleven of the 26 patients had an initial period of drug responsiveness, with a seizure-free period of 7.45 years, and thus justifying the late surgery. However, the remaining 15 patients were drug responsive for only 12 months, and another explanation for the prolonged delay before respective surgery needs further investigation. These 15 patients were referred to our center very late, because the surgical option was considered as last resort by their relatives or referring physicians. Therefore, we believe that every effort should be made to further disseminate our findings about the potential benefits of early surgical resections even in these challenging cases. In fact, despite the lack of an MRI-visible lesion, we obtained a good outcome in more than half of our patient population. None of the potential predictors resulted to be significantly associated with the outcome on seizures in the present study. This is probably due to the small sample size. In fact, it must be remembered that the lack of significance does not imply absence of effect, but only that it was not possible to reject the null hypotheses.25 Further along this line, some considerations can be done comparing the mean values or proportions of our sample. The age at seizure onset was the only variable showing a trend toward significance despite the small sample size (p = 0.1). The first seizure occurred later in patients with a good outcome; the mean age at seizure onset was 11.9 years and 8.9 years for patients with good and poor outcome, respectively. This is probably due to the

383

384

IVc  Temporal Lobe Epilepsy and Surgical Approaches longer duration of epilepsy for poor-outcome patients, but, again, the small sample size does not allow us to further analyze the data with a multivariate model that could control the confounding effect. Similarly, the patients who had a considerable seizure-free period had a poor outcome. Again, this could be likely justified by longer illness duration. Apparently, in contrast with most previously published studies, the patients who underwent video-SEEG monitoring had a more favorable outcome. Usually, SEEG-studied patients have a less favorable outcome because they are more complex as compared to patients who had no previous invasive recordings. In addition, the MRIs are often noninformative. In the present study, all the patients had a negative MRI, and therefore the implantation of intracerebral electrodes provided a diagnostic advantage. This suggests that invasive recordings are still the best option when imaging does not provide any information. Some of the patients in our series were studied with FDGPET scan, and their seizure outcome was better than the ones who were not. This type of metabolic imaging likely provides a critical information in the absence of MRI-visible abnormalities, but unfortunately, FDG-PET was only recently available at our center. The advantage of having informative imaging is also supported by the case we illustrated, where an advanced image processing technique revealed a possible lesional area in an

­therwise MRI-negative patient. Such special techniques, o including voxel-based or surface-based analysis of several ­morphometric brain properties, can be drastically fruitful in a number of MRI-negative patients.8,​26,​27,​28,​29,​30 Finally, it should be noted that pathological examinations in 22 out of 26 cases showed no histological abnormality, despite the fact that all the surgical specimens were entirely examined by the pathologist.

„„ Conclusion Despite some well-recognized negative predictors of outcome on seizures, such as the absence of MRI-visible abnormalities and a long epilepsy duration before surgery, more than half of the studied patients had a good outcome, testifying one more time that seizure freedom can be obtained in many of the most challenging cases if an appropriate diagnostic and therapeutic workflow is adopted. Further studies are needed for the assessment of prognostic factors in this particular population of pediatric patients operated on in the temporal lobe without any MRI-visible lesion. Considering the rarity of such patients, a multicenter study is indicated.

References 1. Wiebe S, Blume WT, Girvin JP, Eliasziw M; Effectiveness and Efficiency of Surgery for Temporal Lobe Epilepsy Study Group. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med 2001;345(5):311–318 2. Munari C, Bancaud J. The role of stereo-electroencephalography (SEEG) in the evaluation of partial epileptic seizures. In: Porter RJ, Morselli PL, eds. The Epilepsies. Bodmin: Butterworth & Co. (Publishers) Ltd; 1985:267–306 3. Spencer S, Huh L. Outcomes of epilepsy surgery in adults and children. Lancet Neurol 2008;7(6):525–537 4. Téllez-Zenteno JF, Hernández Ronquillo L, Moien-Afshari F, Wiebe S. Surgical outcomes in lesional and non-lesional epilepsy: a systematic review and meta-analysis. Epilepsy Res 2010;89(2–3):310–318 5. Bancaud J, Talairach J, eds. La Stéréo-ÉlectroEncéphaloGraphie Dans L’épilepsie. Paris: Masson & Cie, Editeurs; 1965 6. Cardinale F, González-Martínez J, Lo Russo G. SEEG, Happy Anniversary! World Neurosurg 2016;85:1–2 7. Colombo N, Tassi L, Deleo F, et al. Focal cortical dysplasia type IIa and IIb: MRI aspects in 118 cases proven by histopathology. Neuroradiology 2012;54(10):1065–1077 8. Cardinale F, Francione S, Gennari L, et al. SUrface-PRojected FLAIR analysis: a novel tool for advanced imaging of epilepsy. World Neurosurg 2017;98:715–726

new three-dimensional technique and literature review. World Neurosurg 2015;84(2):358–367 13. Jenkinson M, Beckmann CF, Behrens TE, Woolrich MW, Smith SM. FSL. Neuroimage 2012;62(2):782–790 14. Fischl B. FreeSurfer. Neuroimage 2012;62(2):774–781 15. Fedorov A, Beichel R, Kalpathy-Cramer J, et al. 3D Slicer as an image computing platform for the Quantitative Imaging Network. Magn Reson Imaging 2012;30(9):1323–1341 16. Balestrini S, Francione S, Mai R, et al. Multimodal responses induced by cortical stimulation of the parietal lobe: a stereo-­ electroencephalography study. Brain 2015;138(Pt 9): 2596–2607 17. Trébuchon A, Chauvel P. Electrical stimulation for seizure induction and functional mapping in stereoelectroencephalography. J Clin Neurophysiol 2016;33(6):511–521 18. Cossu M, Fuschillo D, Cardinale F, et al. Stereo-EEG-guided radio-frequency thermocoagulations of epileptogenic greymatter nodular heterotopy. J Neurol Neurosurg Psychiatry 2014;85(6):611–617 19. Cossu M, Fuschillo D, Casaceli G, et al. Stereoelectroencephalography-guided radiofrequency thermocoagulation in the epileptogenic zone: a retrospective study on 89 cases. J Neurosurg 2015;123(6):1358–1367

9. Cossu M, Lo Russo G, Francione S, et al. Epilepsy surgery in children: results and predictors of outcome on seizures. Epilepsia 2008;49(1):65–72

20. Engel JJ, Van Ness PC, Rasmussen TB, Ojemann LM. Outcome with respect to epileptic seizures. In: Engel Jr. J, ed. Surgical Treatment of the Epilepsies. 2nd ed. New York, NY: Raven Press, Ltd.; 1993:609–621

10. Cardinale F, Miserocchi A, Moscato A, et al. Talairach methodology in the multimodal imaging and robotics era. In: Scarabin JM, ed. Stereotaxy and Epilepsy Surgery. Montrouge: John Libbey Eurotext; 2012:245–272

21. R Core Team. (2017). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available at: https://www.R-project.org/. Accessed June 21, 2017

11. Cardinale F, Cossu M, Castana L, et al. Stereoelectroencephalography: surgical methodology, safety, and stereotactic application accuracy in 500 procedures. Neurosurgery 2013;72(3):353–366, discussion 366

22. McIntosh AM, Wilson SJ, Berkovic SF. Seizure outcome after temporal lobectomy: current research practice and findings. ­ ­Epilepsia 2001;42(10):1288–1307

12. Cardinale F, Pero G, Quilici L, et al. Cerebral angiography for multimodal surgical planning in epilepsy surgery: description of a

23. Miserocchi A, Cascardo B, Piroddi C, et al. Surgery for temporal lobe epilepsy in children: relevance of presurgical evaluation and analysis of outcome. J Neurosurg Pediatr 2013;11(3):256–267

39  Surgical Management of MRI-Negative Temporal Lobe Epilepsy 24. Muhlhofer W, Tan YL, Mueller SG, Knowlton R. MRI-­negative temporal lobe epilepsy—What do we know? Epilepsia 2017; 58(5):727–742

28. Hong SJ, Kim H, Schrader D, Bernasconi N, Bernhardt BC, ­Bernasconi A. Automated detection of cortical dysplasia type II in MRI-negative epilepsy. Neurology 2014;83(1):48–55

25. Altman DG. Statistics and ethics in medical research. VII—­ Interpreting results. BMJ 1980;281(6255):1612–1614

29. Hong SJ, Bernhardt BC, Caldairou B, et al. Multimodal MRI profiling of focal cortical dysplasia type II. Neurology 2017;88(8):734–742

26. Duncan JS, Winston GP, Koepp MJ, Ourselin S. Brain i­maging in the assessment for epilepsy surgery. Lancet Neurol 2016;15(4):420–433 27. Bernasconi A, Bernasconi N, Bernhardt BC, Schrader D. Advances in MRI for ‘cryptogenic’ epilepsies. Nat Rev Neurol 2011;7(2):99–108

30. Wang ZI, Jones SE, Jaisani Z, et al. Voxel-based morphometric magnetic resonance imaging (MRI) postprocessing in MRI-negative epilepsies. Ann Neurol 2015;77(6):1060– 1075

385

40

  Surgical Management of Insular–Opercular Epilepsy in Children Alexander G. Weil and Sanjiv Bhatia

Summary Insular or insular–opercular epilepsy (IOE) is an underreported form of epilepsy, particularly in childhood. A high index of suspicion of insular epilepsy is imperative, as failure to recognize the insula as part of the epileptogenic zone is a well-recognized cause of persistent disabling seizures following frontal (FLE), temporal (TLE), or parietal lobe epilepsy (PLE) surgery in children. In fact, a third of pediatric IOE patients present after having failed prior resective epilepsy surgery. Advances in the presurgical workup and neurosurgical techniques over recent decades have allowed centers to safely perform invasive investigation and resective surgery for epilepsy originating from the insular and perisylvian region. Over the last decade, several groups have reported their experience with insular epilepsy surgery in pediatric cohorts. Epilepsy surgery is effective in well-selected candidates, and surgical morbidity is relatively low in experienced centers. Detailed working knowledge of insular and perisylvian anatomy and sound surgical technique are essential. This chapter reviews pertinent insular–perisylvian anatomy and function, provides an up-to-date summary of the presurgical workup, indications and different methods for extraoperative invasive investigation, and describes nuances in resective surgical technique for pediatric insular epilepsy. It also summarizes the available evidence in favor of surgery for IOE in childhood. Keywords:  insula, insular–opercular epilepsy, perisylvian refractory epilepsy

„„ Introduction IOE or insular–perisylvian epilepsy is defined as epilepsy originating from the insula and its surrounding frontal, temporal, or parietal cortices.1,​2,​3,​4,​5,​6,​7 Although insular seizures were initially reported in 1954 by Penfield and Faulk,8 interest and surgery for insular epilepsy was abandoned for over 30 years due to the high initial surgical morbidity rates.9 Since its modern description in 2004,3 many centers around the world have reported on the identification and successful surgical management of insular epilepsy and IOE.3,​6,​10,​11,​12,​13,​14,​15,​16,​17,​18 Although extratemporal epilepsy (ETE) represents the most common form of epilepsy during childhood and adolescence,19,​20 most reports on IOE pertain to the adult population.3,​6 Until recently, IOE has remained poorly defined in children, underreported in the

literature, and previously not mentioned in pediatric epilepsy textbooks. IOE has, however, been increasingly recognized as a type of ETE or temporal plus epilepsy in childhood, particularly over the last decade.1,​2,​10,​13,​14,​21,​22,​23,​24 Identification of IOE is imperative, as unrecognized insular epilepsy has been found to be a significant cause of failure following surgery for refractory epilepsy of the frontal, parietal, and temporal lobe.3,​4,​5,​6,​25,​26,​27 Early intervention in cases of drug-resistant epilepsy is particularly important in children, as persistent recurrent seizures have a negative impact on the developing brain, leading to reduced cognition, school performance, daily functioning, and quality of life.28,​29,​30 This holds true for insular epilepsy as well, as the majority of these patients have cognitive dysfunction and many experience significant improvement following surgery.13,​14 Like other forms of drug-resistant ETE, insular epilepsy surgery can cure or reduce seizures, reduce antiepileptic drug (AED) use, improve neurocognitive development, and improve functional and health-related quality of life parameters.31,​32,​33,​34,​35 Like other forms of ETE, however, IOE represents a challenging form of epilepsy as the epileptogenic zone is usually multilobar (extending to contiguous frontal, parietal, and/or temporal cortices), can involve functional areas (e.g., Wernicke, Broca, sensorimotor hand or face), and can be difficult to localize, map out, and adequately resect.1,​2,​13,​21,​22 The deep location of the insula below the opercula, draped by the overlying middle cerebral artery (MCA) branches and the underlying critical subcortical structures, such as the internal capsule, arcuate fasciculus, and basal ganglia, render resective insular surgery challenging.12,​36 Recent advances in the presurgical evaluation and surgical technique have aided in mapping epileptogenic zone and carrying out insular–opercular resections. IOE surgery has been shown to harbor a very favorable efficacy to morbidity profile in both children and adults alike.

„„ Insular–Opercular Surgical and Functional Anatomy The Insula Pediatric epilepsy surgeons must have a detailed working knowledge of the insular lobe and perisylvian structures in order to safely perform IOE surgery.7,​36,​37 The insula, or island of

40  Surgical Management of Insular–Opercular Epilepsy in Children

Fig. 40.1  (a–c) Insular–opercular anatomy and relationship to sylvian vessels. (Reproduced with permission from Ture et al.36)

Reil, is the fifth cerebral lobe and is an inverted pyramid-shaped structure located in the depth of the sylvian fissure (Fig. 40.1). The insular cortex is made up of seven gyri in most patients, including three short gyri (anterior, middle, and posterior) anteriorly and two long gyri (anterior, posterior) posteriorly, in addition to the accessory and transverse gyri, which are located at the most anteroinferior portion of the insula at the insular pole. The insula is divided into a larger anterior and smaller posterior portion by the central insular sulcus, which extends obliquely from the superior peri-insular sulcus to the limen insulae at approximately the same level (up to 5 mm anterior) as the central sulcus. The insula is covered by the fronto-­orbital, frontoparietal, and temporal opercula,36 and it is separated from these structures by the anterior, superior, and inferior peri-insular sulci, respectively. The peri-insular sulci surround the entire insula except for at the limen insulae, transverse gyrus, and accessory gyrus. The transverse gyrus is continuous with the posterior fronto-orbital cortex and the accessory gyrus with the suborbital gyrus. The peri-insular sulcus represents an important surgical anatomical landmark during subpial operculoinsulectomy, as opercular resection extends into the insula. The limen insulae, located below the insular apex, also an important surgical landmark, corresponds to the location of the genu of the MCA before its bifurcation. The M2 segment of MCA branches cover and supply the insula until the peri-insular sulcus, where they become the M3 segment of the MCA. Long insular perforator arteries (LIAs) from M2 segment of MCA arise from the posterosuperior portion of the insula and supply the corona radiata.7,​36,​38 Long medullary arteries (LMAs) from MCA at the opercular region also supply the corona radiata.

The Opercula There are three major opercula covering the insular lobe. These include the fronto-orbital, frontoparietal, and the temporal operculum.36 The fronto-orbital operculum consists of the posterior orbital gyrus, the posterior portion of the lateral orbital gyrus, and the inferior frontal gyrus (pars orbitalis). The f­ ronto-orbital operculum overlays the anterior insula and is separated from the insula by the anterior peri-insular sulcus. The frontoparietal operculum is separated from the fronto-orbital operculum by the anterior ascending ramus of the superficial sylvian fissure that demarcates the pars triangularis and pars opercularis of the inferior frontal gyrus, the inferior portions of the precentral and postcentral gyri, and the superior portion of the

supramarginal gyrus. The frontoparietal operculum covers the superior insula and is separated from the insula by the superior peri-insular sulcus. The temporal operculum, which covers the inferior insula, is composed of the superior temporal gyrus, the posterior temporal pole, and the inferior portion of the supramarginal gyrus and is separated from the insula by the inferior peri-insular sulcus. Knowledge of the relationship of insular anatomy to the more superficial sylvian fissure and opercula is important during surgical treatment of IOE.7 It is important to be able to visualize which portion of the insula will be exposed during sylvian fissure exposure or following opercular retraction or resection.2 These relationships have been studied in several adult cadaveric studies. The apices of the pars orbitalis, pars triangularis, and pars opercularis all converge and lie superficial to the anterior short gyrus. The superior peri-insular sulcus is reached following retraction or resection of the pars orbitalis, pars triangularis, and pars opercularis of the inferior frontal gyrus.7,​36 More specifically, the pars triangularis covers the upper anterior short gyrus, and the pars orbitalis covers the upper part of the anterior short gyrus and the adjacent part of the anterior limiting sulcus. The pars opercularis covers the superior portion of the anterior and middle short gyri. The precentral gyrus is located above the upper part of the posterior short gyrus and the superior limiting sulcus. The postcentral gyrus covers the superior portion of the posterior and anterior long gyri.7 Thus, opening the posterior ramus of the superficial sylvian fissure, adjacent to the pars triangularis, will expose the anterior or middle short gyri and anteroinferior insula. The anterior horizontal and ascending rami will expose the anterior short gyrus (upper portion) and anterior and/or superior peri-insular sulcus. Opening the posterior ramus behind this point, however, will expose the posterior short and long gyri. While removing insular cortex, it is imperative to know and visualize the critical subcortical structures that lie underneath. The insular cortex overlays the extreme capsule, claustrum, external capsule, lentiform nucleus (putamen and globus pallidus), caudate nucleus, arcuate fasciculus, internal capsule, and thalamus.7,​36,​39,​40 The genu of the internal capsule is located below the posterior short gyrus. The lentiform nucleus lies deep to the middle part of the middle short gyrus (MSG), posterior short gyrus, and middle part of the posterior long gyrus (PLG). The central sulcus covers posterior third of lentiform nucleus. The posterior arm of internal capsule and thalamus is deep to lentiform at level of anterior long gyrus (ALG) and PLG. The limen insulae is medial to the temporal operculum. The

387

388

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches inferior limiting sulcus was located medial to the superior temporal sulcus. The supramarginal gyrus surrounds the upturned posterior end of the sylvian fissure. This gyrus lies superficial to the junction of the superior and inferior limiting sulci at the posterior end of the insula (Fig. 40.1).

Insular Function and Connectivity Our understanding of the intrinsic function and connectivity of the insula has come from studies in old world macaque monkeys, semiological descriptions of surgically confirmed insular seizures, electrocortical stimulation of depth electrodes placed within the insula in epileptic patients, and neuroimaging research.3,​4,​6,​41,​42,​43,​44 However, because of its deep location and relatively small size (2% of cortical surface), it has been less extensively studied and is less well described than the other cerebral lobes.43 Knowledge of insular function and connectivity provides insight into understanding both insular seizure semiology and IOE epileptogenic networks. The insula harbors widespread reciprocal structural and functional connections with diverse brain regions including afferent and efferent input mainly from the dorsal thalamus, amygdala, and various cortices of the frontal, parietal, and temporal lobe.42,​45,​46,​47 The insula has been functionally segmented into 2 to 13 functional subdivisions depending on the technique used.44,​48 The structural and functional subdivision of the insula have been traditionally divided using in vivo diffusion tensor imaging (DTI) and resting state functional MRI (fMRI) into two subdivisions with an anteroposterior gradient in insular connectivity, where anterior insula projects primarily to the frontal, orbitofrontal, olfactory, anterior temporal, and anterior cingulate cortex, while the posterior insula projects to parietal, posterior temporal, and sensorimotor areas in addition to a transitional insular zone with overlapping connections.44,​48,​49,​50 More recent functional imaging studies have allowed the insula to be subdivided into a tripartite organization with overlapping and unique connectivity profiles, including the dorsal anterior insula (dAI) that harbors connections with pre-genu anterior cingulate and frontal areas supporting higher-level cognitive processes, the ventral anterior insula (vAI) that harbors connections with limbic areas supporting affective processes, and the posterior insula (PI) that harbors connections with sensorimotor areas supporting interoceptive processes.42,​44,​51 This subdivision is referred to as cognition-­ emotion-interoception tripartite parcellation of the insula.44 In a meta-analysis of all functional neuroimaging studies, Kuth et al divided the insula into four functional subdivisions: in addition to the previously mentioned sensorimotor region located in the mid-PI, socioemotional region in the anterior-ventral insula, and cognitive anterior-dorsal region, they describe a discrete ­chemical-sensory central-olfactogustatory region.52 One of the main roles of the insula is sensorimotor processing including viscerosensory function (painful paresthesia), autonomic control (heart rate and blood pressure), and interoception, which describes the sensation of physiological condition of the body, such as thirst, palpitations, or gastric distention. The insula has a significant role in somatosensory processing and pain, and stimulation of the insula can elicit ipsilateral, contralateral, or bilateral paresthesias or constriction sensations. The posterior insula has a role in thermosensory function and pain perception. The insula is also involved in central auditory

processing. The insula harbors chemosensory function, including olfaction and primary gustatory area, which is believed to be colocated in the midinsula.42,​51 The parietal operculum and retroinsular region are involved in vestibular processing. The insula also harbors socioemotional processing functions, particularly as they pertain to self-awareness and introspection, such as emotional experience and empathy. The insula is involved in social cognition and has been implicated in risky decision making.42 Finally, the insula is involved in high-order cognitive functions, including language and verbal memory. In addition, the dAI integrates external sensory information with internal emotional and bodily state signals to coordinate brain network dynamics and to initiate switches between the default mode network and central executive network.53

„„ Surgical Pathology The pathological substrate underlying pediatric IOE is consistent with pediatric ETE in that most patients suffer from malformations of cortical development.54,​55 Few studies have focused solely on the pathological findings underlying IOE. In the eight largest surgical series of refractory pediatric IOE, pathology revealed focal cortical dysplasia in 50 to 90% of patients. The remaining patients suffered from a variety of epileptogenic lesions, including other malformations of cortical development, cortical gliosis, low-grade neoplasms (ganglioglioma, pilocytic astrocytoma, dysembryoplastic neuroepithelial tumor, oligodendroglioma), tuberous sclerosis, and other abnormalities (e.g., Rasmussen’s encephalitis).2,​10,​13,​14,​21,​22,​23,​24 In patients with concomitant temporal lobe resection, hippocampal sclerosis has been identified, which is not surprising, as insular epilepsy can coexist with mesial temporal sclerosis and TLE.5,​14,​26

„„ Patient Selection and Surgical Indications Surgical Candidate Identification and Selection Patients with suspected IOE should generally be deemed to have drug-resistant epilepsy after failure of two AEDs to control seizures.56 Once patients are deemed medically refractory, a comprehensive noninvasive presurgical workup is warranted to localize and map the epileptogenic zone and functional brain regions around the perisylvian region (e.g., language and motor).57,​58 The goal is to determine both candidacy and likelihood of seizure freedom for resective epilepsy surgery. Many cases of IOE are nonlesional, and the exact location and extent of the epileptogenic zone can be difficult to define noninvasively. In children, the majority of suspected IOE cases require invasive monitoring to confirm both insular epilepsy and adequately map the epileptogenic zone.

Preoperative Evaluation Comprehensive preoperative evaluation of refractory IOE patients includes history and physical examination, routine ictal

40  Surgical Management of Insular–Opercular Epilepsy in Children and interictal scalp electroencephalogram (EEG), long-term video-EEG monitoring, MRI, and neuropsychological studies. Scalp video-EEG and brain MRI represent the core presurgical tests.57 Other functional imaging studies, such as positron emission tomography (PET) and single-photon emission computed tomography (SPECT) and magnetoencephalogram (MEG), are second-tier tests, which can be performed in select cases. These second-tier tests have advantages and limitations and each one harbors relative value in the presurgical workup of these patients. fMRI may be performed for eloquent cortex mapping in select cases, particularly in cases where stereoelectroencephalography (SEEG) invasive investigation is planned. Pediatric insular epilepsy typically starts in the first years of life (Table 40.1). Although most patients are neurologically intact, some may present with preexisting impairment related to the hemispheric pathology, such as motor impairment, apraxia, or hemineglect. Preoperative cognitive status can be quite variable. Although some patients may have normal intellectual quotient and school performance, the majority of pediatric insular epilepsy patients have decreased cognitive function affecting a wide array of domains (memory, executive function, etc.) and varying levels of learning disabilities (Table 40.1). Insular epilepsy semiology has been well defined in the adult population.3,​6 Insular epilepsy semiology can mimic FLE, TLE, and PLE.6,​26,​27,​59 As can be concluded from the discussion of the role of the insula in various body functions detailed above, typical semiology suggestive of insular onset or involvement includes somatosensory symptoms, viscerosensory phenomenon, profound salivation, palpitations, pallor, pupillary changes,

and other autonomic disturbances, which may involve ictal motor or temporal lobe semiology (e.g., language and auditory). However, not all patients have typically florid symptoms. Thus, it is important to be cognizant of insular involvement in cases of failed periopercular epilepsy, FLE, PLE, or TLE surgery. In fact, approximately one-third of reported cases of pediatric insular epilepsy have failed prior lobectomy, most commonly frontal, followed by temporal and parietal lobe surgery. Because many of the insular seizure manifestations are subjective symptoms (e.g., somatosensory auras and viscerosensory auras), the assessment of IOE is particularly challenging in children who, because of developmental immaturity or cognitive limitations, may not be able to report these subjective manifestations.1,​2,​13,​14 The typical somatosensory auras (e.g., unpleasant paresthesias involving large territories) or viscerosensory–visceromotor manifestations (e.g., laryngeal constriction) have not been reported in younger children (e.g., under age of 7); however older children and particularly adolescents may report these symptoms. Indirect evidence of somatosensory or viscerosensory auras, such as strange r­eactions, arrest, or suffering, should thus be sought when considering insular involvement. Early motor manifestations (e.g., tonic and myoclonic) and neurovegetative symptoms are frequent in pediatric insular epilepsy and are suggestive of insular onset. Interestingly, up to 50 to 75% of patients undergoing IOE surgery have typical insular-like presentation in some studies.13,​14,​22 Long-term video-EEG monitoring usually helps lateralize the ictal-onset zone and provides some localizing value.6,​58,​59 In cases of pediatric IOE, ictal scalp EEG lateralizes to the

Table 40.1  Characteristics of patients undergoing resective surgery for drug-resistant pediatric insular–opercular epilepsy reported in the literature

Author, year

Patient number

Mean age at surgery (range)

Prior surgery: no. (%) and type

Early insular semiology

Lesional MRI, no. (%)

Ictal scalp v-EEG lateralized/localized (%)

Ahmed et al 201824

6

13.3 (5–16)

4 (66%) 2 ATL, 2 FPT

> 50%

3 (50%)

Lateralized (100%)

Freri et al 201714

16

12 (6–17)

0

< 50%

14 (88%)

Lateralized (100%), localized: perisylvian (64%), ­suprasylvian (36%)

Perry et al 201723

20

12.8 (6.1–18)

10 (50%) 6 FL; 2 STG-ins; 1 ATL; 1 PL

NA

6 (30%)

Lateralized (60%)

Alomar et al 201822

8

11.3 (3–18)

3 (37%) 1 Radiosurgery, 2 SDDE

NA

NA

NA

Ikegaya et al 201838

3

2.9 (2.5–3.3)

0

> 30%

3 (100%)

Lateralized (67%)

Weil et al 20161

13

8 (5–16)

3 (23%)

< 30%

9 (69%)

Lateralized (100%)

Dylgjeri et al 201413

10

6.4 (1.7– 13.7)

1 (10%)

> 50%

8 (80%)

Lateralized (90%)

7 (1–17)

0

> 30%

6 (100%)

Lateralized to frontotemporal 2/5 (40%); diffuse-generalized 2/5 (40%)

von Lehe et al 6 200910 Park et al 200921

6

4.2 (0.6–7)

0

NA

6 (100%)

Lateralized (100%)

Afif et al 200866

2

11 (9–13)

NA

NA

2 (100%)

NA

Abbreviations: ATL, Anterior Temporal Lobe; FPT, Frontoal-Parietal-Temporal; FL, Frontal Lobe; NA, not available; ND, not done; STG-ins, Superior Temporal Gyrus-Insula; PL, Parietal Lobe; v-EEG, video-electroencephalography.

389

390

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches involved hemisphere in the vast majority of cases and usually localizes to the frontal, temporal, and/or central leads, for example, perisylvian region.1,​13,​14,​22,​58 In the context of typical semiology and perisylvian structural abnormality on MRI, this can be helpful. Because of the location of the insula below the frontal, parietal, and temporal cortices, the spatial resolution of scalp EEG is limited, and it cannot differentiate insular seizures from those arising from surrounding frontal, temporal, or parietal cortices. In a study of nine adult patients with invasive EEG and surgically confirmed IOE (e.g., Engel class 1a outcome after operculoinsulectomy), anterior insular epilepsy harbored frontotemporal ictal discharges, whereas posterior epilepsy was associated with temporal ictal discharges.60 Scalp interictal EEG discharges are less useful in IOE than in epilepsy originating from other lobes.61 Children with pediatric IOE harbor more diffuse epileptogenic networks than their adult counterparts, and although patients can have focal findings, many pediatric patients have lateralized multifocal or nonlateralized interictal abnormalities on scalp EEG.13,​14 Multifocal or bilateral interictal spikes in video-EEG involving the frontal lobe can even be suggestive of insular onset.13 Although patients may have good seizure outcome despite nonlocalizing or bilateral generalized ictal activity, this finding has generally been associated with poorer seizure outcome.23 In the largest pediatric series to date, 30% of patients had either nonlocalizing or bilateral ictal activity, and 80% of these cases had poor seizure outcome.23 The presence of a structural abnormality involving the ­insula and/or opercula on MRI in a patient with refractory epilepsy strongly suggests the presence of IOE, particularly if the patient has typical insular-like seizure semiology and congruent noninvasive tests. The majority of surgically treated pediatric IOE cases reported to date are lesional, with MR abnormalities either limited to the insula, involving the insula and peri-insular structures, although MRI may show abnormalities involving peri-insular structures with a normal-appearing insula.1,​13,​14 However, MRI-negative IOE is not uncommon, and focal cortical dysplasia is the underlying pathological substrate in almost all reported pediatric cases of MRI-negative IOE.1,​13,​14 One recent pediatric study reported a 70% rate of nonlesional pediatric insular epilepsy.23 The absence of a lesion on MRI may reduce the likelihood of seizure freedom, as it renders localizing and assessing the extent of the epileptogenic zone more challenging. In the largest pediatric series reported to date, most report relatively poor seizure outcome (Engel class 3 or 4) among nonlesional cases. While one study reported only 25% poor outcome in nonlesional cases,2 poor outcome has been reported in at least half of nonlesional cases in most studies.13,​14,​23 These findings are not surprising as the epileptogenic zone and the extension of the surgical resection go in general beyond the well-defined MRI abnormality, particularly in children younger than 3 years of age.62,​63 Ictal or interictal regional cerebral blood flow (rCBF) SPECT or interictal FDG-PET have limited utility and have not been systematically utilized in the noninvasive presurgical workup of pediatric refractory epilepsy patients with suspected insular– opercular seizures.1,​13,​14 Interictal FDG-PET (II PET) rarely shows a well-localized hypometabolism in the insula or perisylvian region—more often, extrainsular, nonlocalized, bilateral, or more extensive multifocal changes are seen.13,​23 Interictal FDG-PET can be misleading in a quarter of patients with insular epilepsy. Noncongruent II PET hypometabolism should not deter from insular resection, as 25 to 40% of patients with nonlocalized, multifocal,

or bilateral II PET abnormalities have Engel class 1a outcome after insular–opercular resection.23 Ictal rCBF SPECT can show concordant hyperperfusion localized to the insular–opercular region in a large proportion of patients and represents a valuable tool, particularly in nonlesional cases.1,​2 In children, however, it often reveals abnormalities in the ipsilateral hemisphere outside the insula or nonlocalized bilateral abnormalities.14,​23 While concordant ictal SPECT may aid in the localization of IOE, discordant results should not deter from considering insular epilepsy and sampling the insula with invasive investigation. In a recent study in 20 epilepsy patients, 44% had Engel class 1a despite bilateral or nonlocalized hyperperfusion on ictal SPECT.23 MEG has emerged as a useful tool in refractory insular epilepsy, and a tight cluster in the insular–opercular region has correlated with control of seizures.64 In adult insular epilepsy, three patterns of MEG spike sources have been identified, including anterior operculoinsular clusters, posterior operculoinslar clusters, and diffuse perisylvian clusters.64 Two recent pediatric insular epilepsy studies have shown that MEG cluster localization can be useful for guiding invasive EEG monitoring. Furthermore, surgical resection or ablation of the epileptogenic zone defined by concordant MEG spike sources and invasive EEG monitoring were associated with favorable seizure outcome in most patients.23,​24

Indication for Invasive Monitoring Although noninvasive tests may rarely be sufficient to define the epileptogenic zone in cases of lesional IOE in which the lesion is clearly defined on imaging with congruent noninvasive tests, invasive investigation with insular–opercular coverage is usually indicated to confirm and map IOE. In fact, most ­reported (over 70%) pediatric IOE cases have undergone two-stage ­ resective surgery with invasive investigation to confirm insular e ­ pilepsy, define the epileptogenic zone, and map function (Fig. 40.2, Fig. 40.3, Fig. 40.4, Fig. 40.5, and Fig. 40.6). ­Single-stage insular epilepsy surgery can be performed when a well-defined legion (e.g., epileptogenic tumor) is a ­ ssociated with multiple concordant noninvasive tests and functional mapping is either performed noninvasively (e.g., fMRI) or not required based on presumed epileptogenic zone. The invasive intracerebral exploration is mandatory in the absence of a clear epileptogenic lesion, but also in the pres-

Fig. 40.2  Illustration demonstrating different modalities of insular electrodes insertion. (Reproduced with permission from Weil et al 2016.1)

40  Surgical Management of Insular–Opercular Epilepsy in Children

Fig. 40.3  Illustrative case of primarily orthogonal SEEG implantation with transopercular insular electrodes. A 10-year-old boy with perisylvian-like epilepsy semiology characterized by hypersalivation followed by tonic posture of four extremities with insular, perisylvian (superior temporal gyrus, inferior orbital gyrus, frontoparietal operculum), and mesial temporal signal abnormalities on MRI and hypometabolism in the perisylvian region on PET scan (a). SEEG invasive investigation with five transopercular insular electrodes seen on CT scan (b) and coregistered MRI (c, d) including three electrodes through the inferior frontal gyrus (PF, MF, AF) and two through the temporal lobe (PT, MP) in addition to amygdala (AMG). (d) Ictal onset zone (red dotted line) involving the frontoparietal and temporal operculum and insula is confirmed. (e) An additional oblique parasagittal (S) electrode was placed in the anterior insula covering the short gyri. (f) Postoperative MRI revealing perisylvian and insular resection.

ence of an ill-defined lesion to elucidate the correct location and extension of the epileptogenic zone and its relationships with functional areas. The insula should be considered in any case of failed FLE, TLE, or PLE, particularly if seizure semiology or noninvasive studies suggest insular involvement. In older adolescents and adult patients, there are five scenarios that are typically encountered in the epilepsy surgery conference that should raise suspicion of IOE and warrant invasive investigation.65 These include (1) MRI-negative FLE and PLE; (2) perisylvian epilepsy; (3) nonlesional sleep hypermotor epilepsy; (4) temporal plus epilepsy (TLE with early semiology suggesting insular involvement); and (5) lesional insular epilepsy in whom further epileptogenic zone and functional mapping are warranted.6,​11,​65 Using insular implantation criteria mentioned above, the rate of detection of insular seizures is variable, ranging from 10 to 16% up to 37% in some centers.11,​13,​15,​22,​66,​67 In younger children, insular epilepsy is commonly perisylvian but often involves extended, multifocal epileptogenic networks in adjacent frontal, temporal, or parietal lobe.

Options for Invasive Investigation The insula, perisylvian region, and adjacent cerebral lobes can be investigated using one of two well-described invasive sampling methods, either the SEEG approach, in which ­insular and ­extrainsular depth electrodes are placed with frame-based stereotaxy through small entry holes, or a direct open approach, in which depth electrodes are inserted under direct visualization into the insular cortex after a craniotomy following microsurgical splitting of the sylvian fissure in a ­ ssociation with hemispheric grids and strips with or without frameless ­neuronavigation.1,​11,​18,​68 A hybrid method has also been described in which SEEG is performed in combination with subdural strips through a mini-­craniotomy.11 The choice of insular and perisylvian implantation method depends on the expertise at each center and available resources (Table 40.2). Most centers use the SEEG approach, 13,​18,​22 in which depth electrodes can placed in the insular cortex using either an orthogonal trajectory perpendicular to the insula passing through the opercula (Fig. 40.2 and Fig. 40.3),

391

392

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 40.4  Illustrative case of parietal oblique parasagittal combined with transopercular SEEG electrode implantation in a patient with bitemporal and insular epilepsy. A 16-year-old adolescent boy had previously undergone two-stage right focal superior temporal gyrus corticectomy (a), had persistent seizures characterized by behavioral arrest, head turning, and autonomic disturbances. Ictal SPECT shows increased uptake (b) in right hemisphere including perisylvian region and insula. (c–e) Phase 1 SEEG investigation of insula was performed with parietal parasagittal insular electrodes (RFP, RTP), which recorded ictal activity from right insula, right hippocampal and temporal lobe, and left hippocampal and temporal lobe (f). The patient subsequently underwent palliative temporal lobectomy and insular corticectomy.

Fig. 40.5  Illustrative case of open direct orthogonal insular depth electrode placement. A patient with tuberous sclerosis (TSC) and a frontoinsular tuber undergoing invasive implantation of an orthogonally placed electrode after sylvian fissure splitting. After head positioning (a), the sylvian fissure is dissected to expose the insula (b), frameless stereotactic neuronavigation is used to identify desired site of electrode placement (c–e). Perisylvian subdural grids are then placed, and MRI (f–I) confirms adequate localization of electrode.

40  Surgical Management of Insular–Opercular Epilepsy in Children

Fig. 40.6  Illustrative case of open direct parasagittal transinsular apex depth electrode placement. A 10-year-old patient with prior single-stage left anterior temporal resection sparing mesial structures with immediate seizure recurrence. Invasive investigation with a parasagittal transinsular apex electrode. Following sylvian fissure dissection, the insular apex is exposed, and the electrode is inserted in a posterosuperior direction (a). In addition, left hippocampus depth and subdural electrodes covering orbitofrontal region, frontal operculum, and superior temporal gyrus as seen on intraoperative photograph (b) and postoperative coregistered MRI (c). ECoG identified the ictal-onset zone involving the left orbitofrontal cortex, left hippocampus, and left anterior insular region. Following functional mapping, which confirmed expressive language and tongue motor areas outside the ictal onset zone (d), the patient underwent resective surgery as seen on postoperative MRI (e).

18 an oblique parasagittal trajectory parallel to the insula in the sagittal plane (Fig. 40.4),66,​68 or a combination of both.18,​22 Additional depth electrodes are placed in areas of suspected epileptogenicity (Fig. 40.3 and Fig. 40.4). There are several advantages of SEEG insular sampling compared to the open method. The SEEG method is ideal when bilateral coverage is warranted to lateralize seizure onset, which may represent up to 14% of suspected pediatric insular epilepsy cases (Fig. 40.4).11,​22 SEEG also better captures seizure activity in three dimensions, as it offers better coverage in deep cortical and subcortical areas, such as the medial ­temporal structures.11,​69,​70 SEEG techniques avoid a craniotomy and are associated with 1.3% overall morbidity and generally lower surgical risks than open subdural electrode method.71,​72 Although there are concerns over hemorrhagic complications related to SEEG, the rate of hemorrhage is very low (< 0.1%), and most hemorrhagic complications are related to the entry point rather than the target.22,​73 There have been no reported complications (e.g., intracerebral hemorrhage within insula or opercula) directly related to SEEG-guided insular electrodes.22 When selecting a trajectory for SEEG-guided insular electrodes, one should consider the relative advantages and disadvantages of both the orthogonal transopercular and frontal or parietal oblique parasagittal SEEG approaches.68 Many centers, including ours, perform invasive investigation using a combination of orthogonal and parasagittal oblique insular electrodes in each patient (Fig. 40.2, Fig. 40.3, and Fig. 40.4).

In the orthogonal approach, insular depth electrodes are placed through the frontal, temporal, and/or parietal operculum, allowing simultaneous coverage of the insula and the operculum. This method is particularly useful for perisylvian epilepsy with suspected opercular involvement, which represents a very large proportion of pediatric insular epilepsy patients (Fig. 40.3). It offers excellent mediolateral coverage of the insular cortex. However, because the insular cortex is quite thin, this method only allows for one to two recording contacts per electrode within the insula, and a higher number of electrodes may be necessary in order to achieve good insular coverage.3,​13,​22,​27 Another disadvantage of this method is MCA branches overlying the insular cortex, which limits electrode placement within regions directly under these vessels.36 Concerns regarding transient neurological deficits related to transgression of the eloquent opercular cortex with the electrode or injury to MCA vessels have not been borne out in the literature, and these electrodes are in fact very safe.18,​22 The safety of this method relies on adequate perioperative planning and technique and avascular electrode trajectory avoiding MCA vessels.18,​22 By contrast, the parasagittal oblique method, which is performed through a frontal or parietal approach, allows for a higher amount of contacts within the insula per electrode (better contact to electrode ratio), with up to six to eight contact points from a posterior approach within the insula, which allows the greatest insular coverage.22,​66,​68 The parietal approach

393

394

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches Table 40.2  Methods of insular sampling in pediatric insular epilepsy

Technique

Insular electrode

Advantages

Disadvantages

Frame-based SEEG

TPO depth electrode

Utilizes a non-eloquent corridor Avoids craniotomy, sylvian fissure d ­ issection, opercular retraction Avoids passage through eloquent o ­ percula and avoids risk of MCA injury Follows long axis of insula: high contact-to-electrode ratio Greatest insular coverage Ideal when bihemispheric coverage is ­warranted More suitable with suspected posterior insular or parietal seizure focus

Worse mediolateral insular coverage Limited anterior insular coverage Requires computer-based registration Requires stereotactic frame placement Requires postoperative scanning (coregistered to MRI) to confirm electrode placement

TFO depth electrode

Same as TPO except more suitable with suspected anterior insular or frontal seizure focus

Worse mediolateral coverage Limited posterior insular coverage Worse overall insular coverage than TPO Less contacts/electrode than TPO a­ pproach Requires computer-based registration Requires stereotactic frame placement Requires postoperative scanning (coregistered to MRI) to confirm electrode placement

Combined TFO/TPO

Combined advantages of TFO and TPO

Worse mediolateral coverage Requires computer-based registration Requires stereotactic frame placement Requires postoperative scanning (coregistered to MRI) to confirm electrode ­placement

OTO with stereotactically coregistered vascular imaging

Most well-established method Good opercular coverage Good medial and lateral insular coverage Provides landmark for subpial insular resection

Theoretical risk of MCA vascular injury or sulcal injury Lower contact-to-electrode ratio (n = 2), thus requiring more electrodes Lower insular coverage, particularly anteroinferior insula from overlying MCA More time consuming Less hemispheric coverage

Combined TFO/TPO/ OTO

Combined advantages of TPO, TFO, and OTO

Requires computer-based registration Requires stereotactic frame placement Requires postoperative scanning (coregistered to MRI) to confirm electrode ­placement

Transsylvian orthogonal depth combined with hemispheric subdural strips and grid

Allows extensive ipsilateral hemispheric/­ opercular coverage (Ideal for functional mapping) Good medial + lateral insular coverage Electrode used as landmark during 2nd phase for subpial insular resection

Not indicated when bilateral coverage required Requires craniotomy, sylvian fissure dissection, opercular retraction Risk of MCA vascular injury, opercular ­retraction injury Higher complication rate Lower contact-to-electrode ratio (n = 2) compared to oblique (parasagittal) techniques

Transsylvian translimen parasagittal combined with hemispheric subdural strips and grids

• Allows extensive ipsilateral ­hemispheric/ opercular coverage • (Ideal for functional mapping) • Follows long axis of insula: high ­contact-to-electrode ratio

• Not indicated when bilateral coverage required • Requires craniotomy, sylvian fissure ­dissection, opercular retraction • Risk of MCA vascular injury, opercular retraction injury

Frameless stereotacy + direct through an open ­craniotomy

Abbreviations: MCA, middle cerebral artery; SEEG, stereoelectroencephalography; TFO, transfrontal oblique; TPO, transparietal oblique; OTO, orthogonal transopercular.

40  Surgical Management of Insular–Opercular Epilepsy in Children thus allows the exploration of multiple distinct insular gyri from the anterior and/or posterior insula with one single electrode66 and is useful when delineating the epileptogenic zone and can enable precise tailoring of resection when necessary. In addition to the excellent insular coverage, the major advantage of the frontal or parietal parasagittal oblique trajectory is its trajectory through non-eloquent brain and avoidance of the theoretical risk of MCA vessel injury. The frontal trajectory should be favored for anterior insular coverage or suspected frontal epilepsy, whereas posterior parasagittal electrode should be favored when posterior insular epilepsy and/or parietal involvement is suspected. The major disadvantage of SEEG is that there is less spatial resolution for outlining the ­epileptogenic zone and mapping eloquent functions on the cortical surface. This can be overcome, however, if multiple depth electrodes are placed, and the findings of functional imaging taken into consideration.74 The open technique of insular–opercular invasive investigation, which involves the placement of insular electrodes in addition to hemispheric strip and grids, has been performed by several groups.1,​21,​24,​67 Insular electrodes may be placed ­orthogonally11 or parasagittally through the insular apex1 after dissecting the sylvian fissure widely. The open approach is best reserved for unilateral cases, and the main advantage resides in the concomitant placement of extensive strip and grid coverage of frontal, parietal, and temporal cortices (Fig. 40.5 and Fig. 40.6).1 This technique provides excellent coverage of lateral parietal, lateral frontal, interhemispheric frontal–parietal, or lateral and inferior temporal neocortices. The hidden surfaces of the frontal, parietal, and temporal opercula are not adequately sampled, although novel hybrid electrodes have addressed this issue.75 This technique is thus ideal when a greater sampling of these extrainsular neocortical structures is required. In addition, it is optimal for functional mapping, particularly in the dominant hemisphere (Fig. 40.5 and Fig. 40.6). It is important to note that while several pediatric series have reported complication-free open invasive investigation of insular epilepsy,1,​21,​24 the advantages of the open approach must be weighed against the well-known risks associated with subdural grids and strips, which can occur in up to one-fifth of patients and include cerebral edema, subdural hemorrhage, and infection.72,​76,​77 An adult series on open insular–opercular invasive investigation reported 12.5% rate of transient neurological complication from insular electrode placement (one from electrode migration, another related to opercular retraction) and 19% complications for the subdural hemispheric coverage.11

„„ Surgical Considerations for Invasive Investigation Open Method The general steps are similar for subdural cortical electrodes whether placing transsylvian parasagittal transinsular apex or orthogonal insular electrodes. Frameless neuronavigation is utilized as an adjunct to confirm desired entry point, whether placing orthogonal or parasagittal electrodes. With the patient supine, the head is rotated 45 to 60 degrees to the contralateral

side and fixed in place with a three-point Mayfield head holder.1,​11 A large unilateral fronto-parieto-temporal scalp ­incision and craniotomy are performed centered on the perisylvian region (Fig. 40.5). The lesser wing of the sphenoid is drilled to facilitate exposure of the sylvian fissure and obtain access to the insular apex. Using the surgical microscope, the sylvian fissure is opened using well-known microsurgical techniques, avoiding fixed retractors and preserving the M3 MCA arteries and veins. Sylvian fissure opening should be tailored to the type of electrode implantation. In cases of parasagittal transinsular apex electrode implantation,1 opening the posterior ramus below the pars triangularis will expose the anterior or middle short gyri and anteroinferior insula, which leads down to the insular apex (Fig. 40.6). The relatively avascular insular apex and pole are exposed. The insular apex is the most prominent laterally projecting surface on the insula just above and behind the insular pole, the most anteroinferior point of the insula.1 The insular depth electrode can be inserted starting at an avascular surface of the insular apex above the MCA branches after pial incision with a microblade. The electrode is directed posterosuperiorly to follow the sagittal axis of the insula, staying parallel to the insular cortex in the subpial region. This can also be performed using ultrasound guidance. Drilling down the lesser wing of sphenoid helps in holding the depth electrode in the direction of the insula prior to insertion. A second, anteroinferiorly directed electrode can be placed for coverage of the more posterior insular cortex. For orthogonal insular electrodes, image guidance can be utilized to identify the site of desired sampling, and the fissure can be selectively opened at these locations. It is important to identify an avascular surface on the insular cortex (Fig. 40.5).11 Following insular implantation, Papaverine (30 mg/mL) can be placed over the sylvian vessels to prevent spasm of the M2 and M3 vessels. Subdural grids and strips are then placed over the adjacent opercula and cortical convexities of the frontal, parietal, and temporal lobes based on the presurgical evaluation (Fig. 40.5 and Fig. 40.6). The insular electrodes are sutured to the dura mater to prevent electrode migration.

SEEG Method The goal of SEEG is to place electrodes through an avascular trajectory and reach targets that are identified by analysis of noninvasive data.74 Some authors perform insular–opercular invasive investigation using the initial technique described by Talairach and Bancaud,3,​69 in which stereoscopic teleangiography is performed in frame-based stereotactic conditions, particularly for orthogonal (perpendicular to sagittal plane) transopercular electrodes. However, many centers have modified the workflow using CT angiography or MR angiography (MRA) and MR venography (MRV) with gadolinium.22 We perform orthogonal and oblique insular SEEG using frame-based CT or MRI on the day of surgery, coregistered to preoperative volumetric MRI, MRA, and MRV. The DICOM images are imported into BrainLab software, and electrode trajectories are planned using the iPlan software (Brainlab AG, Feldkirchen, Germany) to avoid vascular injury. In the posterior parietal oblique parasagittal approach, an entry point at the parieto-occipital junction targeting the anterior portion of insula is used, allowing the

395

396

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches greatest coverage of the insula. In the anterior frontal oblique parasagittal approach, electrodes are placed through the superior frontal gyrus into the anterior insula in a medial to lateral direction. By contrast, orthogonal transopercular insular electrodes are placed through the frontal, parietal, or temporal opercular lobes. The risk for vascular injury is in theory highest for this orthogonal approach, and meticulous trajectory planning is imperative. Many of the reported hemorrhagic complications occur at the entry point, and the vascular structures at the cortical surface should not be overlooked. Electrodes are then implanted using standard frame-based stereotactic coordinates after drilling of the skull, inserting the bone anchor, coagulating the dura, and followed by the placement of the SEEG electrode to the desired target. The electrode is secured to the screw with a tight seal in order to prevent cerebrospinal fluid leak. Many centers have now started using robotic assistance to facilitate placement of these electrodes. A follow-up volumetric MRI or CT is done immediately and fused to the preoperative MRI to localize the exact placement of the electrodes. The patient is then monitored in the Epilepsy Monitoring Unit. Illustrative cases of SEEG insular exploration are shown in Fig. 40.3 and Fig. 40.4.

„„ Insular–Opercular Resective Surgery Patient Positioning, Incision, and Craniotomy Insular–opercular resective surgery is performed through a pterional-based frontotemporal craniotomy. Frameless neuronavigation is a useful adjunct to correlate intraoperative anatomy with preoperative MRI. Coregisteration of anatomical MRI with noninvasive workup, including fMRI, PET, SPECT, and MEG cluster is very helpful. Tractography is also useful to avoid inadvertent injury to the pyramidal tract (motor) and arcuate fasciculus (language). With the patient supine, the head is rotated 40 to 60 degrees to the contralateral side. A bolster pad may be placed under the ipsilateral shoulder. About 15 degrees of extension is placed on the neck, and the head is elevated so that the zygoma is the highest point. In patients in whom a prior craniotomy has not been performed, an inverse question mark incision is performed followed by pterional craniotomy tailored to the size of perisylvian corticectomy. After a frontotemporal craniotomy, the lesser sphenoid wing is drilled down to the superior orbital fissure. This step is important to optimize exposure of the anterior insula and reduce brain retraction during insulectomy. After a C-shaped dural opening, the typical hemispheric anatomy can be seen. The subsequent steps depend on whether pure insulectomy or insular–opercular resection is to be performed.

Transsylvian Selective Insulectomy Cases of isolated insular cortex epilepsy without opercular involvement have been rarely reported in the pediatric population (Table 40.1; Fig. 40.7).10,​21 In children, pure insulectomy can be considered in lesional cases (within the insula) in which noninvasive and/or invasive data are concordant and

localize to the insula in isolation. In cases of pure insulectomy, a transsylvian approach with splitting of the fissure followed by insulectomy should be performed (Fig. 40.7). Sylvian fissure opening should be carried in a stepwise approach, starting from the superficial sylvian cistern and extending down to the operculoinsular cistern (deep sylvian fissure). Because the fissure is widest adjacent to the pars triangularis apex, sylvian fissure dissection should usually be started from this point. The fissure can then be dissected posteriorly, followed by anteriorly.78 Sylvian fissure dissection is tailored to the location of the insulectomy. When complete insulectomy is performed, a wide opening of the fissure is warranted. However, when partial insulectomy is performed, a limited corresponding portion of the fissure may be split. Detailed anatomical knowledge of opercular–insular topography correlation is imperative. 7,​36,​79 The sylvian fissure is split using sharp dissection technique. The fissure is widest at the level of the pars triangularis and dissection is usually started here.80 Stepwise dissection is c­ arried out using well-known techniques, in which MCA branches are followed down to open the operculoinsular cistern, all MCA branches and as many sylvian veins as possible are preserved.78 Insular cortex resection is performed by first dissecting the sylvian fissure into the superior and inferior sulcus. The cortex between the M2 branches of the MCA is coagulated and incised. The superficial cortex is resected in a subpial manner stopping just short of the white matter to avoid damage to the basal ganglia or subcortical pyramidal tracts. The M2 perforators going to the insula can be difficult to preserve, but the main arterial supply to the basal ganglia is from the lateral lenticulostriate vessels. Hence, complications can be avoided by limiting the resection of insular cortex to the white matter. Some authors have recommended a transopercular approach (through nonepileptic opercular tissue) for insular epilepsy to avoid fissure splitting and risk of vasospasm and opercular retraction.17 However, this approach results in a high rate of transient facial weakness and may put functional areas at risk. We recommend splitting the fissure and using this approach to complete insulectomy.81,​82

Subpial Operculoinsulectomy Because almost all pediatric cases of insular epilepsy involve the opercula, resection of the insular cortex can usually be carried out in a subpial manner following opercular topectomy (Fig. 40.8).2,​81 Opercular and perisylvian cortices are resected based on preoperative noninvasive and invasive workup maps. Standard microsurgical techniques should be used, including preservation of en passage cortical M2 and M3 MCA branches and as many veins as possible. Functional areas should also be preserved, including Broca’s area, Wernicke’s area, the inferior parietal lobule on the dominant side for language, and the sensorymotor area of the hand and face when possible. Following opercular resection, the resection of cortex can be carried down below the circular sulcus to the insular cortex in a subpial manner. In cases of IOE in which a large insulectomy is warranted but only a small opercular resection is required, it may be necessary to perform transsylvian insulectomy following splitting of the fissure. In addition, in cases that have had an open transsylvian placement of electrodes, insulectomy may be performed through this previously dissected corridor. The insular depth electrode can help guide the resection of involved

40  Surgical Management of Insular–Opercular Epilepsy in Children

Fig. 40.7  Types of insular resection: (a) selective insulectomy; (b) operculoinsulectomy; (c) orbitoinsulectomy; and (d) insulectomy combined with a lobectomy. (Reproduced with permission from Bouthillier and Nguyen 2017.81)

insular cortex. Insular corticectomy can also be combined with temporal lobectomy and opercular corticectomy (Fig. 40.7).81

Alternative Procedures Stereotactic Ablative Insulectomy Stereotactic techniques represent appealing minimally invasive alternatives to open insular resection, particularly in the setting of reoperation cases or patients who have undergone invasive investigation with SEEG. Insulectomy has been performed using SEEG-guided radiofrequency thermoablation and, more recently, MR-guided laser interstitial thermal therapy (MRgLITT).22,​83,​84,​85 SEEG-guided radiofrequency thermoablation for insular epilepsy should be considered a palliative procedure.83,​85 Experience with this technique is limited (only five patients reported to date), and seizure freedom is obtained in 20% and seizure reduction in 60% of cases.84,​85,​86 Over the last 5 years, MRgLITT has gained popularity in epilepsy surgery, particularly for deep-seated lesions or reoperation cases.87 There are three reports of MRgLITT for the treatment of insular epilepsy, including a recent series of 20 pediatric patients.22,​23,​84 The laser fiber may be placed through anterior, posterior, or combined oblique parasagittal approach similar to the approach used during SEEG. The efficacy of MRgLITT is slightly lower than open surgery, with 63% obtaining favorable seizure outcome at 1 year and 50% seizure free at 2 years. The lower seizure freedom rate may be related to incomplete

­ blation of the epileptogenic zone, as most patients in this a pediatric series underwent purely insular ablation rather than insular–opercular ablation. The major advantage of this approach include its minimally invasive nature, with reduced postoperative pain and length of hospitalization, as the vast majority of patients can be discharged home within 48 hours.23 The adverse event profile is also very favorable, as only 20% experience transient hemiparesis.23 MRgLITT is well adapted for patients who have undergone SEEG insular investigation and is an appealing option for patients at risk for multiple surgeries, such as tuberous sclerosis complex patients with IOE related to insular tuber, and we have performed a two-stage laser ablation for this indication with good results.

Insular Epilepsy Neuromodulation Responsive neurostimulation (RNS), an FDA-approved device that delivers electrical stimulation in response to recorded electrographic seizures in a real-time manner, has emerged as an adjunctive therapy for adult partial-onset seizures with frequent disabling seizures and no more than two seizure foci.88 RNS has been performed for insular epilepsy with the rationale of avoiding the surgical risk associated with open resection.89 Although four of the seven patients were “­responders” with at least 50% seizure reduction, there were two ­complications (hydrocephalus, wound infection) in this study. Because open insular resection is generally not associated with long-term deficits, RNS should be considered a

397

398

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 40.8  Technique for subpial insulectomy following opercular resection (a) in which opercular resection is mapped out based on invasive investigation identification of the epileptogenic zone (b), followed by resection of the opercula respecting the pial plane and en passage MCA branches (c, d) until the semicircular sulcus is reached and the insula removed guided by the depth electrodes (e, f).

last-­resort palliative alternative in adult patients who are not ­candidates for open surgery or if MRgLITT is unavailable.

„„ Outcome Seizure Outcome Resective surgery for IOE results in favorable seizure freedom, particularly when compared to seizure-free rates of pediatric ETE. In the pediatric population undergoing open resection, the reported rate of seizure freedom (Engel class 1 or 2 outcome) is approximately 70% (Table 40.3).1,​10,​13,​14,​21,​24,​38 These rates of seizure control are comparable to largest published series on IOE81 and pediatric ETE.19,​20 Persistent seizures have been ­related to residual insular cortex or unresected peri-insular tissue, which may involve functional zones such as orbitofrontal or posterior perisylvian functional language or motor areas.81 In our series, 30% (4 of 13 patients) had persistent seizures following­ insular–opercular resective surgery, three of whom achieved better seizure control with completion of insulectomy.2,​81 This finding was observed in a series of six pediatric patients in whom most (67%) required repeat insula resection to obtain seizure freedom.24 Two patients in another pediatric series had persistent seizures related to incompletely resecting the planned resection volume as shown on postoperative MRI.13

Surgical Morbidity Despite historical concerns regarding safety, many epilepsy centers have demonstrated that insular–perisylvian e ­pilepsy surgery can be carried out with low morbidity. Although

removing the insula may in theory cause deficits related to its intrinsic function, there is no evidence of permanent neurological deficits caused directly by insular cortex removal. One of the main concerns of insular epilepsy surgery is the risk for permanent disabling motor impairment. Although evidence from semiological description of seizures and stimulation studies suggests motor function within the insula, this is not universally accepted. The potential causes of motor impairment include ischemic damage to subcortical corticospinal tract from injury to lateral lenticulostriate arteries (LSAs) that supply the internal capsule, injury to LIA arising from M2 MCA branches over the posterior insula or LMA arising from opercular region supplying the corona radiata. Ischemic damage to primary motor cortex from MCA branch vasospasm or injury or direct injury to motor cortex from retraction or resection may also be possible. In pediatric series, the rate of transient and permanent contralateral motor impairment is variable, ranging from 0 to 31% and 0 to 20%, respectively.2,​10,​13,​14,​21 The majority of large pediatric surgical series report no permanent motor impairment (Table 40.3). Damage to the LSA, though rare, is a well-defined cause of permanent contralateral weakness in insular surgery.90 Transient hemiparesis is usually caused by remote corona radiata infarct secondary to inadvertent and almost unavoidable LIA or LMA sacrifice during insular corticectomy.38,​82 Up to 60% of patients have subcortical stroke on postoperative MRI, however only 40% experience clinically significant deficits (Fig. 40.9).82 Vasospasm of MCA branch vessels is another possible cause of transient contralateral motor weakness. Alternatively, retraction or resection of the frontal opercula may result in transient facial or brachiofacial weakness. These motor deficits can be mild or severe and are almost universally transient, particularly

16 (100%) insular– opercular

16 (100%) insular– opercular

13 (65%) insula; 7 (35%) insular–opercular

8 (100%) insular– opercular

3 (100%)

12 (92%)

10 (100%)

2 (33%)

6 (100%)

2 (100%)

6

16

20

8

3

13

10

6

6

2

Ahmed et al 201824

Freri et al 201714

Perry et al 201723

Alomar et al 201822

Ikegaya et al 201838

Weil et al 20162

Dylgjeri et al, 2014

von Lehe et al 200910

Park et al 200921

Afif et al 200866

Resection, No with extrainsular resection

No.

Author, year

Engel 1, n (%)

NA

1.5 (0.5–2.75)

62 (12–164)

27.6 (8–47)

35 (6–60)

NA

11.3 (3–17)

20.4 (7–39(

Median 39 (24–119)

1 (50%)

5 (83%)

5 (83%)

7 (70%)

9 (69%)

2 (67%)

1 (13%)

10 (50%)

9 (56%)

1 (50%)

5 (83%)

5 (83%)

8 (80%) Different seizures in 1

10 (77%)

2 (67%)

5 (62%)

11 (55%)

11 (69%)

4 (67%)

Engel 1 + 2, n (%)

1 (50%)

1 (17%)

1 (17%)

2 (20%)

3 (23%)

1 (33%)

3 (38%)

9 (45%)

5 (31%)

2 (33%)

Engel 3 +4

Seizure outcome (after repeat insular surgery)

2.8 (0.8–6.8) 4 (67%)

Mean follow-up months (range)

NA

NA

NA

Improved 8 (80%); unchanged 2 (20%)

NA

NA

NA

NA

Improved 6 (37.5%); unchanged 10 (62%)

NA

Cognitive outcome

NA

NA

1 (17%)

NA

2 (15%)

1 (33%)

1 (5%)

6 (30%)

5 (31%)

1 (17%)

Transient motor deficit

NA

NA

0 (0%)

2 (20%)

0 (0%)

0 (0%)

3 (15%) Improvement not specified

0 (0%)

1 (6%)

0 (0%)

Permanent motor deficit

Neurological morbidity

Table 40.3  Surgical outcome of patients undergoing resective surgery for drug-resistant pediatric insular–opercular epilepsy reported in the literature

NA

NA

0 (0%)

0 (0%)

0 (0%)

0 (0%)

0 (0%)

1 (5%)

2 (13%)

0 (0%)

Transient language impairment

NA

NA

0 (0%)

0 (0%)

0 (0%)

0 (0%)

0 (0%)

0 (0%)

0 (0%)

0 (0%)

Permanent language impairment

0 (0%)

0 (0%)

1 (17%)

0 (0%)

4 (31%)

0 (0%)

0 (0%)

4 (20%)

0 (0%)

2 (33%)

Reoperation following first insulectomy

40  Surgical Management of Insular–Opercular Epilepsy in Children

399

400

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 40.9  Subcortical infarct related to inadvertent damage to long M2 perforator following insular corticectomy as seen on illustration (a) and postoperative MRIs (b, c). (Reproduced with permission from Finet et al 2015.82)

when caused by LIA perforator corona radiata stroke or frontal opercular injury.82 Recovery usually takes a few weeks but may take up to 6 months. Surgeon experience likely plays an important role in preventing both transient and permanent neurological deficits. Careful preservation of the MCA branches and lateral lenticulostriate vessels and restricting insular resection to the cortex without extending down into the white matter of the extreme capsule may help prevent permanent motor deficit. A technique preserving posterior insular cortex below the superior peri-insular sulcus has been reported; however this may result in leaving behind residual lesion and poor outcome in one of three patients operated with this technique.38 Although permanent language impairment is rare (unreported in pediatric literature), transient language dysfunction may result from retraction or resection injury to language ­pathways in the frontal operculum, superior temporal gyrus, or inferior parietal lobule of the dominant hemisphere.

Neuropsychological Outcome In spite of the cognitive functions attributable to the insula and the potential impact on the developing brain, there is very limited data on the impact of insulectomy on neurocognitive function or social or emotional problems, particularly in children. Many patients will experience improvements in cognition following insular epilepsy surgery. In a series 16 pediatric IOE patients, Freri et al report improved cognition in 37% of patients and unchanged cognition in the remaining 63% of patients following resection.14 No patient experienced deterioration of standard neuropsychological tasks postoperatively. Dylgjeri et al reported neuropsychological improvement in 90% of patients undergoing insular epilepsy surgery, however, behavior was only improved in 20%.13 In adults, the available evidence suggests that insular– perisylvian resection can be performed without major permanent cognitive decline in most patients.91 In a series of 18 adult patients with drug-resistant IOE, variable cognitive changes were observed; however, the only significant deterioration occurred in a color naming task, which relies on oromotor speed and l­exical

access.92 Interestingly, a subset of adult patients undergoing partial or complete insulectomy develops subtle impairments in social cognition, empathy, and emotional processing.91,​93,​94,​95

„„ Conclusion Insular epilepsy is now widely recognized as a distinct form of surgically treatable pediatric focal epilepsy. A high index of suspicion for insular epilepsy is important, and good working knowledge of the insular–perisylvian anatomy is imperative to safely navigate the intricate anatomy of the perisylvian region. Insular–opercular epilepsy represents a challenging form of focal pediatric epilepsy for many reasons. The typical subjective semiology of insular epilepsy may not be reported in younger children. Scalp EEG may be associated with multifocal or even bilateral epileptic activity and other noninvasive tests such as PET and SPECT have limited value. Malformations of cortical development are the underlying pathological substrate in the majority of cases, and epileptogenic zone is almost universally multilobar, involving the insula and at least one other cerebral lobe. In addition, because a significant number of these cases are either nonlesional or involve epileptogenic zones extending beyond the visible lesion on MRI, assessing the extent of the epileptogenic zone is also challenging in these cases, as can be attested by the high rate of reoperation in these cases. Safe insular–opercular invasive investigation, whether performed open through a craniotomy or using SEEG techniques, relies on good surgeon anatomical knowledge and experience, meticulous planning, and advanced neurosurgical and perioperative imaging techniques. Despite these challenges, resective surgery offers good seizure freedom rates at low permanent morbidity profile, and the available evidence suggests that many patients experience cognitive improvement postoperatively. In upcoming years, refinement of presurgical ­noninvasive tests will improve our ability to identify and evaluate pediatric insular epilepsy to optimize patient selection and surgical results.

40  Surgical Management of Insular–Opercular Epilepsy in Children

References 1. Weil AG, Fallah A, Lewis EC, Bhatia S. Medically resistant pediatric insular-opercular/perisylvian epilepsy. Part 1: invasive monitoring using the parasagittal transinsular apex depth electrode. J Neurosurg Pediatr 2016;18(5):511–522

23. Perry MS, Donahue DJ, Malik SI, et al. Magnetic resonance ­imaging-guided laser interstitial thermal therapy as treatment for intractable insular epilepsy in children. J Neurosurg Pediatr 2017;20(6):575–582

2. Weil AG, Le NM, Jayakar P, et al. Medically resistant pediatric insular-opercular/perisylvian epilepsy. Part 2: outcome following resective surgery. J Neurosurg Pediatr 2016;18(5):523–535

24. Ahmed R, Otsubo H, Snead C III, et al. Diagnostic evaluation and surgical management of pediatric insular epilepsy utilizing magnetoencephalography and invasive EEG monitoring. Epilepsy Res 2018;140:72–81

3. Isnard J, Guénot M, Sindou M, Mauguière F. Clinical manifestations of insular lobe seizures: a stereo-electroencephalographic study. Epilepsia 2004;45(9):1079–1090 4. Isnard J, Mauguière F. [The insula in partial epilepsy] Rev Neurol (Paris) 2005;161(1):17–26 5. Nguyen DK, Nguyen DB, Malak R, Bouthillier A. Insular cortex epilepsy: an overview. Can J Neurol Sci 2009;36(Suppl 2):S58–S62 6. Nguyen DK, Nguyen DB, Malak R, et al. Revisiting the role of the insula in refractory partial epilepsy. Epilepsia 2009;50(3):510–520 7. Tanriover N, Rhoton AL Jr, Kawashima M, Ulm AJ, Yasuda A. Microsurgical anatomy of the insula and the sylvian fissure. J Neurosurg 2004;100(5):891–922 8. Penfield W, Faulk ME Jr. The insula; further observations on its function. Brain 1955;78(4):445–470 9. Silfvenius H, Gloor P, Rasmussen T. Evaluation of insular ablation in surgical treatment of temporal lobe epilepsy. Epilepsia 1964;5:307–320 10. von Lehe M, Wellmer J, Urbach H, Schramm J, Elger CE, Clusmann H. Insular lesionectomy for refractory epilepsy: management and outcome. Brain 2009;132(Pt 4):1048–1056 11. Surbeck W, Bouthillier A, Weil AG, et al. The combination of subdural and depth electrodes for intracranial EEG investigation of suspected insular (perisylvian) epilepsy. Epilepsia 2011;52(3):458–466 12. Malak R, Bouthillier A, Carmant L, et al. Microsurgery of epileptic foci in the insular region. J Neurosurg 2009;110(6):1153–1163 13. Dylgjeri S, Taussig D, Chipaux M, et al. Insular and insulo-opercular epilepsy in childhood: an SEEG study. Seizure 2014;23(4):300–308 14. Freri E, Matricardi S, Gozzo F, Cossu M, Granata T, Tassi L. Perisylvian, including insular, childhood epilepsy: presurgical workup and surgical outcome. Epilepsia 2017;58(8):1360–1369 15. Desai A, Jobst BC, Thadani VM, et al. Stereotactic depth electrode investigation of the insula in the evaluation of medically intractable epilepsy. J Neurosurg 2011;114(4):1176–1186 16. Dobesberger J, Ortler M, Unterberger I, et al. Successful surgical treatment of insular epilepsy with nocturnal hypermotor seizures. Epilepsia 2008;49(1):159–162 17. Gras-Combe G, Minotti L, Hoffmann D, Krainik A, Kahane P, ­Chabardes S. Surgery for nontumoral insular epilepsy explored by stereoelectroencephalography. Neurosurgery 2016;79(4): 578–588 18. Salado AL, Koessler L, De Mijolla G, et al. sEEG is a Safe Procedure for a Comprehensive Anatomic Exploration of the Insula: A Retrospective Study of 108 Procedures Representing 254 Transopercular Insular Electrodes. Oper Neurosurg (Hagerstown) 2018;14(1):1–8 19. Ansari SF, Maher CO, Tubbs RS, Terry CL, Cohen-Gadol AA. ­Surgery for extratemporal nonlesional epilepsy in children: a meta-analysis. Childs Nerv Syst 2010;26(7):945–951 20. Englot DJ, Breshears JD, Sun PP, Chang EF, Auguste KI. Seizure outcomes after resective surgery for extra-temporal lobe epilepsy in pediatric patients. J Neurosurg Pediatr 2013;12(2):126–133 21. Park YS, Lee YH, Shim KW, et al. Insular epilepsy surgery under neuronavigation guidance using depth electrode. Childs Nerv Syst 2009;25(5):591–597 22. Alomar S, Mullin JP, Smithason S, Gonzalez-Martinez J. Indications, technique, and safety profile of insular stereoelectroencephalography electrode implantation in medically intractable epilepsy. J Neurosurg 2018;128(4):1147–1157

25. Levitt MR, Ojemann JG, Kuratani J. Insular epilepsy masquerading as multifocal cortical epilepsy as proven by depth electrode. J Neurosurg Pediatr 2010;5(4):365–367 26. Ryvlin P, Kahane P. The hidden causes of surgery-resistant temporal lobe epilepsy: extratemporal or temporal plus? Curr Opin Neurol 2005;18(2):125–127 27. Ryvlin P, Minotti L, Demarquay G, et al. Nocturnal hypermotor seizures, suggesting frontal lobe epilepsy, can originate in the insula. Epilepsia 2006;47(4):755–765 28. Vasconcellos E, Wyllie E, Sullivan S, et al. Mental retardation in pediatric candidates for epilepsy surgery: the role of early seizure onset. Epilepsia 2001;42(2):268–274 29. Czochańska J, Langner-Tyszka B, Losiowski Z, Schmidt-Sidor B. Children who develop epilepsy in the first year of life: a prospective study. Dev Med Child Neurol 1994;36(4):345–350 30. Dwivedi R, Ramanujam B, Chandra PS, et al. Surgery for drugresistant epilepsy in children. N Engl J Med 2017;377(17): 1639–1647 31. Puka K, Smith ML. Predictors of long-term quality of life after pediatric epilepsy surgery. Epilepsia 2015;56(6):873–881 32. Tavares TP, Puka K, Smith ML. Emotional functioning: longterm outcomes after pediatric epilepsy surgery. Epilepsia 2015;56(5):745–753 33. Titus JB, Lee A, Kasasbeh A, et al. Health-related quality of life before and after pediatric epilepsy surgery: the influence of seizure outcome on changes in physical functioning and social functioning. Epilepsy Behav 2013;27(3):477–483 34. Wyllie E, Comair YG, Kotagal P, Bulacio J, Bingaman W, Ruggieri P. Seizure outcome after epilepsy surgery in children and adolescents. Ann Neurol 1998;44(5):740–748 35. Mühlebner A, Gröppel G, Dressler A, et al. Epilepsy surgery in children and adolescents with malformations of cortical development—outcome and impact of the new ILAE classification on focal cortical dysplasia. Epilepsy Res 2014;108(9):1652–1661 36. Türe U, Yaşargil DC, Al-Mefty O, Yaşargil MG. Topographic anatomy of the insular region. J Neurosurg 1999;90(4):720–733 37. Guenot M, Isnard J, Sindou M. Surgical anatomy of the insula. Adv Tech Stand Neurosurg 2004;29:265–288 38. Ikegaya N, Takahashi A, Kaido T, et al. Surgical strategy to avoid ischemic complications of the pyramidal tract in resective epilepsy surgery of the insula: technical case report. J Neurosurg 2018;128(4):1173–1177 39. Varnavas GG, Grand W. The insular cortex: morphological and vascular anatomic characteristics. Neurosurgery 1999;44(1):127–136, discussion 136–138 40. Naidich TP, Kang E, Fatterpekar GM, et al. The insula: anatomic study and MR imaging display at 1.5 T. AJNR Am J Neuroradiol 2004;25(2):222–232 41. Stephani C, Fernandez-Baca Vaca G, Maciunas R, Koubeissi M, Lüders HO. Functional neuroanatomy of the insular lobe. Brain Struct Funct 2011;216(2):137–149 42. Uddin LQ, Nomi JS, Hébert-Seropian B, Ghaziri J, Boucher O. Structure and function of the human insula. J Clin Neurophysiol 2017;34(4):300–306 43. Augustine JR. Circuitry and functional aspects of the insular lobe in primates including humans. Brain Res Brain Res Rev 1996;22(3):229–244 44. Nomi JS, Schettini E, Broce I, Dick AS, Uddin LQ. Structural Connections of Functionally Defined Human Insular Subdivisions. Cereb Cortex 2018;28(10):3445–3456

401

402

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches 45. Mathern GW, Giza CC, Yudovin S, et al. Postoperative seizure control and antiepileptic drug use in pediatric epilepsy surgery patients: the UCLA experience, 1986–1997. Epilepsia 1999;40(12):1740–1749

66. Afif A, Chabardes S, Minotti L, Kahane P, Hoffmann D. Safety and usefulness of insular depth electrodes implanted via an oblique approach in patients with epilepsy. Neurosurgery 2008;62(5, Suppl 2):ONS471–ONS479, discussion 479–480

46. Mesulam MM, Mufson EJ. Insula of the old world monkey. III: efferent cortical output and comments on function. J Comp Neurol 1982;212(1):38–52

67. Desai A, Bekelis K, Darcey TM, Roberts DW. Surgical techniques for investigating the role of the insula in epilepsy: a review. Neurosurg Focus 2012;32(3):E6

47. Mesulam MM, Mufson EJ. Insula of the old world monkey. I. Architectonics in the insulo-orbito-temporal component of the paralimbic brain. J Comp Neurol 1982;212(1):1–22

68. Robles SG, Gelisse P, El Fertit H, et al. Parasagittal transinsular electrodes for stereo-EEG in temporal and insular lobe epilepsies. Stereotact Funct Neurosurg 2009;87(6):368–378

48. Glasser MF, Coalson TS, Robinson EC, et al. A multi-modal parcellation of human cerebral cortex. Nature 2016;536(7615):171–178

69. Talairach J, Szikla G. Application of stereotactic concepts to the surgery of epilepsy. Acta Neurochir Suppl (Wien) 1980;30:35–54

49. Deen B, Pitskel NB, Pelphrey KA. Three systems of insular functional connectivity identified with cluster analysis. Cereb Cortex 2011;21(7):1498–1506

70. Cardinale F, Cossu M, Castana L, et al. Stereoelectroencephalography: surgical methodology, safety, and stereotactic application accuracy in 500 procedures. Neurosurgery 2013;72(3):353–366, discussion 366

50. Cerliani L, Thomas RM, Jbabdi S, et al. Probabilistic tractography recovers a rostrocaudal trajectory of connectivity variability in the human insular cortex. Hum Brain Mapp 2012;33(9): 2005–2034 51. Uddin LQ, Kinnison J, Pessoa L, Anderson ML. Beyond the tripartite cognition-emotion-interoception model of the human insular cortex. J Cogn Neurosci 2014;26(1):16–27 52. Kurth F, Zilles K, Fox PT, Laird AR, Eickhoff SB. A link between the systems: functional differentiation and integration within the human insula revealed by meta-analysis. Brain Struct Funct 2010;214(5–6):519–534 53. Uddin LQ. Salience processing and insular cortical function and dysfunction. Nat Rev Neurosci 2015;16(1):55–61 54. Blumcke I, Spreafico R, Haaker G, et al; EEBB Consortium. Histopathological findings in brain tissue obtained during epilepsy surgery. N Engl J Med 2017;377(17):1648–1656 55. Frater JL, Prayson RA, Morris HH III, Bingaman WE. Surgical pathologic findings of extratemporal-based intractable epilepsy: a study of 133 consecutive resections. Arch Pathol Lab Med 2000;124(4):545–549 56. Cross JH, Jayakar P, Nordli D, et al; International League against Epilepsy, Subcommission for Paediatric Epilepsy Surgery. Commissions of Neurosurgery and Paediatrics. Proposed criteria for referral and evaluation of children for epilepsy surgery: recommendations of the Subcommission for Pediatric Epilepsy Surgery. Epilepsia 2006;47(6):952–959 57. Jayakar P, Gaillard WD, Tripathi M, Libenson MH, Mathern GW, Cross JH; Task Force for Paediatric Epilepsy Surgery, Commission for Paediatrics, and the Diagnostic Commission of the International League Against Epilepsy. Diagnostic test utilization in evaluation for resective epilepsy surgery in children. Epilepsia 2014;55(4):507–518 58. Obaid S, Zerouali Y, Nguyen DK. Insular Epilepsy: Semiology and Noninvasive Investigations. J Clin Neurophysiol 2017;34(4): 315–323 59. Proserpio P, Cossu M, Francione S, et al. Insular-opercular seizures manifesting with sleep-related paroxysmal motor behaviors: a stereo-EEG study. Epilepsia 2011;52(10):1781–1791 60. Levy A, Yen Tran TP, Boucher O, Bouthillier A, Nguyen DK. ­Operculo-insular epilepsy: scalp and intracranial electroencephalographic findings. J Clin Neurophysiol 2017;34(5):438–447 61. Laoprasert P, Ojemann JG, Handler MH. Insular epilepsy surgery. Epilepsia 2017;58(Suppl 1):35–45 62. Taussig D, Dorfmüller G, Fohlen M, et al. Invasive explorations in children younger than 3 years. Seizure 2012;21(8):631–638 63. Taussig D, Montavont A, Isnard J. Invasive EEG explorations. Neurophysiol Clin 2015;45(1):113–119 64. Mohamed IS, Gibbs SA, Robert M, Bouthillier A, Leroux JM, Khoa Nguyen D. The utility of magnetoencephalography in the presurgical evaluation of refractory insular epilepsy. Epilepsia 2013;54(11):1950–1959 65. Ryvlin P, Picard F. Invasive Investigation of Insular Cortex Epilepsy. J Clin Neurophysiol 2017;34(4):328–332

71. Mullin JP, Shriver M, Alomar S, et al. Is SEEG safe? A systematic review and meta-analysis of stereo-electroencephalography-­ related complications. Epilepsia 2016;57(3):386–401 72. Hader WJ, Tellez-Zenteno J, Metcalfe A, et al. Complications of epilepsy surgery: a systematic review of focal surgical resections and invasive EEG monitoring. Epilepsia 2013;54(5): 840–847 73. Bourdillon P, Ryvlin P, Isnard J, et al. Stereotactic electroencephalography is a safe procedure, including for insular implantations. World Neurosurg 2017;99:353–361 74. Nowell M, Rodionov R, Diehl B, et al. A novel method for implementation of frameless StereoEEG in epilepsy surgery. Neurosurgery 2014;10(Suppl 4):525–533, discussion 533–534 75. Bouthillier A, Surbeck W, Weil AG, Tayah T, Nguyen DK. The hybrid operculo-insular electrode: a new electrode for i­ntracranial investigation of perisylvian/insular refractory epilepsy. Neurosurgery 2012;70(6):1574–1580, discussion 1580 76. Yang PF, Zhang HJ, Pei JS, et al. Intracranial electroencephalography with subdural and/or depth electrodes in children with epilepsy: techniques, complications, and outcomes. Epilepsy Res 2014;108(9):1662–1670 77. Wellmer J, von der Groeben F, Klarmann U, et al. Risks and benefits of invasive epilepsy surgery workup with implanted subdural and depth electrodes. Epilepsia 2012;53(8):1322–1332 78. Maekawa H, Hadeishi H. Venous-preserving sylvian dissection. World Neurosurg 2015;84(6):2043–2052 79. Benet A, Hervey-Jumper SL, Sánchez JJ, Lawton MT, Berger MS. Surgical assessment of the insula. Part 1: surgical anatomy and morphometric analysis of the transsylvian and transcortical approaches to the insula. J Neurosurg 2016;124(2):469–481 80. Ngando HM, Maslehaty H, Schreiber L, Blaeser K, Scholz M, Petridis AK. Anatomical configuration of the Sylvian fissure and its influence on outcome after pterional approach for microsurgical aneurysm clipping. Surg Neurol Int 2013;4:129 81. Bouthillier A, Nguyen DK. Epilepsy surgeries requiring an operculoinsular cortectomy: operative technique and results. Neurosurgery 2017;81(4):602–612 82. Finet P, Nguyen DK, Bouthillier A. Vascular consequences of operculoinsular corticectomy for refractory epilepsy. J Neurosurg 2015;122(6):1293–1298 83. Catenoix H, Mauguière F, Montavont A, Ryvlin P, Guénot M, Isnard J. Seizures outcome after stereoelectroencephalography-guided thermocoagulations in malformations of cortical development poorly accessible to surgical resection. Neurosurgery 2015;77(1):9–14, discussion 14–15 84. Hawasli AH, Bandt SK, Hogan RE, Werner N, Leuthardt EC. Laser ablation as treatment strategy for medically refractory dominant insular epilepsy: therapeutic and functional considerations. Stereotact Funct Neurosurg 2014;92(6):397–404 85. Catenoix H, Mauguière F, Guénot M, et al. SEEG-guided thermocoagulations: a palliative treatment of nonoperable partial epilepsies. Neurology 2008;71(21):1719–1726

40  Surgical Management of Insular–Opercular Epilepsy in Children 86. Guénot M, Isnard J, Ryvlin P, Fischer C, Mauguière F, Sindou M. SEEG-guided RF thermocoagulation of epileptic foci: feasibility, safety, and preliminary results. Epilepsia 2004;45(11): 1368–1374 87. Lewis EC, Weil AG, Duchowny M, Bhatia S, Ragheb J, Miller I. MR-guided laser interstitial thermal therapy for pediatric drug-­ resistant lesional epilepsy. Epilepsia 2015;56(10):1590–1598

91. Boucher O, Rouleau I, Escudier F, et al. Neuropsychological performance before and after partial or complete insulectomy in patients with epilepsy. Epilepsy Behav 2015;43:53–60 92. Jones CL, Ward J, Critchley HD. The neuropsychological impact of insular cortex lesions. J Neurol Neurosurg Psychiatry 2010;81(6):611–618

88. Morrell MJ; RNS System in Epilepsy Study Group. Responsive cortical stimulation for the treatment of medically intractable partial epilepsy. Neurology 2011;77(13):1295–1304

93. Von Siebenthal Z, Boucher O, Rouleau I, Lassonde M, Lepore F, Nguyen DK. Decision-making impairments following insular and medial temporal lobe resection for drug-resistant epilepsy. Soc Cogn Affect Neurosci 2017;12(1):128–137

89. Smith JR, Fountas KN, Murro AM, et al. Closed-loop stimulation in the control of focal epilepsy of insular origin. Stereotact Funct Neurosurg 2010;88(5):281–287

94. Boucher O, Rouleau I, Lassonde M, Lepore F, Bouthillier A, Nguyen DK. Social information processing following resection of the insular cortex. Neuropsychologia 2015;71:1–10

90. Lang FF, Olansen NE, DeMonte F, et al. Surgical resection of intrinsic insular tumors: complication avoidance. J Neurosurg 2001;95(4):638–650

95. Hébert-Seropian B, Boucher O, Sénéchal C, et al. Does unilateral insular resection disturb personality? A study with epileptic patients. J Clin Neurosci 2017;43:121–125

403

41

  Focal Cortical Dysplasia: Histopathology, Neuroimaging, and Electroclinical Presentation Olesya Grinenko and Imad M. Najm

Summary Focal cortical dysplasia (FCD) is a common pathological substrate in patients with focal epilepsies. FCD lesions were found in 50 to 75% of surgically treated epilepsies. Seizures due to FCD can manifest at any time during lifespan (from early infancy to late adulthood) but tend to more commonly occur in young children (before 5 years of age). Both clinical and histopathological data suggest that FCD is a heterogeneous group of congenital neurodevelopmental disorders. Type I FCD (in particular type Ia) is highly reminiscent of the fetal histological/architectural pattern (that is distinctly microcolumnar) seen in the first half of gestation (about 22 weeks in the human fetus), while type II FCD is characterized by the presence of abnormal cells (dysmorphic neurons with or without balloon cell) and is associated with several somatic and germline mutations in mTOR pathway. ­Magnetic resonance imaging (MRI) provides a noninvasive window for the in vivo detection of FCD. Epilepsies associated with FCD represent a broad spectrum of diseases with distinct electroclinical presentations that vary from focal epilepsy with normal development and preserved intellect to severe epileptic encephalopathies (generalized epilepsy phenotype) with neurodevelopmental delay. Seizures are usually frequent, highly stereotyped, and commonly manifest early in life. Epilepsy due to FCD may initially respond to medical treatment but eventually becomes pharmacoresistant and needs surgery in the majority of cases. Presurgical evaluation has several essential steps that include: (1) confirmation of epilepsy diagnosis and excluding genetic and metabolic causes, (2) definition of pharmacoresistance and the timing of presurgical evaluation (in children with neurodevelopmental delay, it should be started as soon as possible), (3) localization of the epileptogenic zone (with noninvasive techniques first and, if needed, with intracranial electroencephalography [EEG]), and (4) establishment of the anatomical relationships between epileptogenic and ­functional/eloquent cortices. Keywords:  focal cortical dysplasia, type I focal cortical dysplasia, type II focal cortical dysplasia, balloon cells

„„ Introduction Focal cortical dysplasia (FCD) is a distinct type of brain lesion that underlies the pathogenesis of epilepsy. Crome in 1957 originally described some of the key histopathological features of FCD,1 and

in 1971 Taylor reported on the presence of “focal cortical dysplasia” in 10 patients with medically intractable temporal lobe epilepsy. Taylor hypothesized that FCD is responsible for the electrical and clinical manifestations of focal epilepsy in the patients he studied.2 Over the next four decades, multiple studies confirmed the association between FCD and epilepsy: FCD was found as a pathological substrate in 50 to 75% surgically treated epilepsies.3,​4,​5,​6 Although epilepsy due to FCD can manifest at any age (from early infancy to late adulthood), it is most commonly expressed before 5 years of age.7 Clinical presentation significantly varies: from infrequent seizures with long periods of remission and normal neurodevelopmental function to severe epilepsy with developmental delay. Both clinical and histopathological data suggest that FCD is a heterogeneous group of congenital neurodevelopmental disorders. Type I FCD (in particular type Ia) is highly reminiscent of the fetal histological/architectural pattern (that is distinctly microcolumnar) seen in the first half of gestation (about 22 weeks in the human fetus).8 Recently, significant progress has been made in understanding the pathophysiological mechanisms underlying type II FCDs as several somatic and germline mutations in mTOR pathway were found in patients with type II FCDs (in particular FCD IIb).9,​10,​11,​12

„„ Histopathology and Classification of Focal Cortical Dysplasia Histopathologically, FCDs are characterized by a broad spectrum of architectural and cellular abnormalities that include: • Disrupted cortical architecture ranging from a disruption of only selected laminae or columns to complete loss of architectural cortical organization. yy Abnormal cells such as dysmorphic neurons (also known as neurofilament-accumulating neurons) and balloon cells (massively enlarged ovoid cells without any detectable cellular differentiation pattern). yy Abnormal gray–white matter border and neuronal heterotopia in subcortical white matter; neurons in the cortical lamina 1 and clustering of neurons in the gray matter. Several groups attempted to organize the histopathological features of FCD into uniform classification.13,​14,​15,​16,​17,​18 Palmini et al proposed the first clinically based classification of FCD with the main aim to integrate pathological features with imaging and

41  Focal Cortical Dysplasia: Histopathology, Neuroimaging, and Electroclinical Presentation Table 41.1  ILAE classification of focal cortical dysplasia

FCD type I (isolated)

FCD with abnormal radial cortical lamination (type Ia)

FCD type II

FCD with dysmorphic neurons (type IIa)

FCD type III Cortical lamination (associated with principal abnormalities in the lesion) temporal lobe associated with hippocampal sclerosis (type IIIa)

FCD with abnormal ­tangential cortical ­lamination (type Ib)

FCD with abnormal radial and tangential cortical lamination (type Ic) FCD with dysmorphic neurons and balloon cells (type IIb)

Cortical lamination abnormalities adjacent to a glia or glioneuronal tumor (type IIIb)

Cortical lamination abnormalities adjacent to vascular malformation (type IIIc)

Cortical lamination ­abnormalities adjacent to any other lesion acquired during early life, e.g., trauma, ischemic injury, encephalitis (type IIId)

Source: Data from Blümcke et al.18

clinical correlates.17 In that classification, FCDs were separated from mild malformations of cortical development and were divided into two main types: • Type I FCD (a and b subtypes): It was described as the presence of mild cortical architecture abnormality without dysmorphic neurons or balloon cells. Type Ia is characterized by isolated architectural disorganization, while type Ib is defined as architectural abnormality intermixed with giant, immature neurons (but not dysmorphic neurons). yy Type II FCD (a and b subtypes): It was defined as the presence of a disruption in the cortical architecture intermixed with dysmorphic neurons (type IIa) or the addition of balloon cells to architectural disorganization and dysmorphic neurons (type IIb). The Palmini classification was widely used both in research studies and in clinical practice (more than 900 citations to date), but its vague definition of FCD type I led to poor intraand interrater reproducibility.19 To overcome this shortcoming, the International League Against Epilepsy (ILAE) proposed a revised classification that included revised histopathological definition of type I FCD.18 The ILAE classification also introduced FCD type III as an association of type I with another principal pathology (such as hippocampal sclerosis, benign tumor, vascular malformation, and acquired lesions) (Table 41.1).

„„ Neuroimaging of Focal Cortical Dysplasia Magnetic resonance imaging (MRI) provides a noninvasive window for the in vivo detection of FCD. The use of MRI in clinical practice (in particular in epilepsy surgery) allowed the definition of the imaging characteristics of various types of FCDs.20,​21,​22,​23 The key MRI findings include the following (Fig. 41.1): • Increased cortical thickness and blurring of the gray–white matter junction. yy Increased cortical signal (mainly on T2-weighted and ­fluid-attenuated inversion recovery [FLAIR] sequences). yy “Transmantle sign”: an area of subcortical T2 and FLAIR signal abnormality that extends from the ventricle to the cortex.

yy Locally increased signal in the depth of a sulcus that linearly extends into the adjacent white matter (bottom-of-sulcus dysplasia); commonly seen in the setting of abnormally deep and misshaped sulcus. MRI findings can vary from no clear abnormalities (mainly in type I FCDs), to subtle and focal neocortical FLAIR/T2 signal increase on only one or two cuts (as in “bottom-of-sulcus dysplasia”), to clear tumor-like features with perilesional subtle signal increase and gray–white matter blurring (commonly seen in FCD IIb). In addition to signal change, abnormal sulcal or gyral patterns can be seen in the setting of FCDs,24,​25 but these abnormalities are usually difficult to visualize due to the large variations in normal gyral/sulcal anatomy. The so-called power button sign with greater number of side branches to the central sulcus and its connection to the precentral sulcus is usually seen in type II FCD.26 Another indirect finding that can help detect FCD is a focal or lobar atrophy reflecting decreased myelination. Decreased volume of temporal pole with blurring of the gray–white matter junction is a common feature of temporal lobe FCD.27 It is of note that this finding is nonspecific as it may also reflect degeneration of white matter fiber bundles that accompanies hippocampal sclerosis.28 Hemispheric volume loss may point to the presence of extended FCD type I.29,​30 Combination of several MRI features can suggest specific histopathological subtype: FCD type IIb commonly presents with a transmantle sign and “bottom-of-sulcus” phenotype,24 while only mild or no MRI changes are typically described in the setting of type I FCD.27,​31 As a general rule, pronounced MRI abnormalities usually point to FCD type II.23,​32 Additional postprocessing of the MR image can identify FCD subtype with better precision: in 2017, Hong and co-authors reported automatic profile-based classifier that distinguishes FCD type IIa and IIb with 85% sensitivity.33 Despite major advances in MRI technology over the last three decades, MRI-negative studies (no lesions identified upon visual analysis) are reported in 5 to 40% of type II FCD cases and in 60 to 70% of type I FCD lesions.23,​25,​31,​32,​34 Besides the histopathology of the FCD lesion, the success in the identification of FCD lesions on MRI is affected by other technical and interpretative factors that include MRI acquisition protocol, the patient’s age, the experience of the physician interpreting the data, and the possibility for

405

406

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 41.1  MRI examples illustrating the key imaging findings of focal cortical dysplasias (FCDs). (a) Coronal T1 showing the focal increase of cortical thickness and blurring of gray–white matter junction (FCD type Ia). (b) Axial fluid-attenuated inversion recovery (FLAIR) showing “transmantle sign” (note increased signal and thickness of an underlying cortex) (FCD type IIb). (c) Coronal FLAIR showing “bottom-of-sulcus” dysplasia (note transmantle sign) (FCD type IIb). (d) Axial FLAIR showing tuber-like FCD with variable degrees of signal increase (FCD type IIb).

re-evaluation of the images in the context of clinical findings. The detection can be improved by “expert” MRI assessment and by using epilepsy-dedicated MRI protocols.35 If the initial MRI interpretation is negative, the study should be re-examined in the context of clinical and neurophysiological data as subtle features of FCDs can be overlooked in up to 40% of patients.25 Reassessment of MRI in the context of all clinical data during patient management conferences with a multidisciplinary team can significantly improve the yield of the study in the detection of FCD lesions. In the pediatric population, it is critical to repeat the study, especially in children younger than 3 years, as brain myelination continues after birth and some lesions can become visible only after myelination is complete. Up to 21% of children with initially negative MRI can have positive results in repeated studies.36 MRI strength may increase the yield of FCD detection as 3T MRI scans may identify lesions that were missed on 1.5T scan.37,​38 There is limited information regarding 7T in clinical practice. MRI postprocessing techniques such as voxel-based morphometric (VBM) analysis can also improve FCD detection.39,​40 VBM produces a map of gray–white matter junctions highlighting the blurring of the junction. Despite promising results, VBM should be interpreted with caution and only in the context of the clinical data. If FCD remains undetected even after MRI protocol ­optimiza­tion, additional neuroimaging techniques can improve detection/visualization of some lesions and may guide further

presurgical evaluation. 18-Fluoro-2-deoxy-D-glucose p ­ositron emission tomography (FDG-PET) may identify some MRI-negative FCD lesions. PET shows hypometabolic region(s) in 50 to 90% of patients with FCD.34,​41,​42,​43,​44 Similar to MRI, PET results should be interpreted in the context of clinical and neurophysiological data, as the study can show a more extended hypometabolic area or several areas that may not co-localize with the lesion per se.42 In addition to hypometabolism, hypermetabolic PET areas are reported at times in electrographically active lesions (continuous spikes or seizures).45

„„ Electrographic Signature and Functional Status of Focal Cortical Dysplasia Pioneers of epilepsy surgery reported electrographic features of the epileptogenic brain regions well before the first histopathological description of FCD.46,​47 Subsequent studies co-localized epileptogenicity and structural (histopathological) abnormality.23,​48,​49,​50,​51,​52 The key electrographic signature of FCD (in particular type IIb FCD) in intracranial recording is continuous rhythmical or semi-rhythmical spiking with a ­transition to a fast activity at seizure onset.53,​54,​55 ­Occasionally, scalp electroencephalography (EEG) may reveal continuous rhythmical spiking in the region of FCD.56 But direct cortical and

41  Focal Cortical Dysplasia: Histopathology, Neuroimaging, and Electroclinical Presentation i­ntracortical electrocorticographic recordings (both intra- and extraoperative) show that spikes are mainly generated from the immediately surrounding cortex to the balloon cell–rich nidus of the lesion.55 Therefore, a complete resection of the spiking tissue is critical for good surgical outcome.53,​57 Not every FCD lesion generates rhythmical or semi-rhythmical spiking: only occasional nonspecific spikes can be recorded. These spikes could be part of network-dependent propagation patterns and therefore their localization significance is questionable. In these cases, seizure recording is critical, with scalp EEG first and then, if needed, with intracranial EEG. In intracranial recordings, FCDs have multiple ictal ­signatures.54,​58,​59,​60,​61 The majority of seizures (around 80%) start with fast activity that initiates from rhythmical spikes or bursts of polyspikes62 and may also be accompanied by slow baseline shift.63,​64 Fast activity is usually in the gamma band (between 30 and 120 Hz). Time–frequency analysis of ictal EEG can show a characteristic pattern that comprises banding structure of fast activity associated with suppression of lower frequencies and preictal spike(s).65 An important question for the localization of epileptogenicity in the setting of suspected FCDs (in particular MRI-negative lesions) is how to distinguish area of seizure generation (epileptogenic zone) from areas of seizure propagation/spread. Time of seizure onset can help in some cases, although it highly depends on electrode positions. Visual characteristics of the ictal patterns also may be identical in the epileptogenic zone and regions of propagation.61 In order to address this issue, several methods of intracranial EEG analysis were proposed with the main aim to identify and localize the epileptogenic zone. These include: (1) measurement of changes in energy ratio between fast and slow activities in relation to time of seizure onset (“epileptogenic index”);66 (2) co-localization of fast activity with slow polarization shift and voltage depression;64 and (3) support vector machine–based classifier identifying specific time–frequency pattern.6 These proposed analysis techniques can be used as additional methods, but the clinical decision should be always based on concordance of clinical data with other noninvasive evaluation information (such as scalp video EEG, MRI, PET, magnetoencephalography [MEG], single-photon emission computed tomography[SPECT]) and the intracranial findings. Functional reorganization has been previously described in patients with FCDs (in particular type IIb), but some forms of FCDs (mainly type I and IIa) may continue to harbor function (such as motor, sensory, and language).29,​67 For example, atypical sensorimotor homunculus was described in FCD type IIb and safe resections of the FCD localized in the Rolando area were reported.68,​69 We previously showed that the center (nidus) of balloon cell–containing regions is usually nonfunctional (and nonepileptic); the function is typically displaced to the contiguous and surrounding cortex that harbors other FCD subtypes (types I and IIa) that is both epileptogenic and functional.68 Inadequate mapping of the epileptogenic zone and subsequent incomplete resection of the surrounding epileptogenic area may lead to status epilepticus in the immediate postoperative period.57 Mapping of the epileptogenic and functional characteristics of the lesion and the surrounding areas could be accomplished intra- or extraoperatively following the placement of depth and/or subdural electrodes (for review, see Najm et al70).

„„ Electroclinical Epilepsy Syndromes Associated with Focal Cortical Dysplasia Epilepsies associated with various types of FCDs represent a broad spectrum of diseases with distinct electroclinical presentations that vary from focal epilepsy with normal ­development and preserved intellect to severe epileptic encephalopathy (generalized epilepsy phenotype) with neurodevelopmental delay. Seizures are usually frequent, highly stereotyped, and commonly manifest early in life (in the majority of patients before 5 years of age).7,​23,​34,​71,​72,​73,​74 Perirolandic FCD can manifest as status epilepticus, in particular in the form of epilepsia partialis continua.75 Epileptic encephalopathies are almost always associated with significant developmental delay but the child has a chance to achieve normal development only if epilepsy is successfully treated during the first 2 years after the onset.71,​72,​74 Focal epilepsy presents with focal seizures that may secondarily generalize. Seizures commonly respond to antiepileptic medications at the onset of the disease or during its course as epilepsy remits for several years, but eventually become pharmacoresistant.76 In general, anatomical location of the FCD determines the clinical presentation of seizures (semiology).77 But mechanisms of semiological seizure manifestation are complex and differ between primary and association cortex epilepsies.78 Patient’s age brings additional complexity as semiology may change between ages 1 and 14 years.7 If the epileptogenic lesion is localized in or near primary cortex (visual, sensory, or motor), ictal discharges typically lead to the generation of positive symptoms and/or signs. Epileptogenicity arising from the primary visual cortex leads to complaints of colored circles or twinkling stars in one of the visual field quadrants, 79 while ictal epileptic activation of the postcentral gyrus is associated with tingling sensation in one part of the body (commonly hand or arm), and a primary motor cortex activation leads to clonic jerks that spreads somatotopically (“jacksonian march”). If an epileptogenic lesion is located in an association cortex, clinical semiology is typically expressed following the ictal involvement of multiple brain areas. The pattern of activation is commonly driven by physiological–anatomical connections (subcortical intralobar, interlobar, and/or interhemispheric). For example, prefrontal epilepsy presents with complex gestural motor behavior when ictal discharges activate anterior cingulate, pre-SMA, and lateral premotor cortex.80 Analysis of seizure semiology in focal epilepsy brings valuable information in the process of presurgical evaluation. In addition, high-quality seizure videos with ictal and postictal patient testing significantly improve the yield from the video EEG evaluation. Seizure semiology should be analyzed independently but interpreted in the context of ictal EEG seizure pattern recorded during video-EEG monitoring. In cases of small and deep situated FCD, such as “bottom-of-sulcus” dysplasia, the initial ictal EEG discharges may remain undetected at seizure onset. In these cases, the early scalp EEG seizure patterns are typically due to spread (continuous and/or subcortical) to other surface cortical areas. These patterns are typically mislocalizing and, at times, mislateralizing.

407

408

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches Epileptic encephalopathy may present with generalized seizures, or with a combination of generalized and focal seizures. The clinical epilepsy phenotype could be West syndrome or Lennox–Gastaut syndrome. Less commonly, it presents as severe epilepsy with multiple spike foci81 or Ohtahara syndrome.82 Electrographic findings are generalized or diffuse: hypsarrhythmia in West syndrome, and generalized slow spike and waves/generalized paroxysmal fast activities in Lennox–Gastaut syndrome. Epileptic encephalopathies usually manifest during the first 2 years of life, and clinical presentation may change with patient’s age. For example, it can initially present as West syndrome and progress to Lennox–Gastaut syndrome.71 The majority of patients suffer from progressive neurodevelopmental delay.30,​71 The extent of the lesions together with age at epilepsy onset and its duration determine the severity of the delay.83,​84 The lesion can be striking, such as multilobar FCD type IIb or hemimegalencephaly, or mild, such as hemispheric type I FCD, presenting on MRI with isolated mild decrease in the volume of the affected hemisphere.29,​85 In addition to the developmental delay, patients may also exhibit congenital hemiparesis. Despite generalized clinical and electrographic characteristics, epilepsy surgery is highly effective in these cases. Moreover, shorter epilepsy duration before surgery increases the chance for improved development.71,​86,​87 The identification of the FCD lesion in patients suffering from epileptic encephalopathies is the main driver and the best predictor for a successful epilepsy surgery. Therefore, MRI and other noninvasive imaging techniques are of utmost importance in the presurgical workup of these patients.29,​72,​88,​89 In a small number of highly selected cases, intracranial EEG may be used to tailor the resection.70,​89 In those patients with multilobar/hemispheric unilateral lesion and congenital hemiparesis, a resection or disconnection of the damaged hemisphere can be the optimal choice as it leads to seizure freedom in more than 80% of patients.90,​91 In cases with more localized lesions and without neurological deficit, tailored lobar or multilobar resections should be tried first.71,​92

„„ Surgical Treatment and Outcomes of Epilepsy due to Focal Cortical Dysplasia The majority of patients with epilepsy due to FCD will need surgery as seizures commonly become pharmacoresistant. Presurgical evaluation has four essential steps: • Confirmation of epilepsy diagnosis and the exclusion of genetic and metabolic causes. yy Initiation of the presurgical evaluation immediately after the establishment of pharmacoresistance (in particular in those children with neurodevelopmental delay). yy Localization of the epileptogenic zone (with noninvasive techniques first and, if needed, using intracranial EEG methodologies).

yy Localization and definition of the anatomical relationships between epileptogenic and functional/eloquent cortices. The ILAE defined epilepsy pharmacoresistance as the failure “of adequate trials of two tolerated, appropriately chosen and used antiepileptic drug schedules (whether as monotherapies or in combination).”93 After the establishment of pharmacoresistance, the presurgical evaluation should be initiated as early as possible because longstanding epilepsy lowers chances for postresective seizure freedom92 and, in children with epileptic encephalopathies, leads to irreversible neurodevelopmental delay.71,​86 The primary goal of presurgical evaluation is the localization of the epileptogenic zone (the region of seizure generation). The phase 1 of the presurgical evaluation consists of scalp video EEG monitoring for the capture of stereotypical clinical seizures, and to establish the electroclinical correlations. Video EEG provides valuable information on seizure semiology, as well as localization of epileptiform discharges (identified in 70–100% of patients). The localization and extent of distribution of the interictal spikes as well as ictal patterns commonly extend beyond the MRI-identified epileptogenic lesion.94 In addition, a high-resolution MRI study with dedicated epilepsy protocol is needed. If MRI identifies an FCD lesion, then an anatomo-electroclinical hypothesis can be generated. Concordance of the data allows for surgical planning, but only after establishing anatomical relationships between functional and epileptogenic cortices. In these cases, additional electrocorticographic and functional mapping may be needed. The mapping may be done intraoperatively (if frequent or continuous spiking is seen during scalp EEG monitoring)53 or extraoperatively using prolonged intracranial recordings with depth and/or subdural grid electrodes.68,​69 Between 5 and 40% of type II FCDs and between 70 and 90% of type I FCDs are MRI-negative.23,​25,​31,​32,​34 In these cases, or in those patients where there is discordance in the electroclinical– imaging data acquired during phase 1, additional noninvasive techniques (PET, ictal SPECT, MEG, and MRI postprocessing) may assist in the formulation of a more clear localization hypothesis (phase 2).95-​98 In some patients, phase 2 studies may lead to the identification of an epileptogenic lesion, leading to the design of a resection strategy. In other patients, phase 2 enables the formulation of strong hypotheses for the planning of intracranial studies (phase 3). (Please see illustrative case. ) The following scenarios may be encountered in clinical practice: • Focal epilepsy with clear electroclinical–imaging correlations for an epilepsy arising from a nonfunctional cortex: surgical resection is recommended. yy Focal epilepsy with electroclinical–imaging correlations for an epilepsy arising from a potentially functional cortex: surgical planning and decision are made following ­intraoperative or extraoperative intracranial recordings. yy Focal epilepsy with negative MRI or discordant neurophysiological and clinical data: additional noninvasive evaluation (phase 2) followed by intracranial evaluation (phase 3) is needed. yy Epileptic encephalopathy (generalized epilepsy phenotype): surgery is decided based on MRI findings (mainly) and other

41  Focal Cortical Dysplasia: Histopathology, Neuroimaging, and Electroclinical Presentation noninvasive data (phase 2). In selected cases, chronic intracranial evaluation may be indicated (phase 3). yy The patient is not a candidate for surgery based on results of phase 1, phase 2, and/or phase 3 studies: further trial of antiepileptic medications, ketogenic diet, and neurostimulation treatments may be indicated. The first scenario is the most desirable and leads to the greatest rate of success in postresection seizure control; it predicts seizure freedom in 60 to 80% of the children.99,​100 In the second scenario, there are two possibilities: FCD harbor function (motor, sensory, or language), in which case its resection will lead to significant deficit,67,​91 or FCD led to a reorganization of the functional cortex and, thus, resection is possible but only after precise mapping of the functional and epileptogenic cortex, as incomplete resection of dysplastic cortex can lead to status epilepticus.69 Previous outcome studies showed seizure freedom rates of 50 to 70% if resection is performed in perirolandic cortex, but these resections are associated with high risk for a motor/sensory deficits.68,​69,​101,​102 In the third scenario, the anatomo-­electroclinical hypotheses are ­generated based on the results of phase 1 and phase 2 evaluations and then tested using intracranial studies (phase 3).103 Stereo-EEG (SEEG) can precisely localize epileptogenic zone and, in carefully selected cases, leads to seizure freedom in 50 to 80% of patients.92,​104,​105,​106 But the success of intracranial studies highly depends on the strength of the initial anatomo-electroclinical hypothesis. For these reasons, invasive recordings should not be performed if clear/ testable hypotheses cannot be generated. In cases of epileptic encephalopathies due to FCD (fourth scenario), the seizure semiology, as well as EEG, is generalized (and subsequently not helpful for the localization of the epileptogenic zone), but surgery may still be possible with high success rate if a “lesion” is identified.72 If a child already has congenial hemiparesis and extensive MRI lesion, then hemispherectomy may be considered as a primary surgical option; it will lead to seizure freedom in 80 to 90% of the cases.90,​91 In more limited lesions, tailored uni- or multilobar resection should be tried first, as it allows to achieve seizure freedom in 58 to 65% of these patients with milder deficit.71,​92 The situation is more difficult if MRI shows only mild but extensive/­diffuse abnormalities such as a reduced volume of an abnormal hemisphere in type I FCD. Most of these children will need disconnection of the abnormal hemisphere; although it will lead to seizure freedom in only 30 to 40%,74 it will add motor and cognitive deficits as FCD type I are intrinsically functional. In the last scenario, a patient is excluded from surgical treatment based on the results of phase 1, phase 2, and phase 3 evaluations. Further medical management and neuromodulation techniques such as vagus nerve stimulation, deep brain stimulation, and responsive neurostimulation should be tried (see Guerrini et al107 for review).

„„ An Illustrative Case Fig. 41.2 illustrates a clinical case with three phases of a presurgical evaluation in a patient with epilepsy due to suspected FCD. Clinical history: 17-year-old right-handed male who presented with focal drug-resistant epilepsy since the age of 9 years. The patient would describe an aura of “feeling of emptiness in the head.” During the seizure, he would lose consciousness and exhibits a change in his facial expression with down deflection of the lips bilaterally (“chapeau de gendarme”). He does not have postictal dysphasia or confusion. All seizures were stereotyped and would last about 30 seconds with frequency of up to 20 per day.

Phase 1 Scalp Video EEG showed interictal spikes in the left frontal region. Ictal onset pattern consisted of mixed frequency and amplitude spiking in the left frontal region associated with rhythmical delta activity over the left fronto-temporal and vertex regions. MRI did not reveal any abnormality.

Phase 2 FDG-PET identified hypometabolic area in the anterior part of the left superior frontal gyrus; MEG showed cluster of dipoles in the left anterior mesial part of the left superior frontal gyrus and two dipoles in the right superior frontal gyrus; Ictal SPECT (SISCOM analysis) revealed bilateral areas of hyperperfusion in the cingulate sulcus and superior frontal gyrus. A patient management discussion led to the proposal of an anatomo-electro-clinical hypothesis pointing to the left frontal lobe: prefrontal, frontopolar and mesial frontal regions. An SEEG evaluation (Phase 3) was recommended.

Phase 3 SEEG revealed continuous rhythmical spikes in the contacts sampling left anterior cingulate sulcus (L’1–3) and captured seizures characterized by fast activity arising from the same area but with almost simultaneous spread to neighboring regions. Electrical cortical stimulation of L’2–3 contacts with current intensity of 2mA elicited a stereotypical electro-clinical seizure. The results of the SEEG evaluation were discussed in another patient management conference with conclusion that the epileptogenic zone is localized in the left anterior cingulate sulcus. A surgical resection or laser ablation of this area was recommended. The patient and his family chose to proceed with the laser ablation (postoperative MRIs are shown in Fig. 41.3). Antiepileptic medications were stopped two years after the laser ablation. The patient remained seizure free at last follow-up (39 months post-ablation).

409

410

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 41.2  The clinical case illustrating the various phases of the presurgical evaluation in a patient with epilepsy due to suspected focal cortical dysplasia (FCD): (a) phase 1; (b) phase 2; (c) phase 3.

Fig. 41.3  Postoperative MRIs: laser ablation of the epileptogenic zone (cingulate sulcus/medial aspect of superior frontal gyrus).

41  Focal Cortical Dysplasia: Histopathology, Neuroimaging, and Electroclinical Presentation

References 1. Crome L. Infantile cerebral gliosis with giant nerve cells. J Neurol Neurosurg Psychiatry 1957;20(2):117–124 2. Taylor DC, Falconer MA, Bruton CJ, Corsellis JA. Focal dysplasia of the cerebral cortex in epilepsy. J Neurol Neurosurg Psychiatry 1971;34(4):369–387 3. Wyllie E, Comair YG, Kotagal P, Bulacio J, Bingaman W, Ruggieri P. Seizure outcome after epilepsy surgery in children and adolescents. Ann Neurol 1998;44(5):740–748 4. Sillanpää M, Gissler M, Schmidt D. Efforts in epilepsy prevention in the last 40 years: lessons from a large nationwide study. JAMA Neurol 2016;73(4):390–395 5. Hemb M, Velasco TR, Parnes MS, et al. Improved outcomes in pediatric epilepsy surgery: the UCLA experience, 1986–2008. Neurology 2010;74(22):1768–1775 6. Holthausen H, Pieper T, Winkler P, Bluemcke I, Kudernatsch M. Electro-clinical-pathological correlations in focal cortical dysplasia (FCD) at young ages. Childs Nerv Syst 2014;30(12):2015–2026 7. Fauser S, Huppertz HJ, Bast T, et al. Clinical characteristics in focal cortical dysplasia: a retrospective evaluation in a series of 120 patients. Brain 2006;129(Pt 7):1907–1916 8. Iffland PH II, Crino PB. Focal cortical dysplasia: gene mutations, cell signaling, and therapeutic implications. Annu Rev Pathol 2017;12:547–571 9. Lim JS, Kim WI, Kang HC, et al. Brain somatic mutations in MTOR cause focal cortical dysplasia type II leading to intractable epilepsy. Nat Med 2015;21(4):395–400 10. Jansen LA, Mirzaa GM, Ishak GE, et al. PI3K/AKT pathway mutations cause a spectrum of brain malformations from megalencephaly to focal cortical dysplasia. Brain 2015;138(Pt 6):1613–1628 11. Scerri T, Riseley JR, Gillies G, et al. Familial cortical dysplasia type IIA caused by a germline mutation in DEPDC5. Ann Clin Transl Neurol 2015;2(5):575–580 12. D’Gama AM, Woodworth MB, Hossain AA, et al. Somatic mutations activating the mTOR pathway in dorsal telencephalic progenitors cause a continuum of cortical dysplasias. Cell Reports 2017;21(13):3754–3766 13. Mischel PS, Nguyen LP, Vinters HV. Cerebral cortical dysplasia associated with pediatric epilepsy. Review of neuropathologic features and proposal for a grading system. J Neuropathol Exp Neurol 1995;54(2):137–153

21. Chan S, Chin SS, Nordli DR, Goodman RR, DeLaPaz RL, Pedley TA. Prospective magnetic resonance imaging identification of focal cortical dysplasia, including the non-balloon cell subtype. Ann Neurol 1998;44(5):749–757 22. Lee BC, Schmidt RE, Hatfield GA, Bourgeois B, Park TS. MRI of focal cortical dysplasia. Neuroradiology 1998;40(10):675–683 23. Widdess-Walsh P, Kellinghaus C, Jeha L, et al. Electro-clinical and imaging characteristics of focal cortical dysplasia: correlation with pathological subtypes. Epilepsy Res 2005;67(1–2):25–33 24. Hofman PA, Fitt GJ, Harvey AS, Kuzniecky RI, Jackson G. Bottom-of-sulcus dysplasia: imaging features. AJR Am J Roentgenol 2011;196(4):881–885 25. Mellerio C, Labeyrie MA, Chassoux F, et al. Optimizing MR imaging detection of type 2 focal cortical dysplasia: best criteria for clinical practice. AJNR Am J Neuroradiol 2012;33(10):1932–1938 26. Mellerio C, Roca P, Chassoux F, et al. The power button sign: a newly described central sulcal pattern on surface rendering MR images of type 2 focal cortical dysplasia. Radiology 2015;274(2):500–507 27. Bartolini L, Whitehead MT, Ho CY, et al. Temporal lobe epilepsy and focal cortical dysplasia in children: A tip to find the abnormality. Epilepsia 2017;58(1):113–122 28. Garbelli R, Milesi G, Medici V, et al. Blurring in patients with temporal lobe epilepsy: clinical, high-field imaging and ultrastructural study. Brain 2012;135(Pt 8):2337–2349 29. Krsek P, Maton B, Korman B, et al. Different features of histopathological subtypes of pediatric focal cortical dysplasia. Ann Neurol 2008;63(6):758–769 30. Blümcke I, Pieper T, Pauli E, et al. A distinct variant of focal cortical dysplasia type I characterised by magnetic resonance imaging and neuropathological examination in children with severe epilepsies. Epileptic Disord 2010;12(3):172–180 31. Tassi L, Colombo N, Garbelli R, et al. Focal cortical dysplasia: neuropathological subtypes, EEG, neuroimaging and surgical outcome. Brain 2002;125(Pt 8):1719–1732 32. Colombo N, Tassi L, Deleo F, et al. Focal cortical dysplasia type IIa and IIb: MRI aspects in 118 cases proven by histopathology. Neuroradiology 2012;54(10):1065–1077 33. Hong SJ, Bernhardt BC, Caldairou B, et al. Multimodal MRI profiling of focal cortical dysplasia type II. Neurology 2017;88(8):734–742

14. Barkovich AJ, Kuzniecky RI, Dobyns WB. Radiologic classification of malformations of cortical development. Curr Opin Neurol 2001;14(2):145–149

34. Lerner JT, Salamon N, Hauptman JS, et al. Assessment and surgical outcomes for mild type I and severe type II cortical dysplasia: a critical review and the UCLA experience. Epilepsia 2009;50(6):1310–1335

15. Barkovich AJ, Kuzniecky RI, Jackson GD, Guerrini R, Dobyns WB. A developmental and genetic classification for malformations of cortical development. Neurology 2005;65(12):1873–1887

35. Von Oertzen J, Urbach H, Jungbluth S, et al. Standard magnetic resonance imaging is inadequate for patients with refractory focal epilepsy. J Neurol Neurosurg Psychiatry 2002;73(6):643–647

16. Barkovich AJ, Guerrini R, Kuzniecky RI, Jackson GD, Dobyns WB. A developmental and genetic classification for malformations of cortical development: update 2012. Brain 2012;135(Pt 5):1348–1369

36. Jeon TY, Kim JH, Lee J, Yoo SY, Hwang SM, Lee M. Value of repeat brain MRI in children with focal epilepsy and negative findings on initial MRI. Korean J Radiol 2017;18(4):729–738

17. Palmini A, Najm I, Avanzini G, et al. Terminology and classification of the cortical dysplasias. Neurology 2004;62(6, Suppl 3):S2–S8 18. Blümcke I, Thom M, Aronica E, et al. The clinicopathologic spectrum of focal cortical dysplasias: a consensus classification proposed by an ad hoc Task Force of the ILAE Diagnostic Methods Commission. Epilepsia 2011;52(1):158–174 19. Chamberlain WA, Cohen ML, Gyure KA, et al. Interobserver and intraobserver reproducibility in focal cortical dysplasia (malformations of cortical development). Epilepsia 2009;50(12):2593–2598 20. Bronen RA, Vives KP, Kim JH, Fulbright RK, Spencer SS, Spencer DD. Focal cortical dysplasia of Taylor, balloon cell subtype: MR differentiation from low-grade tumors. AJNR Am J Neuroradiol 1997;18(6):1141–1151

37. Winston GP, Micallef C, Kendell BE, et al. The value of repeat neuroimaging for epilepsy at a tertiary referral centre: 16 years of experience. Epilepsy Res 2013;105(3):349–355 38. Mellerio C, Labeyrie MA, Chassoux F, et al. 3T MRI improves the detection of transmantle sign in type 2 focal cortical dysplasia. Epilepsia 2014;55(1):117–122 39. Wagner J, Weber B, Urbach H, Elger CE, Huppertz HJ. Morphometric MRI analysis improves detection of focal cortical dysplasia type II. Brain 2011;134(Pt 10):2844–2854 40. Wang ZI, Jones SE, Jaisani Z, et al. Voxel-based morphometric magnetic resonance imaging (MRI) postprocessing in MRI-negative epilepsies. Ann Neurol 2015;77(6):1060–1075 41. Chugani HT, Shields WD, Shewmon DA, Olson DM, Phelps ME, Peacock WJ. Infantile spasms: I. PET identifies focal cortical dysgenesis in cryptogenic cases for surgical treatment. Ann Neurol 1990;27(4):406–413

411

412

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches 42. Chassoux F, Rodrigo S, Semah F, et al. FDG-PET improves surgical outcome in negative MRI Taylor-type focal cortical dysplasias. Neurology 2010;75(24):2168–2175

63. Ikeda A, Terada K, Mikuni N, et al. Subdural recording of ictal DC shifts in neocortical seizures in humans. Epilepsia 1996;37(7):662–674

43. Rathore C, Dickson JC, Teotónio R, Ell P, Duncan JS. The utility of 18F-fluorodeoxyglucose PET (FDG PET) in epilepsy surgery. Epilepsy Res 2014;108(8):1306–1314

64. Gnatkovsky V, de Curtis M, Pastori C, et al. Biomarkers of epileptogenic zone defined by quantified stereo-EEG analysis. Epilepsia 2014;55(2):296–305

44. Salamon N, Kung J, Shaw SJ, et al. FDG-PET/MRI coregistration improves detection of cortical dysplasia in patients with epilepsy. Neurology 2008;71(20):1594–1601

65. Grinenko O, Li J, Mosher JC, et al. A fingerprint of the epileptogenic zone in human epilepsies. Brain 2018;141(1):117–131

45. Talanow R, Ruggieri P, Alexopoulos A, Lachhwani D, Wu G. PET manifestation in different types of pathology in epilepsy. Clin Nucl Med 2009;34(10):670–674 46. Jasper HH, Arfel-Capdeville G, Rasmussen T. Evaluation of EEG and cortical electrographic studies for prognosis of seizures following surgical excision of epileptogenic lesions. Epilepsia 1961;2:130–137 47. Bancaud J, Angelergues R, Bernouilli C, et al. Functional stereotaxic exploration (SEEG) of epilepsy. Electroencephalogr Clin Neurophysiol 1970;28(1):85–86 48. Guerrini R, Dravet C, Raybaud C, et al. Epilepsy and focal gyral anomalies detected by MRI: electroclinico-morphological correlations and follow-up. Dev Med Child Neurol 1992;34(8):706–718

66. Bartolomei F, Chauvel P, Wendling F. Epileptogenicity of brain structures in human temporal lobe epilepsy: a quantified study from intracerebral EEG. Brain 2008;131(Pt 7):1818–1830 67. Duchowny M, Jayakar P, Harvey AS, et al. Language cortex representation: effects of developmental versus acquired pathology. Ann Neurol 1996;40(1):31–38 68. Marusic P, Najm IM, Ying Z, et al. Focal cortical dysplasias in e ­loquent cortex: functional characteristics and correlation with MRI and histopathologic changes. Epilepsia 2002;43(1):27–32 69. Serafini S, Komisarow JM, Gallentine W, et al. Reorganization and stability for motor and language areas using cortical stimulation: case example and review of the literature. Brain Sci 2013;3(4):1597–1614

49. Raymond AA, Fish DR. EEG features of focal malformations of cortical development. J Clin Neurophysiol 1996;13(6):495–506

70. Najm IM, Tassi L, Sarnat HB, Holthausen H, Russo GL. Epilepsies associated with focal cortical dysplasias (FCDs). Acta Neuropathol 2014;128(1):5–19

50. Rosenow F, Lüders HO, Dinner DS, et al. Histopathological correlates of epileptogenicity as expressed by electrocorticographic spiking and seizure frequency. Epilepsia 1998;39(8):850–856

71. Kwon HE, Eom S, Kang HC, et al. Surgical treatment of pediatric focal cortical dysplasia: Clinical spectrum and surgical outcome. Neurology 2016;87(9):945–951

51. Otsubo H, Ochi A, Elliott I, et al. MEG predicts epileptic zone in lesional extrahippocampal epilepsy: 12 pediatric surgery cases. Epilepsia 2001;42(12):1523–1530

72. Wyllie E, Lachhwani DK, Gupta A, et al. Successful surgery for epilepsy due to early brain lesions despite generalized EEG findings. Neurology 2007;69(4):389–397

52. Najm IM, Bingaman WE, Lüders HO. The use of subdural grids in the management of focal malformations due to abnormal cortical development. Neurosurg Clin N Am 2002;13(1):87–92, viii–ix

73. Tassi L, Garbelli R, Colombo N, et al. Type I focal cortical dysplasia: surgical outcome is related to histopathology. Epileptic Disord 2010;12(3):181–191

53. Palmini A, Gambardella A, Andermann F, et al. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 1995;37(4):476–487

74. Krsek P, Pieper T, Karlmeier A, et al. Different presurgical characteristics and seizure outcomes in children with focal cortical dysplasia type I or II. Epilepsia 2009;50(1):125–137

54. Chassoux F, Devaux B, Landré E, et al. Stereoelectroencephalography in focal cortical dysplasia: a 3D approach to delineating the dysplastic cortex. Brain 2000;123(Pt 8):1733–1751

75. Palmini A, Holthausen H. Focal malformations of cortical development: a most relevant etiology of epilepsy in children. Handb Clin Neurol 2013;111:549–565

55. Boonyapisit K, Najm I, Klem G, et al. Epileptogenicity of focal malformations due to abnormal cortical development: direct electrocorticographic-histopathologic correlations. Epilepsia 2003; 44(1):69–76

76. Harvey AS, Mandelstam SA, Maixner WJ, et al. The surgically remediable syndrome of epilepsy associated with bottomof-­sulcus dysplasia. Neurology 2015;84(20):2021–2028

56. Brodtkorb E, Andersen K, Henriksen O, Myhr G, Skullerud K. Focal, continuous spikes suggest cortical developmental abnormalities. Clinical, MRI and neuropathological correlates. Acta Neurol Scand 1998;98(6):377–385 57. Sarkis RA, Jehi LE, Bingaman WE, Najm IM. Surgical outcome following resection of rolandic focal cortical dysplasia. Epilepsy Res 2010;90(3):240–247 58. Kutsy RL, Farrell DF, Ojemann GA. Ictal patterns of neocortical seizures monitored with intracranial electrodes: correlation with surgical outcome. Epilepsia 1999;40(3):257–266 59. Spencer SS, Guimaraes P, Katz A, Kim J, Spencer D. Morphological patterns of seizures recorded intracranially. Epilepsia 1992;33(3):537–545 60. Aubert S, Wendling F, Regis J, et al. Local and remote epileptogenicity in focal cortical dysplasias and neurodevelopmental tumours. Brain 2009;132(Pt 11):3072–3086 61. Perucca P, Dubeau F, Gotman J. Intracranial electroencephalographic seizure-onset patterns: effect of underlying pathology. Brain 2014;137(Pt 1):183–196 62. Lagarde S, Bonini F, McGonigal A, et al. Seizure-onset patterns in focal cortical dysplasia and neurodevelopmental tumors: Relationship with surgical prognosis and neuropathologic subtypes. Epilepsia 2016;57(9):1426–1435

77. Lortie A, Plouin P, Chiron C, Delalande O, Dulac O. Characteristics of epilepsy in focal cortical dysplasia in infancy. Epilepsy Res 2002;51(1–2):133–145 78. Chauvel P, McGonigal A. Emergence of semiology in epileptic seizures. Epilepsy Behav 2014;38(38):94–103 79. Bien CG, Benninger FO, Urbach H, Schramm J, Kurthen M, Elger CE. Localizing value of epileptic visual auras. Brain 2000;123 (Pt 2):244–253 80. Bonini F, McGonigal A, Trébuchon A, et al. Frontal lobe seizures: from clinical semiology to localization. Epilepsia 2014;55(2):264–277 81. Buoni S, Zannolli R, Miracco C, et al. Focal cortical dysplasia type 1b as a cause of severe epilepsy with multiple independent spike foci. Brain Dev 2008;30(1):53–58 82. Ohtsuka Y, Sato M, Sanada S, Yoshinaga H, Oka E. Suppression-burst patterns in intractable epilepsy with focal cortical dysplasia. Brain Dev 2000;22(2):135–138 83. Vasconcellos E, Wyllie E, Sullivan S, et al. Mental retardation in pediatric candidates for epilepsy surgery: the role of early seizure onset. Epilepsia 2001;42(2):268–274 84. Korman B, Krsek P, Duchowny M, Maton B, Pacheco-Jacome E, Rey G. Early seizure onset and dysplastic lesion extent independently disrupt cognitive networks. Neurology 2013;81(8): 745–751

41  Focal Cortical Dysplasia: Histopathology, Neuroimaging, and Electroclinical Presentation 85. Hildebrandt M, Pieper T, Winkler P, Kolodziejczyk D, Holthausen H, Blümcke I. Neuropathological spectrum of cortical dysplasia in children with severe focal epilepsies. Acta Neuropathol 2005;110(1):1–11

97. Fernández S, Donaire A, Serès E, et al. PET/MRI and PET/MRI/ SISCOM coregistration in the presurgical evaluation of refractory focal epilepsy. Epilepsy Res 2015;111:1–9

86. Loddenkemper T, Holland KD, Stanford LD, Kotagal P, Bingaman W, Wyllie E. Developmental outcome after epilepsy surgery in infancy. Pediatrics 2007;119(5):930–935

98. Perissinotti A, Setoain X, Aparicio J, et al. Clinical role of subtraction ictal SPECT coregistered to MR imaging and (18)F-FDG PET in pediatric epilepsy. J Nucl Med 2014;55(7):1099–1105

87. Loddenkemper T, Cosmo G, Kotagal P, et al. Epilepsy surgery in children with electrical status epilepticus in sleep. Neurosurgery 2009;64(2):328–337, discussion 337

99. Cossu M, Lo Russo G, Francione S, et al. Epilepsy surgery in children: results and predictors of outcome on seizures. Epilepsia 2008;49(1):65–72

88. Wyllie E, Comair YG, Kotagal P, Raja S, Ruggieri P. Epilepsy surgery in infants. Epilepsia 1996;37(7):625–637

100. Teutonico F, Mai R, Veggiotti P, et al. Epilepsy surgery in children: evaluation of seizure outcome and predictive elements. Epilepsia 2013;54(Suppl 7):70–76

89. Lee YJ, Kang HC, Lee JS, et al. Resective pediatric epilepsy surgery in Lennox-Gastaut syndrome. Pediatrics 2010;125(1):e58–e66 90. Hallbook T, Ruggieri P, Adina C, et al. Contralateral MRI abnormalities in candidates for hemispherectomy for refractory epilepsy. Epilepsia 2010;51(4):556–563 91. Moosa AN, Jehi L, Marashly A, et al. Long-term functional outcomes and their predictors after hemispherectomy in 115 children. Epilepsia 2013;54(10):1771–1779 92. Barba C, Mai R, Grisotto L, et al. Unilobar surgery for symptomatic epileptic spasms. Ann Clin Transl Neurol 2016;4(1):36–45 93. Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug resistant epilepsy: consensus proposal by the ad hoc Task Force of the ILAE Commission on Therapeutic Strategies. Epilepsia 2010;51(6):1069–1077 94. Mostofi K, Marnet D, Derambure P. Surgical management of focal cortical dysplasia. Acta Neurol Belg 2011;111(4):310–316 95. Murakami H, Wang ZI, Marashly A, et al. Correlating magnetoencephalography to stereo-electroencephalography in patients undergoing epilepsy surgery. Brain 2016;139(11):2935–2947 96. Kudr M, Krsek P, Maton B, et al. Ictal SPECT is useful in localizing the epileptogenic zone in infants with cortical dysplasia. ­Epileptic Disord 2016;18(4):384–390

101. Sandok EK, Cascino GD. Surgical treatment for perirolandic lesional epilepsy. Epilepsia 1998;39(Suppl 4):S42–S48 102. Behdad A, Limbrick DD Jr, Bertrand ME, Smyth MD. Epilepsy surgery in children with seizures arising from the rolandic cortex. Epilepsia 2009;50(6):1450–1461 103. Gonzalez-Martinez J, Najm IM. Indications and selection criteria for invasive monitoring in children with cortical dysplasia. Childs Nerv Syst 2014;30(11):1823–1829 104. Dorfmüller G, Ferrand-Sorbets S, Fohlen M, et al. Outcome of surgery in children with focal cortical dysplasia younger than 5 years explored by stereo-electroencephalography. Childs Nerv Syst 2014;30(11):1875–1883 105. Chassoux F, Landré E, Mellerio C, et al. Type II focal cortical dysplasia: electroclinical phenotype and surgical outcome related to imaging. Epilepsia 2012;53(2):349–358 106. Bonini F, McGonigal A, Scavarda D, et al. Predictive factors of surgical outcome in frontal lobe epilepsy explored with stereoelectroencephalography. Neurosurgery 2018;83(2):217–225 107. Guerrini R, Duchowny M, Jayakar P, et al. Diagnostic methods and treatment options for focal cortical dysplasia. Epilepsia 2015;56(11):1669–1686

413

42

  Surgical Approaches in Cortical Dysplasia Jeffrey Bolton, Sanjay P. Prabhu, Eun-Hyoung Park, Scellig S. Stone, and Joseph R. Madsen

Summary The surgical management of focal cortical dysplasia (FCD) lesions has become one of the most important growth areas in the treatment of refractory epilepsy in children across the age spectrum. Progress in recognition and characterization of these regions of abnormal cortex, including advances in ­magnetic resonance imaging (MRI) and interpretation, and improved preoperative mapping of epileptogenic and functional regions have propelled the growth. Ongoing challenges include identification of the distinction between pathological and potentially eloquent tissue, and the best noninvasive and invasive surgical mapping strategies. Improvement in FCD surgery will be critical to the delivery of surgical therapy to increasing numbers of children. Keywords:  focal cortical dysplasia, electrocorticography, invasive EEG, stereo-EEG, Granger causality, laser interstitial thermal therapy

„„ Introduction Focal cortical dysplasia (FCD) is one of the most common etiologies leading to epilepsy surgery in the pediatric population.1,​2 Presentation is often in the first few years of life, leading to severe refractory focal epilepsy and cognitive impairments.3 FCD is subdivided into three broad categories based on pathologic findings. FCD type I is characterized by abnormal cortical lamination, which includes abnormal radial cortical ­lamination (FCD type Ia), abnormal tangential cortical lamination (FCD type Ib), and a third subtype with both abnormal radial and tangential cortical lamination (FCD type Ic). FCD type II is a malformation defined by disrupted cortical lamination and specific cytological abnormalities, which differentiate FCD type IIa (dysmorphic neurons without balloon cells) from FCD type IIb (dysmorphic neurons and balloon cells). FCD type III refers to cortical lamination abnormalities associated with a principal lesion, usually adjacent to or affecting the same cortical area/lobe.4 Four subtypes of type III FCD are described, including: FCD type IIIa, associated with hippocampal sclerosis; FCD type IIIb, associated with tumors; FCD type IIIc, associated with

­ ascular malformations; and FCD type IIId, associated with any v other principal lesion acquired during early life. The detailed anatomical features of the different types and subtypes of FCD can be seen using various magnetic resonance imaging (MRI) sequences. Features of FCD on MRI include cortical thickening (which should be confirmed in at least two planes and on two different imaging sequences), blurring of the gray and white matter junction, abnormal cortical signal (on one or more imaging sequences), T2/FLAIR (fluid-attenuated inversion recovery) hyperintensity in the cortex and adjacent subcortical white matter, and T1 shortening in the cortex and abnormal sulcal/gyral pattern. MRI is able to show abnormalities in the majority of type II dysplasias but only in some of the type I cortical dysplasias. FCD type IIb (with balloon cells) is often characterized by hypo-, de-, or dysmyelination in the subcortical white matter. This manifests on MRI as blurring of the gray and white matter junction, mimicking cortical thickening. Often, a band of hyperintensity on T2-weighted and FLAIR images can be seen extending from the gray and white matter interface to the surface of the ventricles, called the “transmantle sign.” In order to optimally visualize this often-subtle finding, multiplanar thin section images should be performed. On 3-tesla (3T) scans, the volumetric FLAIR sequence with multiplanar reformats is an excellent sequence to look for the white matter hyperintensity characteristic of type II FCDs. Note that while these FLAIR images are sensitive for the white matter abnormality, they are less sensitive for assessment of extent of cortical abnormality. FCD type IIa is more difficult to detect on in vivo MRI than FCD type IIb. Another point to note is that the “transmantle sign” is not specific for FCD type IIb, and can be seen in other types of FCDs as well. Early identification and surgical resection of FCD provides the best chance at both seizure freedom and quality of life in this population.3,​5,​6 Because outcome is linked to completeness of resection, it is crucial that the extent of the lesion is well defined. As described in Chapter 41, this may be accomplished during the presurgical workup using high-resolution imaging (3T MRI with protocols tailored for epilepsy lesion detection), functional studies such as s­ingle-photon emission computed tomography (SPECT) and positron emission tomography (PET), and advanced electrophysiology techniques, magnetoencephalography (MEG), and source a ­ nalysis.7,​8 In addition to defining

42  Surgical Approaches in Cortical Dysplasia the margins of the d ­ ysplasia, its relationship to eloquent cortex must also be factored into surgical planning. This can be accomplished with v ­ arious noninvasive modalities such as functional MRI (fMRI) and transcranial magnetic stimulation (TMS). Having stated this, it should be noted that all FCDs might not be identified preoperatively. Furthermore, boundaries of the lesion may not be clearly defined even when high field strength scans with multichannel coils are used. The complexity of each case depends on numerous factors, but ultimately can be distilled down to two key factors: (1) how well-defined the lesion is on imaging studies; and (2) whether the lesion involves eloquent cortex. With increasing complexity of each case come options for different s­ urgical approaches. We will use real case examples to outline the range of surgical techniques available for treatment of FCD in the pediatric ­population.

„„ Case 1: Well-Defined Focal Cortical Dysplasia Distant from Eloquent Cortex Patient 1 is a 7-year-old right-handed boy with seizures that started around the age of 3. His seizure semiology included pulling of the mouth to the left followed by bilateral arm twitching, correlating with a right frontal ictal onset pattern on scalp electroencephalography (EEG). 3T MRI revealed a discrete area of FCD in the right frontopolar region (Fig. 42.1). Additional studies, including PET, SPECT, MEG, and source analysis, were also concordant. He was taken to the operating room (OR) for a single-stage resection of the dysplasia with frameless neuronavigation guidance, intraoperative ultrasound, and pre- and postresection electrocorticography (ECOG). An MRI obtained in the OR ensured complete resection of the visible dysplasia. Pathology was consistent with FCD type IIb. Case 1 is an excellent example of the ideal surgical candidate with FCD. The lesion was well defined with concordant preoperative data in a location remote from eloquent cortex. Because the success of this surgery is largely dependent on total resection of the dysplasia, intraoperative imaging guidance with stereotactic systems plays a key role. Depending on the location of the dysplasia, the surgeon may not be able to visualize any abnormality on the cortical surface. Many FCDs involve the depth of the sulcus and can be difficult to locate without the use of advanced OR image guidance.9 Intraoperative frameless neuronavigation systems are a useful means of anatomical guidance that register preoperatively acquired MRI and/or computed tomography (CT) imaging in three-dimensional (3D) space. Intraoperative ultrasound can also play a role in identifying a deep lesion, and has the advantage of providing real-time imaging feedback (Fig. 42.2).10,​11 Once the resection is complete, intraoperative MRI is another tool to aid the surgeon in assessing the completeness of the resection.12,​13 The data are mixed on the utility of intraoperative ECOG for single-stage resections.14 In our center, we routinely use pre- and postresection ECOG to help decide the margins of resection. While we previously utilized s­ ingle strips or small

Fig. 42.1  T2-weighted axial MRI showing dysplasia in right frontopolar region (indicated by arrow). Pathology was consistent with focal cortical dysplasia (FCD) type IIb (Case 1).

subdural electrode grids for i­ntraoperative ECOG, we have increasingly found recordings with larger grids and sometimes stereotactically placed depth electrodes to be helpful in preempting the question of whether observed epileptiform activity is spatially limited or temporally intermittent. In some fortunate instances, an electrographic seizure may even be observed. Once the resection is completed, the electrodes are placed around the margins, again looking for any epileptiform discharges. It is typically not advisable to “chase” epileptiform findings, which are fairly distant from the resection cavity, as these may represent part of the more extended seizure network, and not necessarily the epileptic onset zone.14,​15

„„ Case 2: Poorly Defined Lesion near Eloquent Cortex Patient 2 is a 13-year-old right-handed female with seizures that started at age 10, consisting of initial scream, fearful look, aphasia, and right-hand automatisms with decreased responsiveness. Her MRI was inconclusive, but suggestive of FCD involving the anterior left temporal lobe. Preoperative studies suggested seizures arising from left temporal lobe as well as both expressive and receptive language lateralized to the left hemisphere. Being the dominant temporal lobe without well-defined margins of the dysplasia, invasive monitoring with subdural electrodes was pursued. For the

415

416

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 42.2  Intraoperative photos (a–d) and an intraoperative ultrasound image (locations of two depth electrodes ID and SD are indicated by circles) (e) of a patient who had subdural grids/ strips and two depths (10 contacts for superior depth [SD] and 10 contacts for inferior depth [ID]) for intracranial EEG recordings. (f) Subdural grid electrodes and superior depth electrodes (SD) in relation to dysplasia are schematically shown.

­neurophysiologist to interpret the invasive EEG data, computer 3D reconstructions of the implanted electrodes are extremely helpful (Fig. 42.3). This allows for visualization of the relationship of the individual electrodes to the patient’s brain anatomy and cortical malformation. Once the ictal onset zone (IOZ) is defined, cortical mapping can then be performed to exclude any overlap between the IOZ and eloquent cortex.16,​17 In Case 2, ictal onsets were recorded from contacts in the anterior and mesial aspects, distant from the mapped language function (Fig. 42.4a,b). A resection was completed with no postoperative deficit (Fig. 42.4c). Pathology revealed type IIa dysplasia. The International League Against Epilepsy (ILAE) has recently outlined indications for invasive monitoring that include: (1) defining the epileptic onset zone precisely when noninvasive data are inconclusive; (2) resolving divergence of noninvasive data pointing to two or more regions; (3) mapping eloquent cortical function precisely; and (4) secondary indications.18,​19 Case 2 is a good example of a situation with poorly defined lesional margins (indication 1) as well as close proximity to eloquent cortex (indication 3). The modality of invasive monitoring largely depends on the epilepsy center’s experience. Some centers are more comfortable and experienced using subdural electrodes, while others may be more inclined to use s­tereotactic depth

electrode placement and recording (stereo-EEG, or SEEG).19,​20 There are unique advantages of each, largely depending on the specific case. When the dysplasia is poorly defined on preoperative MRI, it can be difficult for the surgeon to know the extent of adequate resection margins. Invasive monitoring provides electrophysiological confirmation of the IOZ.21 The gold standard to identifying the IOZ is recording the patient’s habitual seizures during the invasive monitoring period. This often requires antiepileptic medication withdrawal in order to capture an adequate number of typical seizures. Because of the risk of status epilepticus and surgical complications, such monitoring should be conducted under the close supervision of an epilepsy team in either an intensive care unit or an experienced epilepsy monitoring unit. To minimize complications such as infection, bleeding, and edema, invasive monitoring is typically limited to 7 days or less, allowing scheduling of the second procedure into an appropriate elective surgical opening. Extending the monitoring session, however, is sometimes necessary if adequate seizures have yet to be captured or mapping is incomplete. In current practice, ictal invasive EEG data are mainly used to specify the seizure origin, which is assumed to be the causal part of the seizure network. However, interictal recordings account for the vast majority of data obtained during invasive EEG and

42  Surgical Approaches in Cortical Dysplasia At the conclusion of the monitoring phase, a detailed resection plan should be formulated between the surgeon and ­epileptologist involved with the case. In the OR, prior to the resection, the location of the electrodes should be confirmed as well as noting any accumulation of blood products beneath the grids. The subdural electrodes can remain in place during the resection as an additional anatomic landmark. Postresection ECOG may be performed to assess for any residual epileptiform areas around the margins of the resection.

„„ Case 3: Discrete Deep Lesion Close to Eloquent Cortex

Fig. 42.3  (a) Subdural grid electrode contacts superimposed upon 3D reconstructed brain cortical surface and 3D volume rendered CT (Case 2). (b) Grid labels with the size (AT, PT, FR, and MT; 4 × 8, 2 × 8, 2 × 8, and 1 × 8, respectively) and the anatomical locations of four grids (indicated by dots with colors) are shown.

there would be advantages in deriving information about causal regions from interictal data as well. One of the advantages is that interictal data can be acquired at almost any time, including immediately in the OR, potentially aiding in the identification and visualization of causal networks in a relatively rapid timeframe that does not rely on ictal events. With newer computational techniques such as Granger causality, high-frequency oscillation, and source analysis, additional information about the topography of the epileptic network can be constructed.22,​23,​24 For example, recently it has been shown that Granger causality analysis using 20 minutes of interictal invasive EEG data recorded on the day of subdural grid placement can generate maps statistically comparable with IOZ maps and actual resection maps.24 Fig. 42.5 shows Granger causality maps obtained for Case 2, one of the cases published in Park and Madsen.24

Patient 3 is an 18-year-old right-handed young man with a long history of focal motor seizures involving the right face and hand, oftentimes secondarily generalizing. After numerous normal MRIs, a subtle FCD was appreciated in the precentral gyrus, at the depth of a sulcus. Extensive preoperative motor mapping with fMRI and TMS demonstrated typical contralateral motor function in the left precentral gyrus. Because of proximity to the motor cortex (hand/face region), invasive monitoring with motor mapping was performed. A combination of subdural grid electrodes and depth electrodes were used in order to be able to map motor cortex on the lateral surface, as well as record the s­ eizure onset zone at the depth of the sulcus (Fig. 42.6 and Fig. 42.7). After several typical seizures were captured, motor mapping was performed allowing for a safe approach to the deep-seated dysplasia. He underwent a complete resection of the lesion with minimal postoperative motor deficits. Case 3 demonstrates the challenges when a clearly defined dysplasia is located deep to eloquent cortex. Though not sufficient, preoperative studies localizing eloquent cortex are extremely important when planning such a case. fMRI can aide in the localization of language (both expressive and receptive) and motor and visual function. It has the advantage of being superimposed on the patient’s MRI, allowing for easy visualization of the dysplasia’s proximity to the BOLD (blood oxygenation level dependent) signal for a specific task. TMS is ideal for motor mapping, as it has better spatial resolution than fMRI and can also be superimposed on the patient’s MRI.25 With good preoperative data, an implantation schema can be planned to ensure adequate cortical mapping combined with seizure localization.20 In cases with lesions deep to the cortical surface, a combination of both subdural grids and depth electrodes may ­provide the most comprehensive data. The subdural grid allows for high-­resolution mapping of the cortical surface. This is done in a controlled setting outside of the OR, oftentimes with the help of a neuropsychologist, depending on the testing being performed. With careful stimulation and mapping, the neurophysiologist may be able to provide an area of cortex safe for accessing the deeper lesion. A series of depth electrodes targeting the deep dysplasia can then provide the surgeon with a good sense of the extent of the epileptic onset zone. Newer techniques allow creation of a single multilayered data set with overlay of multiple modalities including the CT with the sEEG electrodes, structural MRI, PET, TMS maps, fMRI, and diffusion ­tractography data.

417

418

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 42.4  Map of electrodes identified as ictal onset and early spread (a), receptive language dysfunction area noted with cortical stimulation (b), and resected area (c) displayed on 3D volume rendered CT (Case 2).

This allows visualization of relationships between the structural lesion, ictally active electrodes, eloquent areas, and vital white matter tracts. Combining these data sets, cortical function map and the seizure onset zone allows for as complete a resection as possible with minimal deficits.16 In the unfortunate event of significant overlap between eloquent cortex and seizure onset zone, the next steps are less clear. A limited resection avoiding any eloquent cortex may

be performed if there is an incomplete overlap of IOZ and eloquent cortex. Such an approach may provide only partial seizure control postoperatively, but for some patients it may at least allow for a reduction in antiepileptic drugs and an improvement in quality of life. Opting for no resection may be the prudent choice when there is near-complete overlap. In these cases, there is now the option for responsive neural stimulation, which is approved for patients 18 years or o ­ lder.26

42  Surgical Approaches in Cortical Dysplasia

Fig. 42.5  Granger causality map: causal connectivity map (a) and causal nodes map (b) produced using interictal invasive EEG data obtained from the day of subdural grids implantation for Case 2 (the results of this case have been published in Park and Madsen24).

In some severe epilepsy cases, the patient and family may be willing to trade some form of deficit for seizure freedom. Partial visual field cuts or weakness of the nondominant hand may be tolerable outcomes for certain patients in exchange for seizure freedom.

„„ Case 4: Deep Lesion with Unclear Margins, Away from Eloquent Cortex Patient 4 is a 12-year-old right-handed young lady with refractory epilepsy since the age of 14 months. Her seizures localized to the left frontal lobe, where a large, poorly defined dysplasia was seen, involving the orbital and mesial frontal regions. Because of the poorly demarcated boundaries and location of the lesion, sEEG was pursued (Fig. 42.8). Several typical seizures were captured during her invasive monitoring session and a resection was planned based on the ictal ­network. This case demonstrates the benefit of sEEG in patients with hard to reach and ill-defined FCDs. The majority of FCD is

­extratemporal and may be difficult to reach with subdural electrodes. Dysplasias in the cingulate, orbital frontal lobe, or insula, for example, may be better sampled with sEEG electrodes.27 In addition to gaining access to more deep-seated and mesial structures, sEEG also allows for more broad coverage over multiple lobes and even bilateral coverage.28 Risks of sEEG implantation are similar to subdural grids, with bleeding being the most worrisome. This can be avoided with detailed vascular imaging preoperatively. Compared to subdural electrodes, which require a craniotomy, sEEG electrodes are well tolerated by the patient and can be removed at the bedside without an additional trip to the OR.29,​30 Some centers prefer to explant the electrodes and then wait a period of time before proceeding with a resection, mainly out of concern for increased risk of infection. It is our experience that patients tolerate going straight to resection following the invasive monitoring without any additional risk. This gives the added benefit of only one hospitalization, and in some cases electrodes can be temporarily left in situ and used as intraoperative reference points during a resection. Cortical mapping can be done with sEEG in a similar fashion as with subdural electrodes. There may be slight limitations in mapping capabilities depending on placement of the electrodes; however, this can be minimized with comprehensive preoperative planning.

419

420

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 42.6  Operative photos of Case 3 showing brain surface before depths implantation (a) and depth electrodes targeting focal cortical dysplasia (FCD) at the bottom of a sulcus dysplasia (cortical entry indicated by circle) (b) and photo with subdural grid overlying depth electrodes (c).

„„ Case 5: Deep Lesion, Clear Margins Patient 5 is a 14-year-old right-handed young man with a right parasagittal parietal FCD leading to refractory seizures. Preresection data localized the seizure onsets to the area of d ­ ysplasia,

Fig. 42.7  (a–c) Subdural grid and depth electrodes implanted for Case 3 displayed on 3D volume rendered CT scans.

which was found to contain no eloquent cortex. In part due to the location of the lesion and the family’s wish for a minimally invasive approach, he underwent magnetic resonance–­guided

42  Surgical Approaches in Cortical Dysplasia

Fig. 42.8  T1-weighted MRIs showing planned seven SEEG depths trajectories (indicated with multiple colors) for Case 4 (the four images were generated by the StealthStation).

Fig. 42.9  Pre- (a) and post-MR-LITT images (b) obtained from Case 5 which include right parasagittal T2-weighted MRI showing MR-LITT probes inserted into region of dysplasia (a) and parasagittal T1-weighted gadolinium-enhanced MRI post-MR-LITT showing ring-enhancing zone of ablation (b).

laser interstitial thermal therapy (MR-LITT) of the dysplasia (Fig. 42.9). MR-LITT is a relative new technique used in the surgical treatment of dysplasia.31 One or more liquid-cooled guide tube(s)

containing a diode laser(s) can be stereotactically passed to target(s) at surgery, using software allowing the surgeon to plan the trajectory (trajectories) in such a way that the defined target(s) may be safely encompassed by the predicted ablation zone(s)

421

422

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches (generally taking the form of cylinders about 15 mm in diameter and of varying length around the diffusing tip of the laser fiber). Once the laser or lasers are placed, the laser is activated and generates thermal energy, monitored by MR-thermography imaging, and one can ablate regions of cortical dysplasia in a relatively controlled fashion, sparing adjacent vessels (which act as heat sinks) and white matter tracts (which can be kept outside of the region of lesional heating). This approach can be very useful for patients with difficult to access lesions, such as those in the insula. It also has the benefit of being minimally invasive, avoiding the need for a traditional craniotomy.32,​33 Optimal candidates are those with well-defined lesions limited to a single gyrus, since cerebrospinal fluid–­containing sulci prevent heat transmission to neighboring gyri. A larger or more complex-shaped dysplasia may require multiple laser trajectories. In cases where some degree of invasive monitoring is needed, it is possible to combine sEEG with MR-LITT.

„„ Outcomes In general, patients undergoing resection of FCD have favorable outcomes. When considering based on subtype, FCD type IIb has been shown to have the best prognosis followed by type

IIa and then type I. It has been postulated that FCD type I may be more widespread than what is visualized radiographically, thus risking incomplete resections. Seizure-free reported outcomes are from 40 to 70% and are inversely proportional to the length of follow-up.34,​35 Completeness of the resection and, in some cases, the presence of an MRI lesion are predictors of good surgical outcome. In cases where seizures continue, it is often worth considering a second surgery, particularly if there is residual dysplasia visible on postoperative imaging. We have found that fusing and co-registering postoperative scans with preoperative studies can be helpful in delineating residual ­dysplasia from postsurgical changes.

„„ Conclusion Treating patients with FCD can be a challenging but ultimately rewarding experience. Each case must be approached in a systematic fashion in order to formulate the ideal surgical plan. As technology has advanced, there are numerous preoperative and perioperative studies available to help define the optimal resection. The future likely will continue to move toward less invasive techniques, with regard to intracranial monitoring of dysplasia as well as ablation.

References 1. Fauser S, Huppertz HJ, Bast T, et al. Clinical characteristics in focal cortical dysplasia: a retrospective evaluation in a series of 120 patients. Brain 2006;129(Pt 7):1907–1916 2. Jayakar A, Bolton J. Pediatric epilepsy surgery. Curr Neurol Neurosci Rep 2015;15(6):31 3. Ramantani G, Kadish NE, Strobl K, et al. Seizure and cognitive outcomes of epilepsy surgery in infancy and early childhood. Eur J Paediatr Neurol 2013;17(5):498–506 4. Blümcke I, Thom M, Aronica E, et al. The clinicopathologic spectrum of focal cortical dysplasias: a consensus classification proposed by an ad hoc Task Force of the ILAE Diagnostic Methods Commission. Epilepsia 2011;52(1):158–174 5. Chen HH, Chen C, Hung SC, et al. Cognitive and epilepsy outcomes after epilepsy surgery caused by focal cortical dysplasia in children: early intervention maybe better. Childs Nerv Syst 2014;30(11):1885–1895 6. Kloss S, Pieper T, Pannek H, Holthausen H, Tuxhorn I. Epilepsy surgery in children with focal cortical dysplasia (FCD): results of long-term seizure outcome. Neuropediatrics 2002;33(1):21–26 7. Papanicolaou AC, Rezaie R, Narayana S, et al. On the relative merits of invasive and non-invasive pre-surgical brain mapping: new tools in ablative epilepsy surgery. Epilepsy Res 2018;142:153–155

12. Sommer B, Grummich P, Coras R, et al. Integration of functional neuronavigation and intraoperative MRI in surgery for drugresistant extratemporal epilepsy close to eloquent brain areas. Neurosurg Focus 2013;34(4):E4 13. Sacino MF, Ho CY, Murnick J, et al. Intraoperative MRI-guided resection of focal cortical dysplasia in pediatric patients: technique and outcomes. J Neurosurg Pediatr 2016;17(6):672–678 14. Greiner HM, Horn PS, Tenney JR, et al. Should spikes on post-­ resection ECoG guide pediatric epilepsy surgery? Epilepsy Res 2016;122:73–78 15. Palmini A, Gambardella A, Andermann F, et al. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 1995;37(4):476–487 16. Terra VC, Thomé U, Rosset SS, et al. Surgery for focal cortical dysplasia in children using intraoperative mapping. Childs Nerv Syst 2014;30(11):1839–1851 17. Yang PF, Zhang HJ, Pei JS, et al. Intracranial electroencephalography with subdural and/or depth electrodes in children with epilepsy: techniques, complications, and outcomes. Epilepsy Res 2014;108(9):1662–1670 18. Jayakar P, Gotman J, Harvey AS, et al. Diagnostic utility of invasive EEG for epilepsy surgery: indications, modalities, and techniques. Epilepsia 2016;57(11):1735–1747

8. Guerrini R, Duchowny M, Jayakar P, et al. Diagnostic methods and treatment options for focal cortical dysplasia. Epilepsia 2015;56(11):1669–1686

19. Brna P, Duchowny M, Resnick T, Dunoyer C, Bhatia S, Jayakar P. The diagnostic utility of intracranial EEG monitoring for epilepsy surgery in children. Epilepsia 2015;56(7):1065–1070

9. Harvey AS, Mandelstam SA, Maixner WJ, et al. The surgically remediable syndrome of epilepsy associated with bottom-of-­ sulcus dysplasia. Neurology 2015;84(20):2021–2028

20. Nowell M, Rodionov R, Zombori G, et al. Utility of 3D multimodality imaging in the implantation of intracranial electrodes in epilepsy. Epilepsia 2015;56(3):403–413

10. Miller D, Knake S, Menzler K, Krakow K, Rosenow F, Sure U. Intraoperative ultrasound in malformations of cortical development. Ultraschall Med 2011;32(Suppl 2):E69–E74

21. Widdess-Walsh P, Jeha L, Nair D, Kotagal P, Bingaman W, Najm I. Subdural electrode analysis in focal cortical dysplasia: predictors of surgical outcome. Neurology 2007;69(7):660–667

11. Chan HW, Pressler R, Uff C, et al. A novel technique of detecting MRI-negative lesion in focal symptomatic epilepsy: intraoperative ShearWave elastography. Epilepsia 2014;55(4):e30–e33

22. Jacobs J, Zijlmans M, Zelmann R, et al. High-frequency electroencephalographic oscillations correlate with outcome of epilepsy surgery. Ann Neurol 2010;67(2):209–220

42  Surgical Approaches in Cortical Dysplasia 23. Jacobs J, Levan P, Châtillon CE, Olivier A, Dubeau F, Gotman J. High frequency oscillations in intracranial EEGs mark epileptogenicity rather than lesion type. Brain 2009;132(Pt 4):1022–1037 24. Park EH, Madsen JR. Granger causality analysis of interictal iEEG predicts Seizure focus and ultimate resection. Neurosurgery 2018;82(1):99–109 25. Picht T, Krieg SM, Sollmann N, et al. A comparison of language mapping by preoperative navigated transcranial magnetic stimulation and direct cortical stimulation during awake surgery. Neurosurgery 2013;72(5):808–819 26. Jobst BC, Kapur R, Barkley GL, et al. Brain-responsive neurostimulation in patients with medically intractable seizures arising from eloquent and other neocortical areas. Epilepsia 2017;58(6):1005–1014 27. Dylgjeri S, Taussig D, Chipaux M, et al. Insular and insulo-opercular epilepsy in childhood: an SEEG study. Seizure 2014;23(4):300–308 28. Gonzalez-Martinez J, Bulacio J, Alexopoulos A, Jehi L, Bingaman W, Najm I. Stereoelectroencephalography in the “difficult to localize” refractory focal epilepsy: early experience from a North American epilepsy center. Epilepsia 2013;54(2):323–330 29. Dorfmüller G, Ferrand-Sorbets S, Fohlen M, et al. Outcome of surgery in children with focal cortical dysplasia younger than

5 years explored by stereo-electroencephalography. Childs Nerv Syst 2014;30(11):1875–1883 30. Cossu M, Cardinale F, Castana L, Nobili L, Sartori I, Lo Russo G. Stereo-EEG in children. Childs Nerv Syst 2006;22(8):766–778 31. Lewis EC, Weil AG, Duchowny M, Bhatia S, Ragheb J, Miller I. MR-guided laser interstitial thermal therapy for p ­ediatric drug-resistant lesional epilepsy. Epilepsia 2015;56(10): 1590–1598 32. Patel P, Patel NV, Danish SF. Intracranial MR-guided laser-induced thermal therapy: single-center experience with the Visualase thermal therapy system. J Neurosurg 2016;125(4):853–860 33. Waseem H, Vivas AC, Vale FL. MRI-guided laser interstitial thermal therapy for treatment of medically refractory non-lesional mesial temporal lobe epilepsy: outcomes, complications, and current limitations: a review. J Clin Neurosci 2017;38:1–7 34. Hudgins RJ, Flamini JR, Palasis S, Cheng R, Burns TG, Gilreath CL. Surgical treatment of epilepsy in children caused by focal cortical dysplasia. Pediatr Neurosurg 2005;41(2):70–76 35. Park CK, Kim SK, Wang KC, et al. Surgical outcome and prognostic factors of pediatric epilepsy caused by cortical dysplasia. Childs Nerv Syst 2006;22(6):586–592

423

43

  Tuberous Sclerosis Complex Jurriaan M. Peters and Mustafa Şahin

Summary Although the multilesional nature of tuberous sclerosis complex (TSC) poses a formidable challenge in pediatric epilepsy surgery, new scientific insights, clinical advances, and technological innovations have changed the field dramatically. Moreover, a direct developmental impact is seen in patients with good surgical outcomes, and early and aggressive surgery may increase that developmental gain. Newer techniques including stereoelectroencephalography and magnetic resonance thermography–guided laser-induced thermal therapy are ­ increasingly used but still need clinical validation. This chapter introduces TSC and its neuroimaging findings. Next, it discusses key challenges specifically associated with pediatric epilepsy surgery in TSC, including the identification of an epileptogenic tuber among many other lesions, the fuzzy boundaries of tubers, and the varying surgical objectives of surgery in challenging cases. A detailed overview of the presurgical workup in TSC is provided, including some newer techniques such as diffusion tensor imaging, α-[11C] methyl-l-tryptophan positron emission tomography, and high-frequency oscillations. Finally, several surgical controversies are highlighted, including whether seizures originate in tubers versus the perituber rim, the potential of seizures to propagate to other tubers, and minimally invasive techniques versus maximally aggressive resections, among other topics. Keywords:  tuberous sclerosis complex, children, epilepsy, seizures, epilepsy surgery, EEG, presurgical workup

„„ Introduction Tuberous sclerosis complex (TSC) is a genetic neurocutaneous disorder, with a prevalence of approximately 1:6,000. Inherited autosomal dominant mutations (~30%) and sporadic mutations (~70%) lead to inactivation of the tumor suppressor genes TSC1 (on chromosome band 9q34) and TSC2 (on chromosome band 16p13.3) and can be identified in 70 to 90% of patients who meet the clinical criteria of TSC.1,​2 Inactivation of TSC1 or TSC2 genes leads to pathologically enhanced activity of the mechanistic target of rapamycin (mTOR) pathway, with subsequent disinhibition of protein

synthesis and cell growth.3 This results in benign hamartomatous malformations in multiple organs, including the heart (rhabdomyoma), kidneys (angiomyolipoma), the lungs (lymphangioleiomyomatosis), retina (hamartoma), and brain (tubers). In the brain, aberrant cellular proliferation, differentiation, and migration lead to various malformations including cortical tubers. These dysplastic and disorganized lesions on the border of gray and white matter are present in more than 80% of patients with TSC, and contain cells with an ambiguous phenotype of both astrocytic and neuronal l­ ineage.4 In addition, white matter radial migration lines, subependymal nodules, and subependymal giant cell a ­ strocytomas can be seen (Fig. 43.1). TSC is diagnosed on the basis of major and minor clinical criteria, with three of the major criteria being based on neuroimaging findings. In 2012, the International Tuberous Sclerosis Consensus updated the TSC diagnostic criteria from 1998. TSC can now be diagnosed via genetic testing if a pathogenic variant in TSC1 or TSC2 genes is found; see Box 43.1.5

„„ Epilepsy and Neurodevelopment Neurological symptoms are the most disabling, as they appear in childhood and include epilepsy, developmental delays and intellectual disability (ID), neurodevelopmental disabilities including behavioral problems, and autism spectrum disorder (ASD). Epilepsy occurs in 80 to 90% and typically develops in the first year of life, and in 50 to 80% of patients, it is refractory to antiepileptic drugs. Seizures are particularly devastating in childhood as they interfere with early and critical stages of neurodevelopment, and affect long-term neurological outcome. The presence of infantile spasms, early age of onset of epilepsy, and seizure frequency are associated with developmental delay, ID, and ASD.6,​7,​8 Improved long-term cognition with early treatment has been reported in small retrospective series, and delayed treatment has been associated with worse neurological­ outcomes.9,​10 While the clinical experience in individual cases is that ­earlier surgery is associated with earlier mitigation of detrimental effects from medications and seizures on development, and therefore with a greater “bang for your buck” in terms of improved neurodevelopmental outcomes, there are no prospective studies of early surgery in TSC specifically.

43  Tuberous Sclerosis Complex

Fig. 43.1  Conventional MRI findings in tuberous sclerosis complex. (a, b) Axial fluid attenuation inversion recovery images. Both patients have subcortical tubers (arrows) of comparable size and distribution (not all tubers shown in current plane), but the first patient (a) has severe autism and no active seizure disorder and is nonverbal, while the other patient (b) has mild motor and language delays, no autism, and refractory seizures despite multiple antiepileptic drugs. (c, d) Axial fluid attenuation inversion recovery images. Hypointense, partially calcified subependymal nodules are seen lining the ependyma (arrowheads) and a subependymal giant cell astrocytoma is seen in (d), at the level of the foramen of Monro (arrow). (e) Axial T2-weighted image shows a radial migration line tracking from the tuber into the deep white matter (arrow, and zoom frame). (f) Axial fluid attenuation inversion recovery image. Cyst-like appearance of a tuber (arrows). (Reproduced with permission from Peters JM, Taquet M, Prohl AK, et al. Diffusion tensor imaging and related techniques in tuberous sclerosis complex: review and future directions. Future Neurol 2013;8:583–597.)

„„ Identification of Epilepsy Surgical Candidates and Challenges in Tuberous Sclerosis Complex Refractory epilepsy is common, and all of those refractory to drugs should be considered. In addition, those with ­controlled seizures but requiring multiple antiepileptic drugs

(­polypharmacy) and with prominent medication side effects should be considered as well. Young patients with highly refractory seizures but with persistent focality on electroencephalography (EEG) can be feasible candidates, and should be considered for surgery early. Although the success of surgery hinges on reliable identification of a focal epileptogenic zone, there is no TSC-specific scheme for epilepsy surgery workup.11 The exact presurgical

425

426

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Box 43.1 Revised Diagnostic Criteria for Tuberous Sclerosis Complex 2012 A. Genetic Diagnostic Criteria The identification of either a TSC1 or TSC2 pathogenic mutation in DNA from normal tissue is sufficient to make a definite diagnosis of TSC. A pathogenic mutation is defined as a mutation that clearly inactivates the function of the TSC1 or TSC2 proteins (e.g., out-of-frame indel or nonsense mutation) or prevents protein synthesis (e.g., large genomic deletion), or is a missense mutation whose effect on protein function has been established by functional assessment. Other TSC1 or TSC2 variants whose effect on function is less certain do not meet these criteria, and are not sufficient to make a definite diagnosis of TSC. Note that approximately 15% of TSC patients have no mutation identified by conventional genetic testing, and a normal result does not exclude TSC, or have any effect on the use of clinical diagnostic criteria to diagnose TSC.

B. Clinical Diagnostic Criteria Major features: • • • • • • • • • • •

Hypomelanotic macules (three or more, at least 5-mm diameter). Angiofibromas (three or more) or fibrous cephalic plaque. Ungual fibromas (two or more). Shagreen patch. Multiple retinal hamartomas. Cortical dysplasias.a Subependymal nodules. Subependymal giant cell astrocytoma. Cardiac rhabdomyoma. Lymphangioleiomyomatosis (LAM).b Angiomyolipomas (two or more).b

Minor features:

• “Confetti” skin lesions. • Dental enamel pits (more than three). • Intraoral fibromas (two or more). • Retinal achromic patch. • Multiple renal cysts. • Nonrenal hamartomas. Definite diagnosis: Two major features or one major feature with two or more minor features. Possible diagnosis: Either one major feature or two or more minor features. Source: Data from Northrup and Krueger.5 a   Includes tubers and cerebral white matter radial migration lines. b   A combination of the two major clinical features (LAM and ­angiomyolipomas) without other features does not meet criteria for a definite diagnosis.

evaluation depends on clinical characteristics, parents and physician preference and experience, institutional practice, and availability of auxiliary tests. Similar factors lead to ­ differences in surgical approach, including single-stage resections, intra- versus extraoperative monitoring with subdural grids and depth electrodes, and more recently stereo-EEG and magnetic resonance imaging (MRI)-guided laser thermal ablation (magnetic resonance thermography–guided laser-induced thermal therapy [LITT]). Finally, the goals of surgery vary, and range from seizure freedom to targeting a specific disabling seizure type. As a result, surgical outcomes vary from 57 to 70% Engel Class I (seizure freedom) and are difficult to compare.12,​13 There are a number of challenges unique to pediatric epilepsy surgery in TSC (Table 43.1), which have led to a wide v ­ ariation

in the use of investigative tools in various combinations in different centers, referred to as “epilepsy surgery recipes galore” by one author.11 Moreover, there is no single modality clearly superior to others, alone or in combination, that has been shown to improve the rate of seizure freedom after surgery.14 First and foremost, the multilesional nature of the disorder creates an extraordinary challenge: which of the many tubers is the culprit, and how are epileptogenic regions differentiated from nonepileptogenic regions? The pathophysiology of epileptogenesis in TSC is not yet fully elucidated. It involves a complex interplay between astrocytosis15 and aberrant neurons,16 ultimately favoring excitation over inhibition. Both astrocytes and neurons contribute to epileptogenesis through a variety of mechanisms on the molecular, cellular, and network levels.17 While a full review is outside the scope of this chapter, it is important to realize that the proposed pathophysiological bases of epileptogenesis have not translated yet into novel imaging techniques, such as new positron emission tomography (PET) ligands or high-resolution MRI markers of epileptogenesis. Currently, concordance of multiple presurgical test modalities, supplemented in some by invasive monitoring, remains the gold standard for identification of the epileptogenic zone in TSC. Second, tubers are poorly delineated. While with lower resolution MRI lesions may appear discrete, higher resolution imaging reveals expansion of pathology beyond the macroscopically evident borders.18 This results in extensive nondiscrete regions of potential epileptogenesis, including multiple tuber complexes which appear merged, expansive white matter abnormality, and areas of transmantle dysplasia. Third, patients with TSC often have multiple seizure types. Some of these seizures may not have any localizing qualities clinically or on EEG (e.g., infantile spasms or tonic seizures), and may not all originate from the same generators. Consequently, seizure freedom may not be a realistic goal. Still, palliative goals can be attainable: reduction of the total seizure burden, targeting a specifically threatening or disabling seizure type (e.g., seizures associated with falls, profound apnea, nocturnal occurrence, or status epilepticus), lowering medication, or reduced risk for SUDEP (sudden unexplained death in epilepsy) can be associated with significant improvement of the quality of life.12 Encouragingly, some patients can undergo resection of multiple tubers, and coexistence of multiple seizure types with early onset and multifocal or generalized EEG abnormality is not necessarily associated with a poor seizure outcome.11,​14 Likewise, the nonspecific seizure semiology at a young age may not provide much localizing information but should not dissuade the pursuit of epilepsy surgery.12

„„ Presurgical Evaluation of Tuberous Sclerosis Complex Imaging Techniques Conventional and High-Resolution Magnetic Resonance Imaging Structural MRI is used for diagnosing and monitoring TSC, but the ability to localize epileptogenic tubers is limited.19

43  Tuberous Sclerosis Complex Table 43.1  Challenges in pediatric epilepsy surgery in tuberous sclerosis complex

Challenge

Approach

Young age, developmental delays, and behavioral problems prevent cooperation with noninvasive mapping and nonsedated studies (e.g., MEG, fMRI, high-density EEG, TMS for mapping)

Behavioral techniques can improve cooperation. Timing of the study at night may allow for sleep and little motion where sedation is not an option. Sedated studies as indicated. Limit duration of studies to acquire essential data only

Nonspecific “bland” semiology in young age not helpful in localizing seizure origin

Concordance of multiple other modalities can be sufficient for localization

Refractory epilepsy often at young age, and developmental gain from seizure freedom is largest at young age—but stereo-EEG is not feasible given skull thickness

Consider classic surgical techniques including intra- and extraoperative monitoring, and single-stage surgery

Multilesional nature of disorder, which one of the many lesions is ­epileptogenic?

Apply extensive multimodal workup to determine epileptogenic lesion. Multiple lesions can be targeted both during invasive monitoring and during surgery

Extensive or poorly delineated regions of potential ­epileptogenesis, including merger of multiple tuber complexes, expansive white matter abnormality, and transmantle dysplasia

Classically, larger volume resections are associated with improved outcomes. Now, stereo-EEG may allow for determination of important nodes in the seizure-generating network, which can be targeted

Generalized or multifocal EEG abnormalities are found

Such EEG abnormalities do not preclude a good surgical outcome

A patient has multiple seizure types or refractory infantile spasms without localizing features. Over time, focal seizures can become more generalized, and result in a diffuse epileptic encephalopathy

Review all EEG studies, including earlier ones, for clues about persistent focality over time. The epileptogenic focus can be stable over the course of years

Seizure freedom may not be realistic goal

Other goals for epilepsy surgery in TSC include reduction of seizure burden, palliation of most severe or debilitating seizures, mitigation of nocturnal seizures and associated SUDEP risk, less medications and side effects postsurgery

Possibility of postsurgical (late) relapse of seizure emanating from same or different regions

Even a partial or temporary relief can be critical if achieved during early years of steep neurodevelopment; seizure freedom or reduced medications can be associated with developmental acceleration

Abbreviations: EEG, electroencephalography; fMRI, functional magnetic resonance imaging; MEG, magnetoencephalography; SUDEP, sudden unexplained death in epilepsy; TMS, transcranial magnetic stimulation; TSC, tuberous sclerosis complex. Source: Data from Northrup and Krueger.5

­ lassically, characteristics like tuber volume, calcification, and C cyst-like degeneration were used. A recent report suggests that use of features typically associated with focal cortical dysplasia, including cortical thickness, perituber cortical abnormalities, transmantle white matter migration lines, and blurring of the gray/white matter junction, may be more predictive.20 Another study introduced the radiological concept of a “tuber center,” a bull’s-eye-like lesion in the center of the tuber, evident in multiple planes, often at the bottom of the sulcus. Several of these centers were the target of depth electrodes during intraoperative monitoring. Based on qualitative and quantitative analysis of the EEG signal, these centers were deemed both the epileptogenic origin and the site for intertuber seizure propagation.21 This work has not yet been reproduced, contradicts reports of neurophysiological silence within the tuber, and was limited by the use of neuronavigational rather than stereotactic guidance for depth placement, with variable outcomes in a small study population. Tubers are not well-delineated, and the true extent of TSC pathology is not evident on conventional structural MRI. With higher resolution imaging, and with higher magnet strength at 3T or 7T, detection of abnormalities like subtle radial migration lines or small pockets of tuber pathology (“microtubers”), with better delineation of lesions, has been reported in comparison to 1.5T MRI.20 Increased incidental findings, sensitivity to artifact, image distortion, and noise levels may hamper the clinical utility of high-resolution scans.

Nuclear Imaging Like in any routine presurgical workup, 18-fluoro-2­­deoxy-D-glucose PET (FDG-PET) is used to detect areas of aberrant hypometabolism during the interictal state. Tracer uptake in close temporal proximity to a seizure may result in ­pathologic hypermetabolism. In TSC, tuber and perituber regions are hypoactive, but nonepileptogenic lesions are not differentiated from epileptogenic regions easily. Registration with MRI and quantification of FDG tracer uptake has demonstrated that ­epileptogenic tubers are associated with a disproportionally large area of PET hypometabolism.22 Single-photon emission computed tomography (SPECT) is different from PET, but also widely used in epilepsy surgery. An ictal scan is performed after a tracer is injected rapidly at onset of the seizure, and a separate interictal scan is done in between seizures. These two images can be subtracted to maximize the contrast between areas of (ictal) increased perfusion and (interictal) decreased perfusion. Subtraction of ictal SPECT co-registered to MRI (SISCOM) allows for quantifying intensity differences and projection of outlier areas (e.g., two standard deviations from the mean) onto the patient’s own MRI. Scant literature is available on the use of SPECT in TSC. In a small series of six patients, complete resection of the SISCOM abnormality was associated with good surgical outcome, and multifocal ictal SPECT findings were associated with a poor outcome.23

427

428

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Neurophysiology Electroencephalography Conventional interpretation of scalp EEG data involves assessment of interictal and ictal activity. All prior EEGs should be taken into account, as seizures can be focal at onset, but when the epileptic network expands as the child ages, these can become (more) generalized and harder to localize. Over time, however, an EEG focus is typically consistent,24 and there is good concordance between ictal and interictal data.25 Thus, multifocal and generalized EEG abnormalities do not preclude successful epilepsy surgery.12,​14 Electrical source imaging (ESI) is a localization technique, which models the cerebral origins of scalp EEG. A forward model is built from the patient’s own MRI, which is segmented into different layers including cortex, white matter, cerebrospinal fluid (CSF), skull, and scalp. Depending on estimated conduction through each of these layers, a theoretic electrical source is placed at each anatomical region to produce its own unique voltage map on the scalp. The inverse modeling reverses this process; the starting point is now an actual EEG voltage map (e.g., from an average of several interictal spikes), and the

model proposes possible solutions of active areas in the brain that can approximate the measured data. As with conventional EEG, there are only 19 to 21 scalp electrodes and thousands of brain areas that can combine in many ways to explain the measured data; in this case, the model is “ill-posed,” i.e., many solutions are mathematically equally feasible. This problem is in part overcome by introducing constraints to the model, based on anatomy and physiology. For example, the solution is only allowed to be in cortical tissue, and must remain orthogonally oriented to the cortical surface to account for the net orientation of pyramidal cells. With such constraints, the number of possible solutions is reduced. With more electrodes, for example, in high-density EEG, and with registration of the electrodes, solutions become more accurate (Fig. 43.1a,b). In TSC, only a few dedicated studies on ESI exist. One could argue that constraining the solution to the cortical ribbon is appropriate, as the tubers and perituber area are considered the epileptogenic zone. In a small sample of 11 patients, high-resolution ESI was concordant with the resection cavity in 5 out of 5 patients who achieved seizure freedom from surgery. With routine scalp EEG, there was concordance in seven out of nine patients.26 With magnetoencephalography (MEG), ­ magnetic source imaging can be done, as outlined below (Fig. 43.2).

Fig. 43.2  Various novel study modalities in a 4-year-old patient with tuberous sclerosis complex. (a) Electric source imaging with an equivalent current dipole. Here the grand average of the electrical sources of the active cortex during the early phase of a seizure is modeled by a single dipole. The negative (round) end of the dipole projects to the active cortex, and is placed deeply so it can project to a large cortical surface (like a flashlight held further away). (b) Distributed electric source imaging solution using low-resolution brain electromagnetic tomography in a patient-specific model. (c) Magnetic source imaging of an interictal spike, using a minimum norm estimate, illustrating a wide area of activity at the peak of the MEG spike. (d) Tractography of the corticospinal tract, illustrating the anatomical overlap with the segmented surgical cavity. The patient had a hemiparesis postsurgically. (e) A different diffusion model is called DIstribution of Anisotropic Microstructural environments with DWI (DIAMOND); here it demonstrates increased heterogeneity of the distribution of diffusion along the axis of the main tensors and increased free diffusion of water, suggesting decreased microstructural integrity beyond conventional DTI. (f) Projection of electrode grids on cortical surface through registration of postoperative CT (good hard tissue contrast like metal and skull) to preoperative MRI (good soft tissue contrast). This method accounts for nonuniform changes in brain shift. CT, computed tomography; DTI, diffusion tensor imaging; DWI, diffusion-weighted imaging; MEG, magnetoencephalography. (Reproduced with permission from Boom M, Raskin JS, Curry DJ, Weiner HL, Peters JM. Technological advances in pediatric epilepsy surgery: implications for tuberous sclerosis complex. Future Neurol 2017;12:101–115.)

43  Tuberous Sclerosis Complex A more detailed discussion on source localization including an example in TSC can be found elsewhere.13

Magnetoencephalography MEG measures small magnetic fields induced by electric currents from synchronized neuronal activity in the brain. Strengths of the MEG include a lack of attenuation or distortion from conduction through CSF, skull, and scalp, resulting in a high signal-to-noise ratio. Together with the high number of sensors registered to the patient’s own MRI, MEG has superior sensitivity and spatial accuracy compared to EEG. Limitations of MEG, however, include its dependence on sophisticated hardware and software, and the considerable expertise required to acquire, process, and interpret the data. In addition, the machine is less mobile than EEG, and pediatric patients may require sedation. Recordings are typically brief to facilitate patient cooperation, and capture only interictal activity. Only a few studies have been done in TSC specifically. MEG was reported as superior to high-density EEG in the localization of epileptic activity in the proximity of a tuber, and had higher interrater agreement.27 A high sensitivity and specificity of MEG in the presurgical workup of TSC was reported by another group, with more focal solutions and closer to tuber lesions compared to ictal EEG.28

Combined Modalities Aside from incorporating all modalities in the decision-­making process regarding epilepsy surgical candidacy and surgical approach, the merger of modalities can increase the overall diagnostic yield. MEG and PET coregistered to MRI were used in a study of 28 operated patients with TSC. Complete resection of the regions implicated by the PET and MEG sources was associated with a good surgical outcome.29 PET can also be combined with structural imaging. Quantification of areas of decreased PET activity and registration with MRI showed that epileptogenic tubers, compared to nonepileptogenic lesions, had larger volumes of associated hypometabolism.22

Other Experimental Modalities α-[11C] methyl-l-tryptophan (AMT), a labeled serotonin precursor, is a relatively new PET tracer. AMT-PET is compared to FDG-PET, and an increased AMT uptake ratio locally implicates the epileptogenic zone in TSC. Conversely, a nonlocalizing AMT-PET is associated with decreased surgical success. In a small study of 17 patients, an empiric cut-off value of the AMT uptake ratio was established,30 and in a large study of 191 patients the method was further validated.31 The technique requires considerable expertise and is neither reimbursed nor widely available. Diffusion-weighted imaging and diffusion tensor imaging are used in the localization of the epileptogenic zone. A higher apparent diffusion coefficient (ADC; comparable to mean diffusivity) in epileptogenic tubers as compared to nonepileptogenic tubers was first reported in a small study of four patients,

almost two decades ago.32 More recently, in a study of 25 operated patients, indeed a higher ADC and higher radial diffusivity of tuber and tuber rim were found in epileptogenic tubers.33 In practice, calculation of the diffusivity values in each tuber (with or without tuber rim) is not trivial and requires image postprocessing tools that are not yet user-friendly (Fig. 43.2d,e). Finally, with sophisticated processing techniques, information can be extracted out of routinely acquired scalp EEG and intracranial EEG. High-frequency oscillations (HFOs) are high-frequency EEG oscillations, which can be either physiological or pathological. The latter are subdivided into ripples (80–250 Hz), fast ripples (250–500 Hz), and ultrafast ripples (500+ Hz), and are considered biomarkers of the epileptogenic zone.34 In a study of 10 patients with TSC, resection of a higher number of electrodes with frequent fast ripples was associated with improved surgical outcome.35 User-friendly tools are becoming available for the real-time interpretation of such fast activity. Further study of the clinical utility is needed, e.g., HFOs in the epileptogenic zone (seizure onset) versus in the irritative zone (seizure spread/interictal abnormalities), HFO propagation, HFOs in depth electrodes, and more, before this technique can be more widely implemented. An example of intracranially placed subdural grids and strips can be found in Fig. 43.2f.

„„ Controversies in Pediatric Epilepsy Surgery in Tuberous Sclerosis Complex The current understanding of epileptogenesis in TSC is advancing, but several key issues need further elucidation, as discussed below.

Tuber-to-Tuber Propagation Tuber-to-tuber propagation has been described in an exquisitely detailed paper from Kannan et al.21 The authors revealed through in-depth analysis of the intracranial EEG of 10 patients that while seizure activity started in one tuber, it could propagate to other tubers. Over a third of the tubers with propagated activity, once triggered, had their own independent ictal pattern. These tubers with such “intraictal activation” also showed a capacity to be the seizure onset zone. Conversely, tubers that only showed propagation of ictal activity but did not have their own ictal EEG signature did not independently generate seizures. With the widespread introduction of s­ tereo-EEG to the field, it will be important to confirm these propagation patterns as markers of epileptogenicity, to distinguish tubers with a capacity for generating seizures from those that merely propagate seizures.

Intra- versus Extratuber Onset of Seizures Whether seizures originate from the tuber or from the perituber tissue is still controversial. A case series of three invasively monitored patients revealed electrical silence during seizures, although only one had ictal data, and the location of the electrodes was not determined stereotactically.36 In another series

429

430

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches of 12 children with TSC, depth electrode contacts were again visually assessed as being part of the tuber or perituber region, and results were mixed; the seizures originated about evenly from perituber and from tuber depth contacts.37 In 17 patients from an Australian series, an intratuber onset was found approximately two to three times more frequently than perituber onset.38 In 10 patients from the same center, quantitative analysis suggested the tubers always led the perituber rim by up to 20 ms.21 Several authors have argued that since tubers are histopathologically heterogeneous and poorly delineated, the distinction between perituber and intratuber regions is arbitrary as small pockets of tuber pathology are scattered beyond the MRI-evident borders.18,​39,​40

Minimally Invasive or Large Resection Volume? The surgical implications of intratuber versus perituber onset of seizures (see above) may be that smaller lesions or highly targeted focal resections are sufficient for seizure freedom. With the introduction of minimally invasive monitoring through stereo-EEG and the use of magnetic resonance thermography–guided LITT, subsections of large tuber conglomerates can be targeted.13 The concept of minimally invasive surgery, however, is in contrast with the resection volume as a major predictor of surgical

success in a large multicenter study of 74 patients,41 suggesting a generous resection including tuber and tuber rim is better. In the literature on pediatric stereo-EEG, however, children with TSC have thus far only been included as part of larger case series, and the collective experience of multiple surgical centers with expertise in TSC needs to be analyzed and reported.13

Can Extensive Preoperative Workup Preclude the Need for Invasive Monitoring? There is considerable variation between epilepsy centers in the approach to invasive monitoring in pediatric epilepsy surgery in TSC. Some centers have advocated for an almost exclusively noninvasive workup, followed by single-stage surgery,29 while others have traditionally used a multistage and even bilateral invasive monitoring strategy.42 As reported surgical outcomes appear similar, it may be that the familiarity, expertise, and cumulative experience of specific centers with TSC epilepsy surgery are more important than the techniques or tools used. A multicenter prospective observational study collecting data on presurgical workup, intraoperative neurophysiology, postoperative imaging, and surgical outcomes would allow for the study of the optimal surgical approach to pediatric epilepsy surgery in TSC.

References 1. Crino PB, Nathanson KL, Henske EP. The tuberous sclerosis complex. N Engl J Med 2006;355(13):1345–1356 2. Davis PE, Peters JM, Krueger DA, Sahin M. Tuberous sclerosis: a new frontier in targeted treatment of autism. Neurotherapeutics 2015;12(3):572–583 3. Lipton JO, Sahin M. The neurology of mTOR. Neuron 2014; 84(2):275–291 4. Ess KC. The neurobiology of tuberous sclerosis complex. Semin Pediatr Neurol 2006;13(1):37–42 5. Northrup H, Krueger DA; International Tuberous Sclerosis Complex Consensus Group. Tuberous sclerosis complex diagnostic criteria update: recommendations of the 2012 International Tuberous Sclerosis Complex Consensus Conference. Pediatr Neurol 2013;49(4):243–254 6. Numis AL, Major P, Montenegro MA, Muzykewicz DA, ­Pulsifer MB, Thiele EA. Identification of risk factors for autism spectrum disorders in tuberous sclerosis complex. Neurology 2011;76(11):981–987 7. Jansen FE, Vincken KL, Algra A, et al. Cognitive impairment in tuberous sclerosis complex is a multifactorial condition. Neurology 2008;70(12):916–923 8. Capal JK, Bernardino-Cuesta B, Horn PS, et al; TACERN Study Group. Influence of seizures on early development in tuberous sclerosis complex. Epilepsy Behav 2017;70(Pt A):245–252 9. Jóźwiak S, Kotulska K, Domańska-Pakieła D, et al. Antiepileptic treatment before the onset of seizures reduces epilepsy severity and risk of mental retardation in infants with tuberous sclerosis complex. Eur J Paediatr Neurol 2011;15(5): 424–431 10. Muzykewicz DA, Costello DJ, Halpern EF, Thiele EA. Infantile spasms in tuberous sclerosis complex: prognostic utility of EEG. Epilepsia 2009;50(2):290–296 11. Gupta A. “Epilepsy surgery recipes galore”: in quest for the epileptogenic tuber in tuberous sclerosis complex. Epileptic Disord 2009;11(1):80–81

12. Gupta A. Epilepsy surgery in tuberous sclerosis complex: in pursuit of the epileptogenic center(s). Epilepsy Curr 2017;17(3):150–152 13. Boom M, Raskin JS, Curry DJ, Weiner HL, Peters JM. Technological advances in pediatric epilepsy surgery: implications for tuberous sclerosis complex. Future Neurol 2017;12(2):101–115 14. Jansen FE, van Huffelen AC, Algra A, van Nieuwenhuizen O. Epilepsy surgery in tuberous sclerosis: a systematic review. Epilepsia 2007;48(8):1477–1484 15. Wong M, Crino PB. Tuberous sclerosis and epilepsy: role of astrocytes. Glia 2012;60(8):1244–1250 16. Talos DM, Kwiatkowski DJ, Cordero K, Black PM, Jensen FE. Cell-specific alterations of glutamate receptor expression in tuberous sclerosis complex cortical tubers. Ann Neurol 2008;63(4):454–465 17. Wong M. Mechanisms of epileptogenesis in tuberous sclerosis complex and related malformations of cortical development with abnormal glioneuronal proliferation. Epilepsia 2008;49(1):8–21 18. Peters JM, Prohl AK, Tomas-Fernandez XK, et al. Tubers are neither static nor discrete: evidence from serial diffusion tensor imaging. Neurology 2015;85(18):1536–1545 19. Krueger DA, Northrup H; International Tuberous Sclerosis Complex Consensus Group. Tuberous sclerosis complex surveillance and management: recommendations of the 2012 International Tuberous Sclerosis Complex Consensus Conference. Pediatr Neurol 2013;49(4):255–265 20. Jahodova A, Krsek P, Kyncl M, et al. Distinctive MRI features of the epileptogenic zone in children with tuberous sclerosis. Eur J Radiol 2014;83(4):703–709 21. Kannan L, Vogrin S, Bailey C, Maixner W, Harvey AS. Centre of epileptogenic tubers generate and propagate seizures in tuberous sclerosis. Brain 2016;139(Pt 10):2653–2667 22. Chandra PS, Salamon N, Huang J, et al. FDG-PET/MRI coregistration and diffusion-tensor imaging distinguish

43  Tuberous Sclerosis Complex ­ pileptogenic tubers and cortex in patients with tuberous sclee rosis complex: a preliminary report. Epilepsia 2006;47(9): 1543–1549 23. Aboian MS, Wong-Kisiel LC, Rank M, Wetjen NM, Wirrell EC, Witte RJ. SISCOM in children with tuberous sclerosis complex-related epilepsy. Pediatr Neurol 2011;45(2):83–88 24. Jansen FE, van Huffelen AC, Bourez-Swart M, van N ­ ieuwenhuizen O. Consistent localization of interictal epileptiform activity on EEGs of patients with tuberous sclerosis complex. Epilepsia 2005;46(3):415–419 25. van der Heide A, van Huffelen AC, Spetgens WP, Ferrier CH, van Nieuwenhuizen O, Jansen FE. Identification of the epileptogenic zone in patients with tuberous sclerosis: concordance of interictal and ictal epileptiform activity. Clin Neurophysiol 2010;121(6):842–847 26. Kargiotis O, Lascano AM, Garibotto V, et al. Localization of the epileptogenic tuber with electric source imaging in patients with tuberous sclerosis. Epilepsy Res 2014;108(2):267–279 27. Jansen FE, Huiskamp G, van Huffelen AC, et al. Identification of the epileptogenic tuber in patients with tuberous sclerosis: a comparison of high-resolution EEG and MEG. Epilepsia 2006;47(1):108–114 28. Wu JY, Sutherling WW, Koh S, et al. Magnetic source imaging localizes epileptogenic zone in children with tuberous sclerosis complex. Neurology 2006;66(8):1270–1272 29. Wu JY, Salamon N, Kirsch HE, et al. Noninvasive testing, early surgery, and seizure freedom in tuberous sclerosis complex. Neurology 2010;74(5):392–398 30. Kagawa K, Chugani DC, Asano E, et al. Epilepsy surgery outcome in children with tuberous sclerosis complex evaluated with alpha-[11C]methyl-L-tryptophan positron emission tomography (PET). J Child Neurol 2005;20(5):429–438 31. Chugani DC, Chugani HT, Muzik O, et al. Imaging epileptogenic tubers in children with tuberous sclerosis complex using alpha[11C]methyl-L-tryptophan positron emission tomography. Ann Neurol 1998;44(6):858–866 32. Jansen FE, Braun KP, van Nieuwenhuizen O, et al. ­Diffusionweighted magnetic resonance imaging and identification of the

epileptogenic tuber in patients with tuberous sclerosis. Arch ­ eurol 2003;60(11):1580–1584 N 33. Yogi A, Hirata Y, Karavaeva E, et al. DTI of tuber and perituberal tissue can predict epileptogenicity in tuberous sclerosis complex. Neurology 2015;85(23):2011–2015 34. Zijlmans M, Jiruska P, Zelmann R, Leijten FS, Jefferys JG, Gotman J. High-frequency oscillations as a new biomarker in epilepsy. Ann Neurol 2012;71(2):169–178 35. Okanishi T, Akiyama T, Tanaka S, et al. Interictal high frequency oscillations correlating with seizure outcome in patients with widespread epileptic networks in tuberous sclerosis complex. Epilepsia 2014;55(10):1602–1610 36. Major P, Rakowski S, Simon MV, et al. Are cortical tubers epileptogenic? Evidence from electrocorticography. Epilepsia 2009;50(1):147–154 37. Ma TS, Elliott RE, Ruppe V, et al. Electrocorticographic evidence of perituberal cortex epileptogenicity in tuberous sclerosis complex. J Neurosurg Pediatr 2012;10(5):376–382 38. Mohamed AR, Bailey CA, Freeman JL, Maixner W, Jackson GD, Harvey AS. Intrinsic epileptogenicity of cortical tubers revealed by intracranial EEG monitoring. Neurology 2012;79(23): 2249–2257 39. Marcotte L, Aronica E, Baybis M, Crino PB. Cytoarchitectural alterations are widespread in cerebral cortex in tuberous sclerosis complex. Acta Neuropathol 2012;123(5):685–693 40. Ruppe V, Dilsiz P, Reiss CS, et al. Developmental brain abnormalities in tuberous sclerosis complex: a comparative tissue analysis of cortical tubers and perituberal cortex. Epilepsia 2014;55(4):539–550 41. Fallah A, Rodgers SD, Weil AG, et al. Resective epilepsy surgery for tuberous sclerosis in children: determining predictors of seizure outcomes in a multicenter retrospective cohort study. Neurosurgery 2015;77(4):517–524, discussion 524 42. Weiner HL, Carlson C, Ridgway EB, et al. Epilepsy surgery in young children with tuberous sclerosis: results of a novel approach. Pediatrics 2006;117(5):1494–1502

431

44

  Resective Epilepsy Surgery for Tuberous Sclerosis Complex Jeffrey S. Raskin, Daniel J. Curry, and Howard L. Weiner

Summary The first tuberous sclerosis complex (TSC) patient was treated in 1880 by Bourneville for the syndrome consisting of seizures, mental retardation, and adenoma sebaceum. The central nervous system can become burdened by three principal intracranial pathologies: cortical or subcortical tubers, subependymal nodules, and subependymal giant cell astrocytomas. Cognitive disability and autism are common phenotypic manifestations in individuals with TSC, and approximately 75% have epilepsy. Surgical epilepsy dogma defines a seizure onset zone as a targetable cortical or subcortical area, which drives hypersynchronous neuronal discharge, causing a seizure. The mechanism for symptomatic localization-related epilepsy is still incompletely understood. Both cortical tubers and surrounding abnormal cortex can comprise the epileptogenic zone. Epileptic patients with diffuse tuber burden are generally thought not to be candidates for resection due to the bilateral and multifocal nature; however, innovative minimally invasive neurosurgical techniques and a change in pediatric epilepsy surgery philosophy are together promoting a renaissance of discovery in efficacious treatment for this once-abandoned patient population. Keywords:  tuberous sclerosis complex, resective surgery, multifocal epilepsy syndrome, stereotactic laser ablation, minimally invasive epilepsy surgery

of tuberin protein.5 In the disease state, altered p ­ rotein function causes constitutive activation of the mTOR pathway leading to cell proliferation and differentiation. The pathobiology of TSC causes metaplasia, which is manifest as benign tubers of the brain, skin, heart, eyes, lungs, and kidneys. The central nervous system can become burdened by three principal intracranial pathologies: cortical or subcortical tubers, subependymal nodules (SENs), and subependymal giant cell astrocytomas (SEGAs). Tubers and SENs are found in 85 to 90% of patients with TSC, while SEGAs occur in 5 to 20% of patients.7,​8 SEGAs can cause hydrocephalus but generally are not associated with epilepsy syndromes; cortical and subcortical tubers and the cortex around them can be epileptogenic and are the targets of epilepsy syndrome in this population.5,​9,​10 Cerebral tubers are histologically characterized by dysmorphic neurons and giant cells as well as pathobiological processes including angiogenesis and dystrophic calcification.11,​12 Based on unique histopathologic characteristics, tubers have been subdivided into types A, B, and C, with subtypes B and C correlating with presurgical radiographic features. This subdivision may serve as a baseline for further granular understanding of the clinic-pathological implications of tuber–perituberal interfaces.

„„ Introduction

„„ Epilepsy Syndromes in Tuberous Sclerosis Complex: Why Is It Difficult to Treat?

The largest clinical series of prospective observation of patients with tuberous sclerosis complex (TSC) includes patients who live outside of the United States, and is structured to study the natural history of the disease across ethnobiological populations.1 In the United States, TSC affects approximately 1 in 6,000 individuals and has a prevalence of about 50,000 to 100,000 persons.2 The first TSC patient was treated in 1880 by Bourneville for the syndrome consisting of seizures, mental retardation, and adenoma sebaceum, previously defined by Vogt in 1908.3 The genetic underpinnings were discovered much later by Gomez in 1969.4 TSC is an autosomal dominant multisystem genetic disorder most commonly inherited sporadically.5,​6 TSC1 is caused by a chromosome 9 mutation affecting the production of hamartin protein; TSC2 is caused by a chromosome 16 error affecting the production

Cognitive disability and autism are common phenotypic manifestations in individuals with TSC, and approximately 75% have epilepsy.8 Epilepsy syndromes are often overlapping and include complex partial, atonic, myoclonic, absence, and generalized tonic–clonic seizures.8 Antiepileptic drug therapy and ketogenic diet are often partially or completely ineffective.13 Surgical epilepsy dogma defines a seizure onset zone (SOZ) as a targetable cortical or subcortical area, which drives hypersynchronous neuronal discharge, causing a seizure. In concordant epilepsy syndromes, which are characterized by a single seizure semiology, resection of the SOZ is highly effective, especially when a radiographic lesion is present. In patients with more complex epilepsy, featuring heterogeneous semiologies, focal resection has been considered antidogmatic.14

44  Resective Epilepsy Surgery for Tuberous Sclerosis Complex The mechanism for symptomatic localization-related ­ pilepsy is still incompletely understood. Both cortical tubers e and surrounding abnormal cortex can comprise the epileptogenic zone.10,​15 Intracranial electroencephalography (iEEG) with depth electrodes placed with spatial optimization to sample tuber center versus tuber rim in children with TSC found that 90% of seizure onset emanated from tuber center.16 Resection of the tuber plus the surrounding cortex is associated with longer seizure-free periods after multivariate analysis.17 Ma and colleagues demonstrated ictal onset at both the tuber and perituberal cortex.18 Also, not all tubers are epileptogenic. Histologic and immunochemical studies of perituberal cortex identified similar but more mild versions of the dysplasia and mTOR signaling deficiency as in tubers themselves, with a concomitant lack of myelination which is hypothesized to underlie the pathophysiology of the ictal onset zone.19 Within this refractory epileptic population, in general, there seem to be two clinically separate patients: (1) restricted tuber burden with seizure onset localized to that lesion; or (2) diffuse tuber burden with unclear SOZ localization.17,​20 Patients with restricted disease are more amenable to surgical intervention and, as in surgery for other epitomas (e.g., cavernoma), surgery is usually effective.9,​21 Epileptic patients with diffuse tuber burden are generally thought not to be candidates for resection due to the bilateral and multifocal nature; however, innovative minimally invasive neurosurgical techniques and a change in pediatric epilepsy surgery philosophy are together promoting a renaissance of discovery in efficacious treatment for this once-abandoned patient population. Catastrophic childhood epilepsy is becoming an increasing target for surgical interventions, especially when several potentially epileptogenic structural lesions are present.22

„„ The History of Resective Surgery TSC patients with refractory epilepsy and their caregivers can have a very poor quality given the high seizure burden, which results in ineffective and costly polypharmacy, frequent doctor and emergency department visits, developmental delay, and lifelong dependent care. Modern surgical goals consistent with the International League Against Epilepsy (ILAE) are no seizures and no side effects.23 In genetic refractory cases like TSC, goals may not be seizure freedom but seizure reduction and palliation from the risks associated with lifelong continued refractory seizures and ineffective medications, including prevention of further brain injury and facilitation of normal brain development. Other goals include facilitation of psychological, social, and educational growth and improvement in overall quality of life. Mindful of these surgical goals, a comprehensive epilepsy center evaluates TSC patients with structural and functional imaging and neuropsychology evaluation to determine the extent of the epileptogenic zone and the functional status of the surrounding cortex. Unifocal onset cases are referred for resection, no identifiable SOZ are without surgical options, and patients with bilateral independent foci can be evaluated for unilateral surgery or bilateral surgery of nonhomologous foci. Surgical management of TSC patients with refractory epilepsy has traditionally been limited because commonly these patients have extensive bilateral tuber burden and multiple

seizure semiologies. Complex intractable TSC patients with ­discordant phase I data were initially treated at the Montreal Neurological Institute with varied surgical design; 6/18 patients underwent palliative corpus callosotomy, with 5 deriving at least some benefit in seizure reduction.24 Outcomes were better in 7 of 12 patients resected, with 5 being seizure-free and 2 having auras only.24 Surgical management of epilepsy has continued to change across the field, with neurosurgical interventions becoming more commonplace, especially for regressing patients or those with epileptic encephalopathy. A change in our understanding of the risk–benefit profile of surgical management had big implications for the refractory TSC patient. In 1993, Kelly and co-workers began to treat the rare TSC patient with regional disease and good concordance of phase I data.25 Their outcomes were excellent at a mean follow-up of 35 months: out of nine patients, six patients experienced seizure freedom and two patients experienced 80% seizure reduction, and only one patient was a nonresponder. This outcome data have been confirmed in several subsequent modern studies intending to reveal seizure outcomes and their predictors focusing on this subpopulation of TSC patients with single tuber/epileptogenic zones in older patients with good outcomes.26,​27 Multiple retrospective population studies identify good outcomes were associated with concordant imaging–EEG findings, unifocal onset seizures, mild developmental delay, and lobar-multilobar resection versus tuberectomy, while poor outcomes were associated with moderate-severe intellectual disability, tonic seizures, early age of seizure onset, history of infantile spasms, and multifocal interictal activity.17,​28 Lessons learned from resecting highly concordant lesions led to expanded indications for phase II monitoring and subsequent resective surgeries, even bilaterally when indicated.14,​29,​30 Romanelli and colleagues reported a single patient who underwent right parietal tuberectomy, followed by left frontal subdural grid (SDG) monitoring and subsequent resection with good result.31 Weiner and colleagues began bilateral SDG, resections, and even reimplantation and rerecording.29 A retrospective study from one single surgeon over 6 years had 25 TSC patients with three-stage surgery including phase II, resection, rerecording, and often bilateral with 84% Engel Class I at 28 months. Carlson and colleagues published on patients with TSC and uncorrelated EEG/imaging who underwent bilateral phase II iEEG (n = 20), with 8 experiencing Engel I or II at 25 months.30 The resective epilepsy surgery for tuberous sclerosis (REST) study demonstrated 65% Engel Class I at 1 year and 50% at 2 years for refractory TSC patients following tuberectomy or more extensive resection.17 Patients with better outcomes were more likely to be younger at age of onset, have localized disease (e.g., a single large tuber), and have undergone resection versus simple tuberectomy.17 However, Kaplan–Meier curve for complete seizure freedom in patients with TSC status post resective surgery is less than 40% at 5 years.32 The outcome of this bilateral surgical approach to TSC patients demonstrates similar seizure-free curves to other poor-risk resective surgeries. Particularly challenging TSC patients have bilateral EEG onset in which clear lateralization is unknown dictate bilateral iEEG monitoring. Resection of eloquent epileptogenic cortex (e.g., perirolandic), large tuber complexes involving multilobular or hemispheric, or resective surgery in children with ­radiologically no visible lesion is challenging in TSC cases. Resective surgery

433

434

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches may still be indicated in a select cohort of TSC children, often with a significantly reduced postoperative seizure b ­ urden.33 Bilateral iEEG and subsequent resection was once antidogmatic and innovative. The surgical approach continues to struggle for efficacy in some patients, particularly those who have silent SOZs due to either developmental delay or localization within associative cortex.17 Furthermore, the complication profile for bilateral recurrent craniotomies is well documented in the literature and includes hypertrophic scar formation, blood loss anemia requiring transfusion, wound infection, osteomyelitis, meningitis, and return to operation room (OR) for ­surgical management of a complication with a rate from 5.7 to over 20%.34,​35,​36

„„ Future Surgical Management Strategies Minimally invasive epilepsy surgery is an increasingly utilized methodology for epilepsy syndromes.37 The paradigm integrates diagnostic capabilities of stereoelectroencephalography (sEEG) with therapeutic treatment by laser interstitial thermal therapy (LITT).38 Pairing stereotaxy and LITT is defined as stereotactic laser ablation (SLA), for which refractory epilepsy in TSC is an emerging indication.38 Improving phase I imaging is helping to identify SOZs, which are directly targetable first by sEEG and then by SLA. The workflow for SLA involves robot-assisted or frame-based placement of depth electrodes at surgical targets suspicious for being an SOZ. The stereotactic trajectory of each electrode is maintained in situ by a transcranial bolt. A patient is then monitored in the epilepsy monitoring unit (EMU) per protocol for phase II monitoring. If an SOZ is identified, the patient is returned to the OR where the electrode is removed and a laser is placed down the previous trajectory maintained by the transcranial bolt to the same depth. The laser can then be used to ablate the SOZ. In failed TSC patients, there are obvious benefits to a surgical paradigm including sEEG and SLA. Minimally invasive nature of sEEG does not require revision craniotomy, obviating the increased surgical risks of pial injury and infection. In some selected cases, sEEG and SLA are used as primary treatment for multifocal epilepsy syndromes, including TSC. As our surgical morbidity decreases and understanding of the condition becomes more sophisticated, epilepsy surgery in children is expanding to include younger children with worse bilateral disease, even in genetic conditions like TSC, and perhaps as a palliative strategy for the long term.

„„ An Illustrative Case DC was a six-year-old left-handed male with clinically definite and genetically negative TSC with medically refractory multifocal epilepsy. He was evaluated in the surgical epilepsy program at Texas Children’s Hospital. His main seizure semiology is characterized by asymmetric tonic seizures, with subtle right arm elevation evolving bilaterally involving leg extension and generalized stiffening lasting 20 seconds to 1 minute, followed by intermittent myoclonic seizures.

Seizures were reported to occur 1 to 3 times daily, and these have occurred up to 30 times a day. This semiology has not changed. He developed infantile spasms at 15 months of age and has only experienced one seizure-free period prior to 3 years of age. An extensive multi-institutional phase I evaluation was performed which resulted in bilateral structural and functional imaging targetable neural correlates (Table 44.1). Inpatient scalp video-EEG (vEEG) monitorization was performed and it was nonlocalizing, identifying ictal and interictal abnormalities including nine typical electroclinical seizures with left frontal lobe onset; other typical seizures were noted to have right frontal or right temporal onset. Interictal abnormalities included multifocal spikes in the bilateral frontotemporal, left frontal, and right temporal regions. The patient was discussed during interdisciplinary epilepsy conference which recommended proceeding with presurgical evaluation for phase II monitoring with depth electrodes followed by laser ablation of main ictal focus if possible (Fig. 44.1). A 10-electrode dynamic sEEG was performed targeting the neural correlates identified from the phase I evaluation (Table 44.2). He remained 1 week in the EMU during which time many typical electroclinical events were recorded. More than 90% of the seizure onset was identified by electrodes 1L SFG aps, 2L MFG aps, and 3L IFG as. Many subclinical events were coming from 4L SMA aems (Fig. 44.1). Due to this relatively tight seizure network around these tubers, the surgical epilepsy team decided he would benefit from removal of his other electrodes and SLA of the electrically active tubers. The patient tolerated this very well and was discharged soon after ablation therapy. At 6 months postablation, he was admitted for vEEG which did not capture any seizure activity. He has been clinically seizure-free and is making developmental gains in speech, handwriting, and strength. This continued to be true at 10 months postablation with continued polypharmacy including Sabril, zonisamide, and Depakote. Table 44.1  Phase I modality and neural correlate

Structural

Functional

Modality

Neocortical targets

T1 +gad

No tuber enhancement

FLAIR

Diffuse bilateral frontal and left temporal subcortical tubers; many subependymal nodules

CT

Left frontal tuber calcification

Task-fMRI

Expected pericentral activation; absent right SLF

rs-fMRI

Left language dominance

MEG

Left frontal spikes, right temporal spike cluster

FDG-PET-CT

Hypometabolism of anterior left frontal lobe, lateral right frontal lobe, anterior right temporal lobe

Abbreviations: CT, computed tomography; FDG-PET-CT, fluorodeoxyglucose positron emission tomography (PET) CT; FLAIR, fluid attenuated inversion recovery; gad, gadolinium; MEG, ­magnetoencephalography; fMRI, functional MRI; rs-fMRI, resting state functional MRI.

44  Resective Epilepsy Surgery for Tuberous Sclerosis Complex

Fig. 44.1  MRI reconstruction demonstrating active electrode contacts from electrodes 1L superior frontal gyrus (SFG) aps, 2L middle frontal gyrus (MGF) aps, and 3L inferior frontal gyrus (IFG) as within the left frontal calcified tuber. Purple: electrodes; orange: calcified tuber; yellow: tubers; red oval: onset.

Table 44.2  SEEG electrode complement

Number and ­lateralization

Neural correlate

Phase I modality

Electrode name

# contacts

1L

SFG

a,p,s

1L SFG aps

8

2L

MFG

a,p,s

2L MFG aps

6

3L

IFG

a,s

3L IFG as

6

4L

SMA

a,e,m,s

4L SMA aems

6

5L

SFS

a,e,p

5L SFS aep

6

6R

Operc

a,e

6R Operc ae

8

7R

S1

a

7R S1

6

8R

OT

e,p,s

8R OT eps

12

9R

MTG

e,Mc

9R MTG eMc

6

10R

ITG

e,Mc

10R ITG eMc

6

Abbreviations: a, anatomy; e, EEG; IFG, inferior frontal gyrus; ITG, inferior temporal gyrus; L, left; m, MEG dipole; MC, MEG cluster; MFG, middle frontal gyrus; MTG, middle temporal gyrus; Operc, operculum; OT, occipital temporal; p, PET; R, right; s, semiology; S1, primary sensory; SFG, superior frontal gyrus; SFS, superior frontal sulcus; SMA, supplementary motor area.

„„ Acknowldgement We would like to acknowledge Dr. Rohini Coorg for her assistance with Fig. 44.1, and all of the members of the­

comprehensive surgical epilepsy team at Texas Children’s Hospital and thank them for the intellectual contributions ­utilized inpreparing this chapter.

References 1. Kingswood JC, d’Augères GB, Belousova E, et al; TOSCA consortium and TOSCA investigators. TuberOus SClerosis registry to increase disease Awareness (TOSCA)—baseline data on 2093 patients. Orphanet J Rare Dis 2017;12(1):2 2. von Ranke FM, Zanetti G, e Silva JL, et al. Tuberous sclerosis complex: state-of-the-art review with a focus on pulmonary involvement. Lung 2015;193(5):619–627

5. Baybis M, Yu J, Lee A, et al. mTOR cascade activation distinguishes tubers from focal cortical dysplasia. Ann Neurol 2004;56(4):478–487 6. Kothare SV, Singh K, Chalifoux JR, et al. Severity of manifestations in tuberous sclerosis complex in relation to genotype. Epilepsia 2014;55(7):1025–1029

3. O’Callaghan FJ, Osborne JP. Advances in the understanding of tuberous sclerosis. Arch Dis Child 2000;83(2):140–142

7. Hersh DS, Chun J, Weiner HL, et al. Longitudinal quantitative analysis of the tuber-to-brain proportion in patients with tuberous sclerosis. J Neurosurg Pediatr 2013;12(1):71–76

4. Gomez MR. Prenatal and neonatal seizure disorders. Postgrad Med 1969;46(1):71–77

8. Stafstrom CE, Staedtke V, Comi AM. Epilepsy mechanisms in neurocutaneous disorders: tuberous sclerosis complex, neuro­

435

436

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches fibromatosis type 1, and Sturge-Weber syndrome. Front Neurol 2017;8:87 9. Arya R, Tenney JR, Horn PS, et al. Long-term outcomes of resective epilepsy surgery after invasive presurgical evaluation in children with tuberous sclerosis complex and bilateral multiple lesions. J Neurosurg Pediatr 2015;15(1):26–33 10. Sosunov AA, McGovern RA, Mikell CB, et al. Epileptogenic but MRI-normal perituberal tissue in tuberous sclerosis complex contains tuber-specific abnormalities. Acta Neuropathol Commun 2015;3:17 11. Jeong A, Wong M. Systemic disease manifestations associated with epilepsy in tuberous sclerosis complex. Epilepsia 2016;57(9):1443–1449 12. Mühlebner A, van Scheppingen J, Hulshof HM, et al. Novel histopathological patterns in cortical tubers of epilepsy surgery patients with tuberous sclerosis complex. PLoS One 2016;11(6):e0157396 13. Fallah A, Weil AG, Wang S, Lewis E, Baca CB, Mathern GW. Cost-utility analysis of competing treatment strategies for drug-resistant epilepsy in children with tuberous sclerosis complex. Epilepsy Behav 2016;63:79–88 14. Weiner HL. Tuberous sclerosis and multiple tubers: localizing the epileptogenic zone. Epilepsia 2004;45(Suppl 4):41–42 15. MajorP,RakowskiS,SimonMV,etal.Arecorticaltubersepileptogenic? Evidence from electrocorticography.Epilepsia2009;50(1):147–154 16. Kannan L, Vogrin S, Bailey C, Maixner W, Harvey AS. Centre of epileptogenic tubers generate and propagate seizures in tuberous sclerosis. Brain 2016;139(Pt 10):2653–2667 17. Fallah A, Rodgers SD, Weil AG, et al. Resective epilepsy surgery for tuberous sclerosis in children: determining predictors of seizure outcomes in a multicenter retrospective cohort study. Neurosurgery 2015;77(4):517–524, discussion 524

Commissions of Neurosurgery and Paediatrics. Proposed criteria for referral and evaluation of children for epilepsy surgery: recommendations of the Subcommission for Pediatric Epilepsy Surgery. Epilepsia 2006;47(6):952–959 24. Guerreiro MM, Andermann F, Andermann E, et al. Surgical treatment of epilepsy in tuberous sclerosis: strategies and results in 18 patients. Neurology 1998;51(5):1263–1269 25. Bebin EM, Kelly PJ, Gomez MR. Surgical treatment for epilepsy in cerebral tuberous sclerosis. Epilepsia 1993;34(4):651–657 26. Madhavan D, Schaffer S, Yankovsky A, et al. Surgical outcome in tuberous sclerosis complex: a multicenter survey. Epilepsia 2007;48(8):1625–1628 27. Jansen FE, van Huffelen AC, Algra A, van Nieuwenhuizen O. Epilepsy surgery in tuberous sclerosis: a systematic review. Epilepsia 2007;48(8):1477–1484 28. Fallah A, Guyatt GH, Snead OC III, et al. Predictors of seizure outcomes in children with tuberous sclerosis complex and intractable epilepsy undergoing resective epilepsy surgery: an individual participant data meta-analysis. PLoS One 2013;8(2):e53565 29. Weiner HL, Carlson C, Ridgway EB, et al. Epilepsy surgery in young children with tuberous sclerosis: results of a novel approach. Pediatrics 2006;117(5):1494–1502 30. Carlson C, Teutonico F, Elliott RE, et al. Bilateral invasive electroencephalography in patients with tuberous scle­ rosis complex: a path to surgery? J Neurosurg Pediatr 2011;7(4):421–430 31. Romanelli P, Weiner HL, Najjar S, Devinsky O. Bilateral resective epilepsy surgery in a child with tuberous sclerosis: case report. Neurosurgery 2001;49(3):732–734, discussion 735 32. Bulacio JC, Jehi L, Wong C, et al. Long-term seizure outcome after resective surgery in patients evaluated with intracranial electrodes. Epilepsia 2012;53(10):1722–1730

18. Ma TS, Elliott RE, Ruppe V, et al. Electrocorticographic evidence of perituberal cortex epileptogenicity in tuberous sclerosis complex. J Neurosurg Pediatr 2012;10(5):376–382

33. Curatolo P, Moavero R. Tubers, epileptogenic foci, and epileptogenic networks in tuberous sclerosis. Epilepsia 2010;51(11): 2357–2359

19. Ruppe V, Dilsiz P, Reiss CS, et al. Developmental brain abnormalities in tuberous sclerosis complex: a comparative tissue analysis of cortical tubers and perituberal cortex. Epilepsia 2014;55(4): 539–550

34. Önal C, Otsubo H, Araki T, et al. Complications of invasive subdural grid monitoring in children with epilepsy. J Neurosurg 2003;98(5):1017–1026

20. Madhavan D, Weiner HL, Carlson C, Devinsky O, Kuzniecky R. Local epileptogenic networks in tuberous sclerosis complex: a case review. Epilepsy Behav 2007;11(1):140–146 21. Liang S, Zhang J, Yang Z, et al. Long-term outcomes of epilepsy surgery in tuberous sclerosis complex. J Neurol2017;264(6): 1146–1154 22. Holthausen H, Pieper T, Kudernatsch M. Towards early diagnosis and treatment to save children from catastrophic epilepsy— focus on epilepsy surgery. Brain Dev 2013;35(8):730–741 23. Cross JH, Jayakar P, Nordli D, et al; International League against Epilepsy, Subcommission for Paediatric Epilepsy Surgery.

35. Lee WS, Lee JK, Lee SA, Kang JK, Ko TS. Complications and results of subdural grid electrode implantation in epilepsy surgery. Surg Neurol 2000;54(5):346–351 36. Wiggins GC, Elisevich K, Smith BJ. Morbidity and infection in combined subdural grid and strip electrode investigation for intractable epilepsy. Epilepsy Res 1999;37(1):73–80 37. Quigg M, Harden C. Minimally invasive techniques for epilepsy surgery: stereotactic radiosurgery and other technologies. J Neurosurg 2014;121(Suppl):232–240 38. LaRiviere MJ, Gross RE. Stereotactic laser ablation for medically intractable epilepsy: the next generation of minimally invasive epilepsy surgery. Front Surg 2016;3:64

45

  Extratemporal Resection and Staged Epilepsy Surgery in Children Daxa M. Patel, Howard L. Weiner, and Robert J. Bollo

Summary In children, the predominance of epilepsy of extratemporal origin is related to developmental brain abnormalities (e.g., focal cortical dysplasia, tuberous sclerosis complex, Sturge–Weber syndrome) and low-grade cortical tumors (e.g., gangliogliomas, dysembryoplastic neuroepithelial tumors, oligodendrogliomas, and astrocytomas). Cortical dysplasia is especially prominent in the pathologic specimens of children who undergo resections for extratemporal epilepsy. Surgery for epilepsy of extratemporal origin poses a number of unique challenges, which require the team of treating physicians to engage in a rigorous preoperative evaluation, often culminating in a technically demanding surgical procedure with traditionally less successful outcomes regarding seizure freedom The surgical plan, ideally devised by a multidisciplinary team in the setting of a comprehensive epilepsy center, is based on a multitude of factors and includes a consideration of underlying pathology, neuroimaging data, electroencephalography (EEG), functional mapping data, and the specific risk–benefit profile of the individual patient. Published rates of seizure freedom after surgery for extratemporal epilepsy vary between 30 and 80%, compared with more than 80% for temporal lobe epilepsy. Keywords:  focal cortical dysplasia, pediatrics, extratemporal, epilepsy surgery, drug-resistant epilepsy, robotic stereo-EEG, laser ablation, cortical mapping, eloquent cortex, developmental brain abnormality

„„ Introduction The safety and efficacy of surgery for temporal lobe epilepsy has been well established and represents the paradigm for resective epilepsy surgery.1,​2,​3,​4,​5 By contrast, surgery for epilepsy of extratemporal origin poses a number of unique challenges, which require the team of treating physicians to engage in a rigorous preoperative evaluation, often culminating in a technically demanding surgical procedure with traditionally less successful outcomes regarding seizure freedom. Whereas anteromesial temporal lobectomy for temporal lobe epilepsy is known to carry a reasonably low morbidity, the resection necessary for extratemporal seizure foci involves larger cortical resection of the areas of ictal onset and seizure propagation anywhere outside of the temporal lobe, which often is adjacent

to or overlapping with functionally significant regions of the brain. In children, resections of extratemporal seizure foci are more common than resections of the temporal lobe. Although much of the literature pertaining to extratemporal epilepsy surgery is focused on frontal lobe epilepsy, a focal epileptic substrate may be located anywhere within the cerebral cortex. The surgical plan, ideally be devised by a multidisciplinary team in the setting of a comprehensive epilepsy center, is based on a multitude of factors and includes a consideration of underlying pathology, neuroimaging data, electroencephalography (EEG), functional mapping data, and the specific risk–benefit profile of the individual patient. Published rates of seizure freedom after surgery for extratemporal epilepsy vary between 30 and 80%, compared with more than 80% for temporal lobe epilepsy.6,​7,​8,​9,​10,​11,​12,​13 A recent meta-analysis of 36 studies including 1,259 children with a mean age of 9.8 years undergoing extratemporal resective surgery for drug-resistant focal epilepsy reported a 56% Engel Class I outcome.14 Regardless of anatomical location, the treatment goals are the same: the reduction or elimination of seizures with minimal morbidity, as well as the preservation or improvement of neurocognitive function. Several published studies have demonstrated both the safety and efficacy of extratemporal epilepsy surgery in children.6,​7,​8,​9

„„ Unique Considerations in Pediatric Extratemporal Epilepsy Surgery Over the past decade, the scientific literature describing extratemporal epilepsy surgery in children has grown enormously. Children as a unique patient population warrant special consideration for a number of reasons. First, the underlying pathology differs from that in adults. The most common cause of intractable partial epilepsy in adults is hippocampal sclerosis, classically treated surgically by anterior temporal lobectomy with amygdalohippocampectomy. In children, however, the predominance of epilepsy of extratemporal origin is related to developmental brain abnormalities (e.g., focal cortical dysplasia, tuberous sclerosis complex, Sturge–Weber syndrome) and low-grade cortical tumors (e.g., gangliogliomas, dysembryoplastic neuroepithelial tumors, oligodendrogliomas, and astrocytomas).15,​16,​17 Cortical dysplasia is especially prominent in the

438

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches pathologic specimens of children who undergo resections for extratemporal epilepsy.16,​17 By contrast, extratemporal seizure foci in adults are more varied, frequently including gliosis and focal cell loss. Compared with those in adults, outcomes reported following extratemporal seizure focus resection are better in children.18,​19,​20,​21,​22,​23,​24 Children also represent a special population because the treating physician must take into account the developmental implications of intervention on the developing nervous system. The developing brain is very sensitive to the detrimental effects of recurrent seizures, which are associated with potentially permanent neuropsychological and cognitive sequelae.25 Conversely, the plasticity of the developing brain also lends itself to better functional recovery after cortical resections that may involve eloquent cortex.15,​26,​27 It is now increasingly appreciated that uncontrolled epilepsy in childhood can have a detrimental effect on a child’s intelligence and cognitive abilities and, moreover, that epilepsy surgery performed in childhood may play a critical role in enhancing development and overall quality of life.25,​28 Surgical treatments, specifically resective surgeries with curative intent, are associated with better development outcomes compared with medical therapy alone in younger patients.1,​15,​18,​27,​29,​30,​31

„„ Presurgical Evaluation Because of the complexity of extratemporal epilepsy, patients who are being considered for surgery are best evaluated in a comprehensive epilepsy center by a multidisciplinary team of epilepsy neurologists, epilepsy surgeons, neuropsychologists, psychiatrists, and social workers. Typically, the preoperative evaluation consists of a comprehensive battery of tests, all designed to localize the epileptogenic zone and define the eloquent cortex. A thorough understanding of the potential strengths and limitations of each of these techniques is critical for appropriate patient selection and satisfactory surgical ­outcomes.

Noninvasive Modalities A detailed description of clinical seizure semiology, structural imaging with magnetic resonance imaging (MRI), and scalp video EEG capturing ictal events are the core of any preoperative epilepsy evaluation. Other noninvasive techniques are also commonly used to further localize the ictal onset zone and map functionally eloquent cortex. These include 7T MRI, functional MRI with EEG (EEG-fMRI), magnetoencephalography (MEG), single-photon emission computed tomography (SPECT), positron emission tomography (PET), and transcranial magnetic stimulation (TMS). Although previous 1.5T and 3.0T MRI devices provide adequate detection of large lesions, advanced 7T MRI structural imaging increases lesion identification and characterization, especially in focal cortical dysplasia.32,​33,​34,​35 Furthermore, they provide visualization of perivascular spaces, which are not only asymmetrically distributed but also more frequent in epileptogenic foci compared with healthy controls.36 Additionally, fMRI measurements at 7T with simultaneous EEG recordings can be

used for investigations of dynamic cortical and subcortical networks of epilepsy.37,​38 Although invasive video-EEG (iVEEG) with intracranial electrodes is thought to be superior in localizing extratemporal ictal onset zones in children undergoing resective surgery for refractory epilepsy, many studies have shown the promise of MEG as a noninvasive technique for localizing the epileptogenic zone.39,​40,​41,​42,​43,​44,​45,​46 This may partially be explained by a differential sensitivity to radial compared with tangential pacemakers, making these techniques complementary.43,​44,​47 The resection of remote ictal zones identified by MEG and electrocorticography (ECoG) does not contribute to epilepsy outcome in patients with tumors; however, in patients with cortical dysplasia, the resection of remote ictal regions may be critical for seizure outcome.40,​48 Furthermore, not only is MEG accurate in localization of the epileptogenic zone, but also the complete resection of MEG dipole clusters is predictive of seizure-free outcome after surgery.49,​50,​51,​52,​53 Peri-ictal SPECT is also a common tool evaluating focal ­metabolic abnormalities in the evaluation of drug-resistant extratemporal epilepsy in children.54 Subtraction ictal SPECT scanning coregistered to MRI (SISCOM) improves the sensitivity of ictal SPECT for localization of the ictal onset zone and has demonstrated clinical utility in guiding placement of intracranial electrodes.55,​56,​57 Further improvement in accuracy of ictal SPECT is achieved by statistical parametric mapping (SPM), where ictal SPECT is compared with normal brain SPECT to identify regions of statistically significant alteration in regional cerebral blood flow related to seizure focus.52 Ictal SPECT and SISCOM may be especially useful in providing preoperative guidance for intracranial electrode placement in children with frontal lobe epilepsy, as rapid seizure spread often renders false-localizing clinical and electrophysiologic data.58,​59 Like SPECT, PET uses radiolabeled tracers to image cerebral glucose metabolism and, less frequently, protein metabolism and gamma-aminobutyric acid and serotonin receptor density.60 Interictal fluorodeoxyglucose (FDG)-PET has a reported sensitivity of 60 to 80% for identification of foci of hypometabolism among patients with chronic, refractory extratemporal epilepsy and a normal structural brain MRI; this sensitivity is similar to that reported for similar patients with ictal SPECT.52,​61,​62,​63 However, similar to MEG and SPECT, the clinical utility of FDG-PET in presurgical seizure focus localization is relatively limited compared with invasive seizure monitoring with intracranial electrodes. Studies in children with tuberous sclerosis complex who have chronic, refractory epilepsy have shown that FDG-PET may complement MRI with diffusion-weighted imaging (DWI) in differentiating epileptogenic from clinically silent tubers.64

Functional Mapping Precise identification of eloquent cortex essential for sensorimotor, language, visual, and memory function and their neuroanatomical relationship to the epileptogenic zone is critical for surgical risk assessment and decision making in extratemporal epilepsy. Several noninvasive imaging techniques, including fMRI, resting-state fMRI, MEG, FDG-PET, and, most recently, TMS, provide accurate maps of primary sensorimotor cortex in children.

45  Extratemporal Resection and Staged Epilepsy Surgery in Children One main disadvantage of fMRI is that it identifies all structures involved in a task instead of those areas that are essential for it, which is frequently challenging in young children. ­Resting-state fMRI is capable of demonstrating functional brain networks without task performance and under sedation, and adds to the previous noninvasive functional monitoring by revealing an association between frontal lobe epilepsy and motor network disruption.65,​66,​67 TMS incorporates the benefits of direct cortical stimulation mapping, identifying essential cortex for a sensorimotor function, in an outpatient setting without task performance.68,​69 Cortical stimulation mapping may also be performed invasively, via direct cortical stimulation following craniotomy, either in the operating room or extraoperatively via implanted subdural electrodes. Wada testing (intracarotid amobarbital test) is useful for establishing the laterality of language and memory function in cooperative children. In addition, a thorough neuropsychological evaluation is essential not only for providing a preoperative baseline but also for potentially providing corroborative data to a suspected ictal focus. Comparison of preoperative and postoperative neuropsychiatric tests may help determine whether ictal behavior has been altered.70,​71

„„ Surgical Techniques The major factors that potentially complicate the surgical intervention in extratemporal epilepsy include the multifocality of seizure foci, the higher incidence of nonlesional MRI-negative epilepsy, and the frequent proximity of the epileptogenic zone to eloquent cortex. As a result, surgical strategies should take into consideration a wide array of therapeutic options, which must be tailored to the individual patient’s risk–benefit profile. When a comprehensive noninvasive preoperative evaluation does not reveal an apparent localized seizure focus, the treating team of specialists is left with a dilemma regarding how to proceed. The options include either no surgery or the use of a palliative procedure, such as vagal nerve stimulation or corpus ­callosotomy, in the appropriate clinical setting. In patients with focal seizure semiology and a discordant or nonlocalizing presurgical evaluation, many centers perform stereoelectroencephalography (SEEG) or stereotactically place multiple intracranial electrodes to evaluate various hypotheses regarding the location of the epileptogenic zone based on noninvasive studies. In other circumstances, further refinement of neuroanatomic relationships between the epileptogenic zone and functional cortex require staged surgical approaches with large subdural electrode arrays. SEEG offers excellent spatial resolution of seizure onset, is minimally invasive, and is generally very well tolerated with a low complication rate.72,​73 Although depth electrodes record local EEG signal around a 2- to 3-mm radius from each contact, they allow surveillance of deep regions, such as the insula and cingulum, that are challenging to sample with subdural electrodes. Additionally, SEEG allows for bilateral and noncontiguous lobar sampling.74,​75 The rate of complications of SEEG is less than 1.3%, and the most common complication is infection.76,​ 77 Although controversy still remains regarding the roles and indications of intracranial EEG recordings via both SEEG and

s­ubdural electrode strategies, some reports suggest SEEG is successful in identifying epileptic zone in 80 to 90% of cases.48 In the context of extratemporal lesional epilepsy where concerns exist that the epileptogenic zone extends beyond the boundaries of the radiographic lesion, the extent of ictal zone can be identified by either intraoperative ECoG, SEEG evaluation around the margins of the lesion, or subdural electrodes placed over the lesion to define this extraoperatively. Intraoperatively, ECoG may be used to assist in the identification of the epileptogenic zone. Limitations of this technique include analysis of interictal data only and significantly diminished interictal spike frequency under general anesthesia. Technically, this procedure is similar in adults and children: electrodes are placed over the putative epileptogenic zone and interictal electrical spike activity is characterized. It is generally reliable if an interictal spike frequency reaches at least one spike per minute. This technique is widely applied in children with brain tumors and refractory epilepsy, where the ictal onset zone is frequently localized to cortex adjacent to the tumor and the resection of perilesional cortex with abnormal interictal spike activity correlates with long-term seizure freedom.78,​79 Although there are clearly limitations to performing intraoperative ECoG in awake pediatric patients, chronic implanted subdural electrodes not only seem to be well tolerated in this population but also may yield more useful data given the longer sampling times and ability to capture ictal data directly.27,​ 80 With the two-stage approach, the epilepsy team is able to develop a map of the ictal zone and the associated epileptogenic network that would need to be addressed surgically (Fig. 45.1). Extraoperative ictal focus mapping via stimulation across subdural electrodes (arrays of 2.5-mm platinum electrodes either 5 or 10 mm apart, mounted on silastic strips, implanted under general anesthesia) is a well-established technique for ictal focus mapping.81,​82 Reports in children with extratemporal epilepsy indicate a sensitivity of approximately 90%.83 A disadvantage to the two-stage approach is the need for a second operation for electrode implantation and inherent increased surgical risks.81 Most centers report an overall complication rate of approximately 10 to 20% among patients with chronic intracranial electrodes.84,​85,​86,​87,​88,​89,​90 Commonly reported complications include cerebrospinal fluid (CSF) leak or positive CSF cultures, usually in the absence of clinically evident meningitis.80,​85,​86,​88,​90 Other reported complications include transient neurologic deficit, epidural or subdural hematoma, and stroke. A reduction in complication rate with increasing surgical experience has been reported.84,​87 Class 2 data indicate dexamethasone may reduce cerebral swelling in children with implanted subdural grid electrode arrays. However, it also decreases the seizure frequency, which may lead to longer extraoperative monitoring periods to capture sufficient data to localize the ictal focus.91 Studies of the p ­ athologic changes seen in cortex underlying subdural arrays have revealed focal, transient aseptic meningitis in all patients. However, the severity of this reaction does not correlate with the incidence of infection or long-term surgical outcome.92 In addition to the ictal data, functional data can be obtained using the same grid with extraoperative cortical stimulation mapping. In many of these patients, the extent of resection of the seizure focus may be limited by the extent of involvement of

439

440

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 45.1  Placement of intracranial electrodes. (a) Custom pediatric 64-contact subdural grid array with electrodes spaced 5 mm apart and cables exiting the lateral surface to minimize mass effect. (b) 64-contact grid (1-cm electrode spacing) is contoured to match regional cortical anatomy and allow circulation of CSF. (c) Grid in situ. (d) Placement of depth electrode perpendicular to the cortical surface after bipolar coagulation of the pial surface.

eloquent cortex (Fig. 45.2). Since awake cortical mapping may not be possible in younger pediatric patients, other modalities such as extraoperative cortical stimulation mapping are frequently necessary to define eloquent cortex, especially language cortex. Once these regions have been satisfactorily mapped, the surgeon can discuss the various treatment options with the patient’s family in light of the specific risk–benefit considerations. The refinement of neuronavigational techniques over the last several years has also aided the epilepsy surgeon. Both frameless and frame-based systems have been used in placing monitoring electrodes, particularly depth electrodes, and to aid in resection. Three-dimensional MRI reconstructions, functional imaging data, and MR angiography have all been used effectively to help delineate the extent of resection.93 For a select group of patients with the most complex manifestations of extratemporal epilepsy, some experienced centers with low complication rates have used a multistage surgical approach.80 This typically consists of electrode implantation, an extraoperative monitoring period, resection and electrode reimplantation, a second monitoring period, and a third operation for electrode removal and further resection if necessary. The additional monitoring period often reveals secondary ictal foci that only become apparent after removal of the primary seizure focus (Fig. 45.3). Criticisms of this philosophy include the additional risk of further invasive monitoring and an additional surgery. In the authors’ experience, a second intraoperative monitoring period followed by a third surgical stage is simpler than a return to the operating room several months later. In this select group of pediatric patients with poor surgical prognostic factors (i.e., multifocal ictal onset, ictal overlap with eloquent cortex, or previous surgical failure), we believe that this additional stage has the potential to improve surgical outcomes. We have postulated that it is difficult to assess adequately the potential of secondary seizure foci to become independent ictal generators once the primary focus has been resected. In addition, acute seizures after ictal focus resection reflect a poor prognosis.94—​96 An alternative method for localizing secondary foci is intraoperative ECoG after resection; however,

this method relies on capturing interictal data under general anesthesia, as opposed to capturing ictal and interictal events over a longer period of time. Nonetheless, the justification for a multistaged surgical procedure remains controversial and necessitates additional study before any conclusions regarding treatment protocols in this difficult subset of patients can be drawn. In addition, the recent popularity of SEEG has made multiple evaluations with intracranial electrodes safer and better tolerated. The development of MRI-guided stereotactic laser ablation has also allowed minimally invasive diagnostic procedures to often be coupled with a minimally invasive treatment strategy.73,​97

Implantation of Subdural Electrodes Grid, strip, and depth electrodes are placed by performing a stereotactically guided craniotomy in the brain region of interest. Typically, patients are placed in cranial pins for skull fixation. Because brain relaxation is critical, patients are placed in the reverse Trendelenburg position and may be hyperventilated for the initial phase of surgery. A pericranial dural graft is frequently harvested prior to the craniotomy to use for watertight closure of the dura at the time of closure and for augmenting the subdural space; alternatively, dural substitutes such as Durepair (Medtronic, Minneapolis, MN) can be used. This has virtually eliminated any incidence of symptomatic mass effect from the subdural electrodes in our experience. Electrodes are placed with frameless stereotactic image guidance, and the location of each electrode is recorded by the surgical team and by the nursing staff. Electrode wires are secured to the dural edges with 4–0 Nurolon or Vicryl interrupted sutures. Wires are tunneled to an adjacent region of the scalp with a trocar device and are secured in place with 4–0 Nurolon or 3–0 silk purse string sutures, and then to one another with a 0 Prolene suture. A Jackson–Pratt drain is left in the subgaleal space for the duration of monitoring to divert CSF from the incision and electrode exit sites. We have also utilized frameless stereotactic image

45  Extratemporal Resection and Staged Epilepsy Surgery in Children

Fig. 45.2  Ictal and functional mapping via intracranial electrodes. (a) Ictal focus map based on accumulated ictal and interictal data demonstrates regions of seizure onset, regional electrodecrements, and interictal discharges in multilobar epilepsy. (b) Corresponding functional map demonstrating language function at the posterior border of electrophysiologically abnormal cortex. All electrodes shown were tested against at least one adjacent electrode. Functional results and clinically relevant electrodes with no evident function are shown.

Fig. 45.3  Multistage approach. Replacement of 64-contact subdural grid array (5-mm electrode spacing) after extratemporal ictal focus resection. The grid is contoured (a) and laid over the resection cavity (b). Postoperative T2-weighted axial MRI (c) demonstrates grid overlying resection cavity after the second stage of a three-stage procedure.

441

442

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches guidance for insertion of depth electrodes through the subdural electrode array. In our experience, the depth electrodes have been particularly useful for monitoring deep lesions and more remote cortex, such as the mesial frontal and parietal areas. Culture swabs of the epi- and subdural spaces are taken at the subsequent surgical stages. MRI scans are obtained postoperatively to document electrode position and evaluate for complications such as intracranial hemorrhage ischemia or significant mass effect. Whereas some surgeons choose to freeze the bone flap during the period of invasive monitoring, it has been our practice to leave the bone flap in situ. In our experience, irrigation is utilized very liberally at each stage of surgery, and surgical gloves are changed at least twice during an epilepsy operation.

Stereoelectroencephalography SEEG electrodes are placed either with a surgical robot (ROSA, Zimmer Biomet, Montpellier, France) or with conventional ­stereotactic techniques (frameless or frame based). For robotic SEEG placement (Fig. 45.4), the patient is placed in three-point cranial fixation using the Mayfield head holder or stereotactic head frame; subsequently, the robot is brought in and aligned with a swivel stud adapter and secured. The table is locked and then unplugged to prevent movement. Next, laser registration is performed. Once this is complete with low error margins, the preoperative electrodes, whose entry point, trajectory, and target were planned preoperatively according to the structural abnormalities or hypothesized seizure foci, are assessed for feasibility. Then the entire scalp is prepped and draped in sterile fashion. The robot is driven to the first electrode placement entry point, with the adaptor for the drill placed at the end of the robot arm. The drill is placed within the adaptor and adjusted to the entry point. Once the entry point is marked, infiltrated with local anesthetic, and incised, a twist drill hole is created, and the drill is removed from the adaptor. A bolt is placed on a screwdriver through the adaptor within the robot arm and screwed into the twist drill hole. Once the bolt is secured, the robot calculates the distance from the top of the adaptor to the

top of the bolt. A moveable adaptor on the screwdriver is then tightened at the top of the robot arm, and the screwdriver is removed. The distance from the adapter to the end of the screwdriver is measured and subtracted from the distance calculated by the robot to get the distance from the top of the bolt to the target. Subsequently, this is measured accurately on the obturator and electrode. Next, the obturator is placed down the bolt and removed, and then electrode is passed, the cap is engaged with the threads on the bolt, the stylet is removed, and the cap is secured. All planned electrode trajectories are placed in similar manner. The drapes are carefully removed, the robot is disconnected from the Mayfield head holder and backed away, and the patient is removed from cranial fixation. The patient is transferred to the intensive care unit for initial postoperative care. The patient can be transferred out of the intensive care to the epilepsy monitoring unit after p ­ ostoperative ­stereotactic imaging, which may be merged to preoperative imaging to assess the accuracy of each electrode. With conventional frameless stereotactic techniques,97,​98,​99 the patient is placed in three-point cranial fixation typically with the head elevated and the neck gently flexed with a gel roll beneath the shoulders, so the surgeon has access to the entire calvaria. Then a frameless stereotactic system, on which several electrode trajectories have previously been planned, is registered to the patient. The patient is prepped and draped in sterile fashion. The stereotactic arm is assembled and brought into the position of first electrode trajectory. Once the arm is aligned with the entry point and target and the trajectory is verified as safe, a reducing tube is placed within the stereotactic arm, and the drill is passed within the reducing tube to the entry point. The point is marked on the scalp, which is infiltrated with local anesthetic, a stab incision is created, and a twist drill hole is created. The drill is removed, and a bolt is securely placed. Subsequently, the entry point to the top of the bolt is marked to determine the depth of the electrode, and the obturator and the electrode catheter are set to this depth. Similar to the robotic method, the obturator is placed down to the target and then removed, and the electrode is passed. Frame-based SEEG methods are similar to previously described frame-based stereotactic methods.77

Fig. 45.4  (a) A patient undergoing robotic-assisted stereotactic depth electrode placement. (b) Postoperative stereotactic CT fused to preoperative MRI using robotic software demonstrating electrode trajectories.

45  Extratemporal Resection and Staged Epilepsy Surgery in Children

Fig. 45.5  Craniotomy for extratemporal seizure focus resection. (a) The lateral decubitus position allows the neck to remain in line with the body, preventing jugular venous compression. (b) Bipolar coagulation of the cortical surface over the area to be resected. (c) Right parietal ictal focus resection. Following pial coagulation, ultrasonic aspiration is used to empty all gray matter from gyri containing the mapped ictal focus as permitted by functional anatomy. (d) Multifocal resection. Multifocal epilepsy involving purely extratemporal or temporal and extratemporal regions of seizure onset is common in the pediatric population. In addition, evidence suggests acute postresection seizures predict a poor prognosis. This supports the strategy of multistage procedures, where secondary ictal foci not evident during the initial invasive monitoring period may be identified in a second monitoring session and treated surgically with a second resection.

Seizure Focus Resection The surgical resection techniques used in epilepsy surgery are unique and critical for its success. The epileptogenic cortex typically involves specific gyri of the brain, and its safe removal therefore requires absolute preservation of arteries and veins in the subarachnoid space, as well as the underlying subcortical white matter. To accomplish the safe resection of this cortical tissue, we have found ultrasonic aspiration (e.g., the CUSA, Integra, Plainsboro, NJ) to be an indispensable tool. The initial cortical incision is performed using bipolar cautery and microscissors to coagulate and open the pia. The CUSA can then be used on a low setting to safely resect the gray matter tissue filling any gyrus in a subpial fashion, with total preservation of the underlying pia and blood vessels. The cortex may also be sharply dissected in a subpial fashion along sulci with sharp dissection with a Penfield or Rhoton dissector. Bleeding from the raw edge of the exposed pia after resection is easily managed by placing small square strips of surgical gauze (Fig. 45.5). When performing extratemporal resections, it is critical to ensure that the resection is completed all the way to the pial surface and that the gyrus in question has been emptied of all gray matter. This is especially true when the surgery calls for resecting the mesial frontal, parietal, or occipital cortex, in which case the surgeon must be sure to visualize the mesial pia to be confident that the resection is complete.

Stereotactic Laser Ablation More recently, MR-guided laser interstitial thermotherapy (MRgLITT) or stereotactic laser ablation offers a minimally invasive treatment option in select cases.73,​97 In this technique, thermal energy delivered via a stereotactically placed cooled laser is used to ablate the epileptogenic zone while monitoring in real time using MR thermography. Although its safety and efficacy in pediatric patients remain uncertain, early results for deep-­seated

ictal generators in children such as hypothalamic hamartomas appear excellent.100,​101,​102,​103 A recent case series on children with epileptic foci treated with LITT demonstrated ­seizure-free outcomes at 1 year without any complications.104 Overall, MRgLITT is associated with less operative risk and shorter length of hospitalization. Although transient neurologic deficits due to edema and damage to surrounding structures have been reported, MRgLITT remains a safe and viable treatment option for selective pediatric patients with extratemporal epilepsy.105

„„ Results Overall, children with drug-resistant epilepsy who have undergone surgery have a significantly higher rate of seizure freedom and better behavior and quality-of-life improvement at 1 year compared with those treated with medical therapy.31 ­ Nevertheless, extratemporal resections remain a significant challenge in pediatric epilepsy surgery. Multiple single-­ ­ institutional cohorts and several systematic reviews indicate that the 1-year seizure-free rate for extratemporal resection is 70%, but there is a gradual decrease in seizure freedom to approximately 50% at 5 years.14,​106,​107,​108,​109 Although seizure freedom rates as high as 80% have been reported in some published pediatric extratemporal epilepsy surgery series, these usually include a large proportion of lesional cases such as neoplasms, which are biased toward a more favorable outcome.10 In general, the success rate for ­pediatric extratemporal epilepsy surgery has been consistently inferior to that for temporal lobe epilepsy. Two long-term follow-up studies in pediatric populations reported Engel Class I outcomes of 78 and 74% for temporal resections and 54 and 60% for extratemporal resections.11,​15 Similarly, pediatric patients with nonlesional epilepsy who undergo surgery do not achieve surgical outcomes comparable with those with lesions noted on MRI.3,​12 However, most reports describe a retrospective

443

444

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches analysis of a small group of patients treated at a single institution. A recent meta-analysis reports a seizure freedom rate of 56% and indicates that pathological structures visualized on MRI, localizing ictal electrophysiologic data, and short duration of seizures with absence of secondary generalization predict long-term seizure freedom.14 Acute postoperative seizures consistently predict a significantly lower rate of long-term seizure freedom.96,​100,​101,​102,​103,​104,​105,​106,​107,​108,​109,​110,​111,​112 In the authors’ report of a consecutive series of patients who underwent multistage surgery, we observed an Engel Class I outcome in 60% of children, with 87% of the group demonstrating worthwhile improvement (Engel Class III or higher).80 In addition, retrospective data suggest similar improvements in quality of life among patients undergoing temporal and extratemporal resections.113 Other data in children have suggested similar long-term outcome compared with adults, despite

a much higher frequency of extratemporal seizures in this population.114 Most authors recommend early surgical intervention in children with refractory seizures.96,​111,​112,​114

„„ Conclusion Despite significant technological advances that have been made in functional neuroimaging, neuronavigation, and neuromonitoring, the epileptogenic zone in pediatric extratemporal ­ epilepsy often remains elusive. With enhanced surgical innovation and new minimally invasive diagnostic and ­therapeutic options, however, we remain optimistic that more effective treatment algorithms are forthcoming that will r­ender patients seizure-free with the lowest possible ­surgical­morbidity.

References 1. Wiebe S, Blume WT, Girvin JP, Eliasziw M; Effectiveness and Efficiency of Surgery for Temporal Lobe Epilepsy Study Group. A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med 2001;345(5):311–318 2. Yasuda CL, Tedeschi H, Oliveira EL, et al. Comparison of shortterm outcome between surgical and clinical treatment in temporal lobe epilepsy: a prospective study. Seizure 2006;15(1):35–40 3. Engel J Jr, van Ness P, Rasmussen T, Ojemann L. Outcome with respect to epileptic seizures. In: Engel J Jr, ed. Surgical Treatment of the Epilepsies. 2nd ed. New York, NY: Raven; 1993:609–621 4. Kellett MW, Smith DF, Baker GA, Chadwick DW. Quality of life after epilepsy surgery. J Neurol Neurosurg Psychiatry 1997;63(1):52–58 5. Birbeck GL, Hays RD, Cui X, Vickrey BG. Seizure reduction and quality of life improvements in people with epilepsy. Epilepsia 2002;43(5):535–538 6. Wyllie E, Lüders H, Morris HH III, et al. Subdural electrodes in the evaluation for epilepsy surgery in children and adults. Neuropediatrics 1988;19(2):80–86 7. Morrison G, Duchowny M, Resnick T, et al. Epilepsy s­urgery in childhood. A report of 79 patients. Pediatr Neurosurg 1992;18(5–6):291–297 8. Rossi GF. Epilepsy in the pediatric age and its surgical treatment. Childs Nerv Syst 1995;11(1):23–28 9. Adelson PD, Black PM, Madsen JR, et al. Use of subdural grids and strip electrodes to identify a seizure focus in children. Pediatr Neurosurg 1995;22(4):174–180 10. Pomata HB, González R, Bartuluchi M, et al. Extratemporal epilepsy in children: candidate selection and surgical treatment. Childs Nerv Syst 2000;16(12):842–850 11. Van Oijen M, De Waal H, Van Rijen PC, Jennekens-Schinkel A, van Huffelen AC, Van Nieuwenhuizen O; Dutch Collaborative Epilepsy Surgery Program. Resective epilepsy surgery in childhood: the Dutch experience 1992–2002. Eur J Paediatr Neurol 2006;10(3):114–123 12. Sinclair DB, Aronyk K, Snyder T, et al. Extratemporal resection for childhood epilepsy. Pediatr Neurol 2004;30(3):177–185 13. Quesney LF. Extratemporal epilepsy: clinical presentation, pre-operative EEG localization and surgical outcome. Acta Neurol Scand Suppl 1992;140(S140):81–94 14. Englot DJ, Breshears JD, Sun PP, Chang EF, Auguste KI. Seizure outcomes after resective surgery for extra-temporal lobe epilepsy in pediatric patients. J Neurosurg Pediatr 2013;12(2):126–133

15. Wyllie E, Comair YG, Kotagal P, Bulacio J, Bingaman W, Ruggieri P. Seizure outcome after epilepsy surgery in children and adolescents. Ann Neurol 1998;44(5):740–748 16. Paolicchi JM, Jayakar P, Dean P, et al. Predictors of outcome in pediatric epilepsy surgery. Neurology 2000;54(3):642–647 17. Blumcke I, Spreafico R, Haaker G, et al; EEBB Consortium. Histopathological findings in brain tissue obtained during epilepsy surgery. N Engl J Med 2017;377(17):1648–1656 18. Cascino GD. Surgical treatment for extratemporal epilepsy. Curr Treat Options Neurol 2004;6(3):257–262 19. Zentner J, Hufnagel A, Ostertun B, et al. Surgical treatment of extratemporal epilepsy: clinical, radiologic, and histopathologic findings in 60 patients. Epilepsia 1996;37(11):1072–1080 20. Cascino GD, Jack CR Jr, Parisi JE, et al. MRI in the presurgical evaluation of patients with frontal lobe epilepsy and children with temporal lobe epilepsy: pathologic correlation and prognostic importance. Epilepsy Res 1992;11(1):51–59 21. Kim DW, Lee SK, Yun CH, et al. Parietal lobe epilepsy: the semiology, yield of diagnostic workup, and surgical outcome. Epilepsia 2004;45(6):641–649 22. Lee JJ, Lee SK, Lee SY, et al. Frontal lobe epilepsy: clinical characteristics, surgical outcomes and diagnostic modalities. Seizure 2008;17(6):514–523 23. Olivier A. Surgery of frontal lobe epilepsy. Adv Neurol 1995;66:321–348, discussion 348–352 24. Jobst BC, Siegel AM, Thadani VM, Roberts DW, Rhodes HC, ­Williamson PD. Intractable seizures of frontal lobe origin: clinical characteristics, localizing signs, and results of surgery. Epilepsia 2000;41(9):1139–1152 25. Berg AT, Zelko FA, Levy SR, Testa FM. Age at onset of epilepsy, pharmacoresistance, and cognitive outcomes: a prospective cohort study. Neurology 2012;79(13):1384–1391 26. Mizrahi EM, Kellaway P, Grossman RG, et al. Anterior temporal lobectomy and medically refractory temporal lobe epilepsy of childhood. Epilepsia 1990;31(3):302–312 27. Centeno RS, Yacubian EM, Sakamoto AC, Ferraz AF, Junior HC, Cavalheiro S. Pre-surgical evaluation and surgical treatment in children with extratemporal epilepsy. Childs Nerv Syst 2006;22(8):945–959 28. Loddenkemper T, Holland KD, Stanford LD, Kotagal P, Bingaman W, Wyllie E. Developmental outcome after epilepsy surgery in infancy. Pediatrics 2007;119(5):930–935 29. Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000;342(5):314–319

45  Extratemporal Resection and Staged Epilepsy Surgery in Children 30. Otsuki T, Kim HD, Luan G, et al; FACE Study Group. Surgical versus medical treatment for children with epileptic encephalopathy in infancy and early childhood: Results of an international multicenter cohort study in Far-East Asia (the FACE study). Brain Dev 2016;38(5):449–460

51. Fischer MJ, Scheler G, Stefan H. Utilization of magnetoencephalography results to obtain favourable outcomes in epilepsy surgery. Brain 2005;128(Pt 1):153–157

31. Dwivedi R, Ramanujam B, Chandra PS, et al. Surgery for drugresistant epilepsy in children. N Engl J Med 2017;377(17): 1639–1647

53. Knowlton RC, Elgavish RA, Bartolucci A, et al. Functional imaging: II. Prediction of epilepsy surgery outcome. Ann Neurol 2008;64(1):35–41

32. Madan N, Grant PE. New directions in clinical imaging of cortical dysplasias. Epilepsia 2009;50(Suppl 9):9–18 33. Obusez EC, Lowe M, Oh SH, et al. 7T MR of intracranial pathology: preliminary observations and comparisons to 3T and 1.5T. Neuroimage 2018;168:459–476 34. Zucca I, Milesi G, Medici V, et al. Type II focal cortical dysplasia: ex vivo 7T magnetic resonance imaging abnormalities and histopathological comparisons. Ann Neurol 2016;79(1):42–58 35. De Ciantis A, Barba C, Tassi L, et al. 7T MRI in focal epilepsy with unrevealing conventional field strength imaging. Epilepsia 2016;57(3):445–454 36. Feldman RE, Rutland JW, Fields MC, et al. Quantification of perivascular spaces at 7T: a potential MRI biomarker for epilepsy. Seizure 2018;54:11–18 37. Andersson P, Pluim JP, Siero JC, Klein S, Viergever MA, Ramsey NF. Real-time decoding of brain responses to visuospatial attention using 7T fMRI. PLoS One 2011;6(11):e27638 38. Nersesyan H, Hyder F, Rothman DL, Blumenfeld H. Dynamic fMRI and EEG recordings during spike-wave seizures and generalized tonic-clonic seizures in WAG/Rij rats. J Cereb Blood Flow Metab 2004;24(6):589–599 39. Oishi M, Kameyama S, Masuda H, et al. Single and multiple clusters of magnetoencephalographic dipoles in neocortical epilepsy: significance in characterizing the epileptogenic zone. Epilepsia 2006;47(2):355–364 40. Otsubo H, Ochi A, Elliott I, et al. MEG predicts epileptic zone in lesional extrahippocampal epilepsy: 12 pediatric surgery cases. Epilepsia 2001;42(12):1523–1530 41. Knowlton RC, Laxer KD, Aminoff MJ, Roberts TP, Wong ST, Rowley HA. Magnetoencephalography in partial epilepsy: clinical yield and localization accuracy. Ann Neurol 1997;42(4): 622–631 42. Minassian BA, Otsubo H, Weiss S, Elliott I, Rutka JT, Snead OC III. Magnetoencephalographic localization in pediatric epilepsy surgery: comparison with invasive intracranial electroencephalography. Ann Neurol 1999;46(4):627–633 43. Wheless JW, Willmore LJ, Breier JI, et al. A comparison of magnetoencephalography, MRI, and V-EEG in patients evaluated for epilepsy surgery. Epilepsia 1999;40(7):931–941 44. Papanicolaou AC, Pataraia E, Billingsley-Marshall R, et al. Toward the substitution of invasive electroencephalography in epilepsy surgery. J Clin Neurophysiol 2005;22(4):231–237 45. Stefan H, Trinka E. Magnetoencephalography (MEG): past, current and future perspectives for improved differentiation and treatment of epilepsies. Seizure 2017;44:121–124 46. Englot DJ, Nagarajan SS, Imber BS, et al. Epileptogenic zone localization using magnetoencephalography predicts seizure freedom in epilepsy surgery. Epilepsia 2015;56(6):949–958 47. Jansen FE, Huiskamp G, van Huffelen AC, et al. Identification of the epileptogenic tuber in patients with tuberous sclerosis: a comparison of high-resolution EEG and MEG. Epilepsia 2006;47(1):108–114 48. Jayabal V, Pillai A, Sinha S, et al. Role of magnetoencephalography and stereo-electroencephalography in the presurgical evaluation in patients with drug-resistant epilepsy. Neurol India 2017;65(Supplement):S34–S44 49. Bagić A. Look back to leap forward: the emerging new role of magnetoencephalography (MEG) in nonlesional epilepsy. Clin Neurophysiol 2016;127(1):60–66 50. Knowlton RC, Razdan SN, Limdi N, et al. Effect of epilepsy magnetic source imaging on intracranial electrode placement. Ann Neurol 2009;65(6):716–723

52. Knowlton RC. The role of FDG-PET, ictal SPECT, and MEG in the epilepsy surgery evaluation. Epilepsy Behav 2006;8(1):91–101

54. Pirotte B, Goldman S, Salzberg S, et al. Combined positron emission tomography and magnetic resonance imaging for the planning of stereotactic brain biopsies in children: experience in 9 cases. Pediatr Neurosurg 2003;38(3):146–155 55. Buchhalter JR, So EL. Advances in computer-assisted single-photon emission computed tomography (SPECT) for epilepsy surgery in children. Acta Paediatr Suppl 2004;93(445):32–35, discussion 36–37 56. Van Paesschen W. Ictal SPECT. Epilepsia 2004;45(Suppl 4):35–40 57. Ahnlide JA, Rosén I, Lindén-Mickelsson Tech P, Källén K. Does SISCOM contribute to favorable seizure outcome after epilepsy surgery? Epilepsia 2007;48(3):579–588 58. Lee SK, Lee SY, Yun CH, Lee HY, Lee JS, Lee DS. Ictal SPECT in neocortical epilepsies: clinical usefulness and factors affecting the pattern of hyperperfusion. Neuroradiology 2006;48(9):678–684 59. Fukuda M, Masuda H, Honma J, Kameyama S, Tanaka R. Ictal SPECT analyzed by three-dimensional stereotactic surface projection in frontal lobe epilepsy patients. Epilepsy Res 2006;68(2):95–102 60. Duncan JD, Moss SD, Bandy DJ, et al. Use of positron emission tomography for presurgical localization of eloquent brain areas in children with seizures. Pediatr Neurosurg 1997;26(3):144–156 61. Juhász C, Chugani HT. Imaging the epileptic brain with positron emission tomography. Neuroimaging Clin N Am 2003;13(4):705– 716, viii 62. Sood S, Chugani HT. Functional neuroimaging in the preoperative evaluation of children with drug-resistant epilepsy. Childs Nerv Syst 2006;22(8):810–820 63. Verger A, Lagarde S, Maillard L, Bartolomei F, Guedj E. Brain molecular imaging in pharmacoresistant focal epilepsy: current practice and perspectives. Rev Neurol (Paris) 2018;174(1–2):16–27 64. Chandra PS, Salamon N, Huang J, et al. FDG-PET/MRI coregistration and diffusion-tensor imaging distinguish epileptogenic tubers and cortex in patients with tuberous sclerosis complex: a preliminary report. Epilepsia 2006;47(9):1543–1549 65. Morioka T, Mizushima A, Yamamoto T, et al. Functional mapping of the sensorimotor cortex: combined use of magnetoencephalography, functional MRI, and motor evoked potentials. Neuroradiology 1995;37(7):526–530 66. Woodward KE, Gaxiola-Valdez I, Goodyear BG, Federico P. Frontal lobe epilepsy alters functional connections within the brain’s motor network: a resting-state fMRI study. Brain Connect 2014;4(2):91–99 67. Vadivelu S, Wolf VL, Bollo RJ, Wilfong A, Curry DJ. Resting-state functional MRI in pediatric epilepsy surgery. Pediatr Neurosurg 2013;49(5):261–273 68. Theodore WH. Transcranial magnetic stimulation in epilepsy. Epilepsy Curr 2003;3(6):191–197 69. Macdonell RA, Jackson GD, Curatolo JM, et al. Motor cortex localization using functional MRI and transcranial magnetic stimulation. Neurology 1999;53(7):1462–1467 70. Jones-Gotman M, Smith M, Zatorre R. Neuropsychological testing for localizing and lateralizing the epileptogenic region. In: Engel J Jr, ed. Surgical Treatment of the Epilepsies. 2nd ed. New York, NY: Raven; 1993:245–261 71. Bernstein JH, Prather PA, Rey-Casserly C. Neuropsychological assessment in preoperative and postoperative evaluation. Neurosurg Clin N Am 1995;6(3):443–454 72. González-Martínez J, Bulacio J, Thompson S, et al. Technique, results, and complications related to robot-assisted stereoelectroencephalography. Neurosurgery 2016;78(2):169–180

445

446

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches 73. Gonzalez-Martinez J, Mullin J, Bulacio J, et al. Stereoelectroencephalography in children and adolescents with difficult-to-localize refractory focal epilepsy. Neurosurgery 2014;75(3):258–268, discussion 267–268 74. Gonzalez-Martinez J, Bulacio J, Alexopoulos A, Jehi L, Bingaman W, Najm I. Stereoelectroencephalography in the “difficult to localize” refractory focal epilepsy: early experience from a North American epilepsy center. Epilepsia 2013;54(2): 323–330 75. Gonzalez-Martinez J, Lachhwani D. Stereoelectroencephalography in children with cortical dysplasia: technique and results. Childs Nerv Syst 2014;30(11):1853–1857 76. Mullin JP, Shriver M, Alomar S, et al. Is SEEG safe? A systematic review and meta-analysis of stereo-electroencephalography-related complications. Epilepsia 2016;57(3):386–401 77. Cardinale F, Cossu M, Castana L, et al. Stereoelectroencephalography: surgical methodology, safety, and stereotactic application accuracy in 500 procedures. Neurosurgery 2013;72(3):353–366, discussion 366 78. Pilcher WH, Silbergeld DL, Berger MS, Ojemann GA. Intraoperative electrocorticography during tumor resection: impact on seizure outcome in patients with gangliogliomas. J Neurosurg 1993;78(6):891–902 79. Ojemann SG, Berger MS, Lettich E, Ojemann GA. Localization of language function in children: results of electrical stimulation mapping. J Neurosurg 2003;98(3):465–470 80. Bauman JA, Feoli E, Romanelli P, Doyle WK, Devinsky O, Weiner HL. Multistage epilepsy surgery: safety, efficacy, and utility of a novel approach in pediatric extratemporal epilepsy. Neurosurgery 2005;56(2):318–334 81. Tharin S, Golby A. Functional brain mapping and its applications to neurosurgery. Neurosurgery 2007;60(4, Suppl 2):185– 201, discussion 201–202 82. Duffau H. Lessons from brain mapping in surgery for low-grade glioma: insights into associations between tumour and brain plasticity. Lancet Neurol 2005;4(8):476–486 83. Bruce DA, Bizzi JW. Surgical technique for the insertion of grids and strips for invasive monitoring in children with intractable epilepsy. Childs Nerv Syst 2000;16(10–11):724–730 84. Hamer HM, Morris HH, Mascha EJ, et al. Complications of invasive video-EEG monitoring with subdural grid electrodes. Neurology 2002;58(1):97–103 85. Johnston JM Jr, Mangano FT, Ojemann JG, Park TS, Trevathan E, Smyth MD. Complications of invasive subdural electrode monitoring at St. Louis Children’s Hospital, 1994–2005. J Neurosurg 2006;105(5, Suppl):343–347 86. Onal C, Otsubo H, Araki T, et al. Complications of invasive subdural grid monitoring in children with epilepsy. J Neurosurg 2003;98(5):1017–1026 87. Rydenhag B, Silander HC. Complications of epilepsy surgery after 654 procedures in Sweden, September 1990– 1995: a multicenter study based on the Swedish National ­Epilepsy Surgery Register. Neurosurgery 2001;49(1):51–56,­ discussion 56–57 88. Simon SL, Telfeian A, Duhaime AC. Complications of invasive monitoring used in intractable pediatric epilepsy. Pediatr Neurosurg 2003;38(1):47–52 89. Swartz BE, Rich JR, Dwan PS, et al. The safety and efficacy of chronically implanted subdural electrodes: a prospective study. Surg Neurol 1996;46(1):87–93 90. Yang PF, Zhang HJ, Pei JS, et al. Intracranial electroencephalography with subdural and/or depth electrodes in children with epilepsy: techniques, complications, and outcomes. Epilepsy Res 2014;108(9):1662–1670 91. Araki T, Otsubo H, Makino Y, et al. Efficacy of dexamathasone on cerebral swelling and seizures during subdural grid EEG recording in children. Epilepsia 2006;47(1):176–180 92. Stephan CL, Kepes JJ, SantaCruz K, Wilkinson SB, Fegley B, Osorio I. Spectrum of clinical and histopathologic responses to intracranial electrodes: from multifocal aseptic meningitis to

multifocal hypersensitivity-type meningovasculitis. Epilepsia 2001;42(7):895–901 93. Maciunas RJ. Computer-assisted neurosurgery. Clin Neurosurg 2006;53:267–271 94. Garcia PA, Barbaro NM, Laxer KD. The prognostic value of postoperative seizures following epilepsy surgery. Neurology 1991;41(9):1511–1512 95. Lüders H, Murphy D, Awad I, et al. Quantitative analysis of seizure frequency 1 week and 6, 12, and 24 months after surgery of epilepsy. Epilepsia 1994;35(6):1174–1178 96. Mani J, Gupta A, Mascha E, et al. Postoperative seizures after extratemporal resections and hemispherectomy in pediatric epilepsy. Neurology 2006;66(7):1038–1043 97. Perry MS, Donahue DJ, Malik SI, et al. Magnetic resonance imaging-guided laser interstitial thermal therapy as treatment for intractable insular epilepsy in children. J Neurosurg Pediatr 2017;20(6):575–582 98. Fujimoto A, Okanishi T, Kanai S, Sato K, Nishimura M, Enoki H. Neuronavigation-guided frameless stereoelectroencephalography (SEEG). Neurol Med Chir (Tokyo) 2017;57(9):496–502 99. Nowell M, Rodionov R, Diehl B, et al. A novel method for implementation of frameless StereoEEG in epilepsy surgery. Neurosurgery 2014;10(Suppl 4):525–533, discussion 533–534 100. Buckley R, Estronza-Ojeda S, Ojemann JG. Laser ablation in pediatric epilepsy. Neurosurg Clin N Am 2016;27(1):69–78 101.  Nowell M, Miserocchi A, McEvoy AW, Duncan JS. Advances in epilepsy surgery. J Neurol Neurosurg Psychiatry 2014;85(11):1273–1279 102. Prince E, Hakimian S, Ko AL, Ojemann JG, Kim MS, Miller JW. Laser interstitial thermal therapy for epilepsy. Curr Neurol Neurosci Rep 2017;17(9):63 103.  Wilfong AA, Curry DJ. Hypothalamic hamartomas: optimal approach to clinical evaluation and diagnosis. Epilepsia 2013;54(Suppl 9):109–114 104. Curry DJ, Gowda A, McNichols RJ, Wilfong AA. MR-guided stereotactic laser ablation of epileptogenic foci in children. Epilepsy Behav 2012;24(4):408–414 105. Hawasli AH, Bagade S, Shimony JS, Miller-Thomas M, Leuthardt EC. Magnetic resonance imaging-guided focused laser interstitial thermal therapy for intracranial lesions: single-institution series. Neurosurgery 2013;73(6):1007–1017 106. D’Argenzio L, Colonnelli MC, Harrison S, et al. Seizure outcome after extratemporal epilepsy surgery in childhood. Dev Med Child Neurol 2012;54(11):995–1000 107. Dorward IG, Titus JB, Limbrick DD, Johnston JM, Bertrand ME, Smyth MD. Extratemporal, nonlesional epilepsy in children: postsurgical clinical and neurocognitive outcomes. J Neurosurg Pediatr 2011;7(2):179–188 108. Englot DJ, Wang DD, Rolston JD, Shih TT, Chang EF. Rates and predictors of long-term seizure freedom after frontal lobe epilepsy surgery: a systematic review and meta-analysis. J Neurosurg 2012;116(5):1042–1048 109. Hauptman JS, Pedram K, Sison CA, et al. Pediatric epilepsy surgery: long-term 5-year seizure remission and medication use. Neurosurgery 2012;71(5):985–993 110. Tanriverdi T, Olivier NP, Olivier A. Quality of life after extratemporal epilepsy surgery: a prospective clinical study. Clin Neurol Neurosurg 2008;110(1):30–37 111. Tonini C, Beghi E, Berg AT, et al. Predictors of epilepsy surgery outcome: a meta-analysis. Epilepsy Res 2004;62(1):75–87 112. Beghi E, Tonini C. Surgery for epilepsy: assessing evidence from observational studies. Epilepsy Res 2006;70(2–3):97–102 113.  Gilliam F, Wyllie E, Kashden J, et al. Epilepsy surgery outcome: comprehensive assessment in children. Neurology 1997;48(5):1368–1374 114. Terra-Bustamante VC, Fernandes RM, Inuzuka LM, et al. Surgically amenable epilepsies in children and adolescents: clinical, imaging, electrophysiological, and post-surgical outcome data. Childs Nerv Syst 2005;21(7):546–551

46

  Supplementary Sensorimotor Area Surgery Jarod L. Roland and Matthew D. Smyth

Summary

„„ Anatomy

The supplementary sensorimotor area (SSMA) is a well-studied brain region that requires special considerations when undergoing resection for epilepsy in pediatric patients. A significant portion of our clinical understanding is extrapolated from adult patients and resections for brain tumors. In this chapter, we begin by reviewing the origins of the SSMA in the medical literature and its anatomical definitions. We then explore the function and the clinical course of deficits associated with resection in this area. This knowledge serves as the basis to interpret seizure semiology and outcomes reported in pediatric epilepsy surgery.

The anatomical definition of the supplementary motor area (SMA) was first described via electrocortical stimulation mapping by Penfield and Welch1 and confirmed by Woolsey and colleagues.2 Subsequent human studies additionally attributed sensory function to this area and, therefore, introduced the modified label of supplementary sensorimotor area (SSMA).3,​4 These terms, SMA and SSMA, are often used interchangeably in the literature and generally refer to the same anatomically defined location. For the remainder of this text, we will use SSMA to refer to this brain region for consistency. The SSMA makes up a portion of the posterior superior frontal gyrus and posterior mesial frontal lobe. The anatomic limits were originally defined by Penfield and colleagues via cortical stimulation studies in primates and humans. The precentral sulcus and cingulate sulcus are well-defined posterior and inferior borders, respectively.1,​5 The lateral and anterior limits are somewhat less well defined, but may include the superior frontal sulcus as the lateral border and up to 5 cm anterior to the precentral sulcus as the anterior border.3 Fig. 46.1 illustrates this region on an average cortical surface model. However, these limits may be subject to further subdivision into areas such as the SMA proper, pre-SMA, and supplementary eye field.6,​7,​8 Structural connections to and from the SSMA have been previously studied with great detail in nonhuman primate models.9 Such animal studies have demonstrated a significant contribution of the SSMA to motor systems. The corticospinal tract carries white matter fibers to the lower motor neurons of the spinal cord, of which approximately 10% arise from the SSMA.9 Furthermore, the SSMA is interconnected with many other regions of the brain. The interconnections provide even greater insight to the function of the SSMA and help explain the

Keywords:  supplementary sensorimotor area, supplementary motor area, SMA syndrome, epilepsy, epilepsy surgery, pediatric, neurosurgery

„„ Introduction Surgery involving the posterior mesial frontal lobe is associated with a syndrome of contralateral hemiparesis and speech deficits. Although deficits may be profound, they are frequently temporary. Rates of occurrence and predictive factors for postoperative deficits are variable in reported literature. However, given the high rate of recovery associated with this syndrome, surgery for medically refractory epilepsy in the mesial frontal lobe is generally well tolerated in pediatric patients. Here, we review the relevant literature regarding the anatomy and function of the supplementary sensorimotor area and its relevance to the neurosurgical treatment of medically refractory epilepsy.

Fig. 46.1  SSMA anatomic location. Lateral (a) and medial (b) renderings of an average brain model. FreeSurfer software was used to render the fsaverage left-hand pial surface model. The supplementary sensorimotor area is shaded in blue.

448

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches deficits that occur after resection. A human study by ­Vergani et al examined the white matter connections between the SSMA and other brain regions.10 They found five main connections both in postmortem dissections and in high-resolution diffusion tensor imaging (DTI) tractography from the Human Connectome Project: (1) U-fibers connecting the precentral gyrus; (2) U-fibers connecting the cingulate gyrus; (3) fibers connecting the pars opercularis; (4) fibers connecting the striatum; and (5) fibers connecting the contralateral SSMA via the corpus callosum.10 Of interest, the fibers connecting the pars opercularis, where Broca’s area is commonly located, may help explain the speech deficits observed in the postoperative SMA syndrome.

„„ Supplementary Sensorimotor Area Function Cortical stimulation and electrophysiologic recordings provide experimental evidence of function attributed to the SSMA. Electrical stimulation experiments demonstrate both induced complex motor and sensory behavior as well as suppression of function, namely speech. A study by Fried et al observed the results of electrocortical stimulation in human subjects undergoing invasive electrocorticography (ECoG) monitoring for epilepsy.11 They found stimulation at 169 electrode sites in the SSMA region of 13 patients produced motor responses most often (63%), followed by sensory perceptions (30%) and speech deficits (12%). The motor results were similar to those previously described by Penfield and Welch.1 However, Fried et al noted that subjective sensory experiences could be easily overlooked at stimulation sites with lower stimulation current thresholds, which would then evoke motor responses at higher stimulation current.11 Lim et al reported their results from studying a similar population of 19 human subjects undergoing invasive ECoG monitoring for epilepsy and provided further evidence for the dual role of sensory and motor function.3 These studies in humans as well as prior experiments in nonhuman primates also support a somatotopic organization to the SSMA.3,​11,​12 It is now well accepted that the SSMA is somatotopically organized with caudal areas representing lower extremities and more rostral areas representing orofacial and upper extremity functions. These experimental observations provide a firm basis to interpret clinical observations of the SMA syndrome resulting from cortical resections in the SSMA. One early series by Laplane et al described three patients suffering SMA syndrome after resection of the mesial frontal lobe. All three patients experienced temporary contralateral motor deficits (including lower extremity, upper extremity, and face) and speech arrest.13 A report by Zentner et al prospectively studied 28 patients in whom a frontodorsal resection was performed.14 They found 89% of their cohort suffered a temporary postoperative deficit of contralateral hemiparesis with or without concomitant speech arrest. Additionally, an effect of hemispheric dominance has also been reported. Speech function is more often found in the dominant hemisphere, while the nondominant hemisphere is more likely to evoke ipsilateral or bilateral motor response.11 The complete SMA syndrome including speech arrest is most often associated with the dominant frontal lobe; however, this is not without exception.

An excellent and detailed review of early studies of the SSMA can be found in Goldberg.15 An in-depth review by Nachev et al expands the view on subdivisions of the greater region referred to as the supplementary motor complex and includes the SSMA, pre-SMA, and supplementary eye field.6 The detailed anatomic and functional study of these areas is possible in animal models and well-designed studies of normal human subjects. However, delineation of these subdivisions is not common in clinical studies of patients with epilepsy.

„„ SMA Syndrome The SMA syndrome refers to the induced functional deficits of contralateral hemiparesis and aphasia that often result from resection of the SSMA. The syndrome was first described by Laplane and colleagues.13 There is a well-described postoperative evolution of symptoms, which is characterized by hemiparesis or motor apraxia and a range of associated speech arrest. These symptoms may be present immediately on recovery from anesthesia or may occur within hours following surgery.16 The course then proceeds to recovery that typically occurs over a time frame of days to months, but classically there is a rapid recovery to near baseline within about 1 week. Involvement of the primary motor area in the resection may cloud the time course of recovery and contribute to persistent postoperative deficits and prolong the recovery period. Kasasbeh and colleagues reviewed a series of 17 pediatric patients who developed SMA syndrome after resection for medically refractory epilepsy involving the SSMA.17 They found 82% had complete recovery at 1 month and 100% by 6 months. A larger study by Alonso-Vanegas et al included 52 adult patients who underwent surgery in the SSMA for epilepsy and found 50% experienced a postoperative SMA syndrome.18 All of their patients also had complete resolution of deficits by 6 months. These results from SSMA resection can be contrasted to those of a similar study by Pondal-Sordo et al that included 52 adult patients undergoing surgical intervention for epilepsy in the primary (opposed to supplementary) sensorimotor area. Last follow-up in their cohort was on average greater than 4 years at which time 50% of patients remained with neurological d ­ eficits.19 Not all studies report complete resolution in all patients. A retrospective review by Kim and colleagues in 2013 identified 43 adult and pediatric patients who underwent resection in the SSMA for epilepsy, of which 23 patients (53.5%) experienced new deficits.20 Of those 23 patients, 3 failed to completely recover after at least 2 years of follow-up. Several groups have reported clinical factors in their case series that are associated with postoperative deficits. Kim et al noted resection of the SSMA and cingulate gyrus to be more predictive of transient deficits compared to resection of the SSMA alone.20 Kasasbeh et al noted a similar relationship in their case series when they measured the distance from the resection site to the cingulate gyrus and to the precentral gyrus, and found statistically significant correlations for both.17 They also identified cases with lesions identified on preoperative imaging to be less likely to develop an SMA syndrome. These findings are similar to those reported in an earlier series of 11 patients undergoing glioma resection, where location and extent of resection were reported to correlate with transient postoperative deficits.21

46  Supplementary Sensorimotor Area Surgery

„„ Functional Mapping The SSMA is located immediately adjacent to the precentral gyrus. The precentral sulcus is the anatomic border separating the SSMA from the primary motor strip. This proximity often leads to the use of adjuvant mapping techniques when surgical resection in the SSMA is indicated. Intraoperative mapping via ECoG or direct stimulation is frequently reported in case series involving SSMA resection.17,​20,​21,​22,​23 Stimulation mapping can be performed during awake surgery or under general anesthesia with monitoring for induced motor activity. Observation of gross movements in the extremities can be performed simply by the surgeon and anesthesiologist or with the aid of electromyography (EMG) recording. Detailed functional mapping helps the surgeon limit the resection to the SSMA, where deficits are expected to be temporary, and avoid the primary motor area, where outcomes are more often permanent.19 Somatosensory evoked potentials can also be used to easily and reliably identify the central sulcus to aid in localizing the primary motor strip. Stimulation is applied peripherally to the median nerve while observing the ECoG for a distance phase reversal at N20 between adjacent electrodes. This phase reversal occurs between electrodes spanning the central sulcus, thereby helping to localize relevant anatomy on the exposed cortical surface. Invasive mapping is the traditional method to which other techniques are often compared. However, less invasive and extraoperative techniques often make a significant contribution to surgical planning and may be sufficient in certain cases.24 Prior to entering the operating room, noninvasive functional imaging can be used to identify the primary motor strip and possibly the SSMA as well. Functional MRI (fMRI) is commonly available at academic neurosurgical institutions. It has the advantage of providing localizing information in MRI space that is familiar to neurosurgeons. Additionally, the functional data can often be

incorporated with neuronavigation for intraoperative guid­ ance. Task-based fMRI (t-fMRI) is typically performed in an MRI scanner by prompting the patient to perform motor tasks, such as finger-tapping or protruding the tongue, in a block design pattern alternating with rest epochs. The choice of task determines the area of the motor homunculus being activated. A task paradigm where foot movement is used may identify the mesial posterior frontal lobe corresponding to the lower extremity area of the motor homunculus. Similarly, hand or tongue tasks will localize the lateral and ventral portions of motor strip, respectively. These task-based paradigms can be used to localize the primary as well as SMAs.25 In 2017, Collinge et al performed a systematic review of the literature on using functional imaging for presurgical mapping in ­pediatric ­epilepsy patients.26 They conclude that there is strong preliminary evidence to support t-fMRI to localize motor function in pediatric patients. In pediatric patients, compliance with task paradigms may be challenging. This is particularly true in those children who are very young, developmentally delayed, or suffering acute cognitive deficits secondary to their pathology. In such cases, resting-state fMRI (r-fMRI) may be of particular value to pediatric patients.27 This is because r-fMRI has the unique advantage of not requiring patient participation in order to localize functional cortical areas as part of a resting state network (RSN). The initial methodological description of r-fMRI by Biswal and colleagues used the primary motor strip as an exemplar area for localization in the absence of task performance.28 Since that time, the field of resting-state imaging has evolved and RSNs, such as the sensorimotor network (SMN), have become well defined. The SMN includes the precentral gyrus, postcentral gyrus, and portions of the SSMA.25,​29,​30 Fig. 46.2 ­illustrates an exemplar pediatric patient that underwent r-fMRI for ­preoperative mapping prior to resection of the right frontal lobe, including the SSMA. More recently, additional methodological techniques have been developed for specifically i­ dentifying the SSMA proper and pre-SSMA in r-fMRI data.31

Fig. 46.2  Exemplar supplementary sensorimotor area (SSMA) resection for epilepsy in a pediatric patient. A 14-year-old female with medically refractory epilepsy originating in the right frontal lobe. (a) Resting-state fMRI was used preoperatively to map the primary sensorimotor and SSMA, outlined in red. Diffusion tensor imaging tractography was also used to identify the right and left corticospinal tracts, outlined in green and blue, respectively. (b) Intraoperative photograph demonstrates a wide resection of the right frontal lobe including the SSMA. (c) Postoperative imaging confirmed area of resection including the mesial frontal lobe structures of the SSMA. She experienced a typical supplementary motor area (SMA) syndrome involving hemiplegia and speech arrest immediately postoperatively that began recovering in 1 week and completely resolved by 1 month.

449

450

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

„„ Seizures Arising from the SSMA Seizures arising from the SSMA may be difficult to localize owing to unusual symptoms such as bilateral limb involvement and preserved consciousness and difficult to interpret electroencephalography (EEG) findings.7,​32 However, identifying a stereotypical semiology may be helpful in a comprehensive clinical evaluation. In 1995, Connolly et al detailed the characteristics of a small cohort of pediatric patients with SSMA seizures. They defined inclusion criteria as seizures involving bilateral tonic posturing of upper and lower extremities with preserved consciousness and lack of p ­ ostictal confusion. In this group, they found speech arrest occurred in eight or nine patients old enough to adequately evaluate speech.33 Unnwongse and colleagues devote a significant portion of their review on mesial frontal lobe epilepsy to the SSMA.7 They describe seizures originating in the SSMA as characterized by asymmetric posturing of one or more extremities that commonly affects both sides of the body. This posturing may include the fencing or M2e posture that involved extension of the contralateral arm and flexion of the ipsilateral arm, and figure-of-four posture where the contralateral arm is extended across the chest and the ipsilateral arm is flexed forming the shape of the digit “4.” Various patterns of vocalization are also commonly reported in seizures arising from the SSMA.7 The extant ­literature reports

a wide variation in semiology with significant overlap from seizures spreading secondarily into the SSMA, which makes localization by clinical observation challenging.

„„ Seizure Outcomes following Surgical Resection Reported seizure outcomes for surgical resection in the SSMA tend to fair well on the Engel classification scale. In the series reported by Kasasbeh et al, 37 (84%) of their pediatric patients achieved an Engel Class I or II outcome at 12-months ­follow-up.17 Alonso-Vanegas et al reviewed their series of 52 adult patients and similarly reported good outcomes, with 92% achieving Engel Class I or II outcomes.18 A series of pediatric and adult patients reported by von Lehe and colleagues in 2012 suggest a possible correlation between seizure freedom and extent of resection in cingulate and mesial frontal cortex.34 They found 43% achieved Engel Class I after a resection limited to the cingulate gyrus compared to 71% with extended resection including the mesial frontal cortex. They noted that postoperative deficits occurred only in those who had resections extending in the SSMA. These data, along with the well-described transient course and high rate of complete SMA syndrome resolution, argue for greater use of extended resections in the cingulate and mesial frontal lobe.

References 1. Penfield W, Welch K. The supplementary motor area of the cerebral cortex; a clinical and experimental study. AMA Arch Neurol Psychiatry 1951;66(3):289–317

12. Mitz AR, Wise SP. The somatotopic organization of the supplementary motor area: intracortical microstimulation mapping. J Neurosci 1987;7(4):1010–1021

2. Woolsey CN, Settlage PH, Meyer DR, Sencer W, Pinto Hamuy T, ­Travis AM. Patterns of localization in precentral and “supplementary” motor areas and their relation to the concept of a premotor area. Res Publ Assoc Res Nerv Ment Dis 1952;30: 238–264

13. Laplane D, Talairach J, Meininger V, Bancaud J, Orgogozo JM. Clinical consequences of corticectomies involving the supplementary motor area in man. J Neurol Sci 1977;34(3):301–314

3. Lim SH, Dinner DS, Pillay PK, et al. Functional anatomy of the human supplementary sensorimotor area: results of extraoperative electrical stimulation. Electroencephalogr Clin Neurophysiol 1994;91(3):179–193 4. Lüders HO. The supplementary sensorimotor area. An overview. Adv Neurol 1996;70:1–16 5. Laich E, Kuzniecky R, Mountz J, et al. Supplementary sensorimotor area epilepsy. Seizure localization, cortical propagation and subcortical activation pathways using ictal SPECT. Brain 1997;120(Pt 5):855–864 6. Nachev P, Kennard C, Husain M. Functional role of the supplementary and pre-supplementary motor areas. Nat Rev Neurosci 2008;9(11):856–869

14. Zentner J, Hufnagel A, Pechstein U, Wolf HK, Schramm J. Functional results after resective procedures involving the supplementary motor area. J Neurosurg 1996;85(4):542–549 15. Goldberg G. Supplementary motor area structure and function: review and hypotheses. Behav Brain Sci 1985;8(4):567–588 16. Duffau H, Lopes M, Denvil D, Capelle L. Delayed onset of the supplementary motor area syndrome after surgical resection of the mesial frontal lobe: a time course study using intraoperative mapping in an awake patient. Stereotact Funct Neurosurg 2001;76(2):74–82 17. Kasasbeh AS, Yarbrough CK, Limbrick DD, et al. Characterization of the supplementary motor area syndrome and seizure outcome after medial frontal lobe resections in pediatric epilepsy surgery. Neurosurgery 2012;70(5):1152–1168, discussion 1168

7. Unnwongse K, Wehner T, Foldvary-Schaefer N. Mesial frontal lobe epilepsy. J Clin Neurophysiol 2012;29(5):371–378

18. Alonso-Vanegas MA, San-Juan D, Buentello García RM, et al. Long-term surgical results of supplementary motor area epilepsy surgery. J Neurosurg 2017;127(5):1153–1159

8. Dale AM, Fischl B, Sereno MI. Cortical surface-based analysis. I. Segmentation and surface reconstruction. Neuroimage 1999;9(2):179–194

19. Pondal-Sordo M, Diosy D, Téllez-Zenteno JF, Girvin JP, Wiebe S. Epilepsy surgery involving the sensory-motor cortex. Brain 2006;129(Pt 12):3307–3314

9. Potgieser AR, de Jong BM, Wagemakers M, Hoving EW, Groen RJ. Insights from the supplementary motor area syndrome in balancing movement initiation and inhibition. Front Hum Neurosci 2014;8:960

20. Kim Y-H, Kim CH, Kim JS, et al. Risk factor analysis of the development of new neurological deficits following supplementary motor area resection. J Neurosurg 2013;119(1):7–14

10. Vergani F, Lacerda L, Martino J, et al. White matter connections of the supplementary motor area in humans. J Neurol Neurosurg Psychiatry 2014;85(12):1377–1385

21. Fontaine D, Capelle L, Duffau H. Somatotopy of the supplementary motor area: evidence from correlation of the extent of surgical resection with the clinical patterns of deficit. Neurosurgery 2002;50(2):297–303, discussion 303–305

11. Fried I, Katz A, McCarthy G, et al. Functional organization of human supplementary motor cortex studied by electrical stimulation. J Neurosci 1991;11(11):3656–3666

22. Yamane F, Muragaki Y, Maruyama T, et al. Preoperative mapping for patients with supplementary motor area epilepsy: multimodality brain mapping. Psychiatry Clin Neurosci 2004;58(3):S16–S21

46  Supplementary Sensorimotor Area Surgery 23. Ibe Y, Tosaka M, Horiguchi K, et al. Resection extent of the supplementary motor area and post-operative neurological deficits in glioma surgery. Br J Neurosurg 2016;30(3):323–329

29. Xiong J, Parsons LM, Gao JH, Fox PT. Interregional connectivity to primary motor cortex revealed using MRI resting state images. Hum Brain Mapp 1999;8(2–3):151–156

24. Liégeois F, Cross JH, Gadian DG, Connelly A. Role of fMRI in the decision-making process: epilepsy surgery for children. J Magn Reson Imaging 2006;23(6):933–940

30. Ma L, Wang B, Chen X, Xiong J. Detecting functional connectivity in the resting brain: a comparison between ICA and CCA. Magn Reson Imaging 2007;25(1):47–56

25. Hiroshima S, Anei R, Murakami N, Kamada K. Functional localization of the supplementary motor area. Neurol Med Chir (Tokyo) 2014;54(7):511–520

31. Kim J-H, Lee J-M, Jo HJ, et al. Defining functional SMA and pre-SMA subregions in human MFC using resting state fMRI: functional connectivity-based parcellation method. Neuroimage 2010;49(3):2375–2386

26. Collinge S, Prendergast G, Mayers ST, et al. Pre-surgical mapping of eloquent cortex for paediatric epilepsy surgery candidates: ­evidence from a review of advanced functional neuroimaging. Seizure 2017;52:136–146 27. Roland JL, Griffin N, Hacker CD, et al. Resting-state functional magnetic resonance imaging for surgical planning in pediatric patients: a preliminary experience. J Neurosurg Pediatr 2017;20(6):583–590 28. Biswal B, Yetkin FZ, Haughton VM, Hyde JS. Functional connectivity in the motor cortex of resting human brain using echo-planar MRI. Magn Reson Med 1995;34(4):537–541

32. Bass N, Wyllie E, Comair Y, Kotagal P, Ruggieri P, Holthausen H. Supplementary sensorimotor area seizures in children and adolescents. J Pediatr 1995;126(4):537–544 33. Connolly MB, Langill L, Wong PK, Farrell K. Seizures involving the supplementary sensorimotor area in children: a video-EEG analysis. Epilepsia 1995;36(10):1025–1032 34. von Lehe M, Wagner J, Wellmer J, Clusmann H, Kral T. Epilepsy surgery of the cingulate gyrus and the frontomesial cortex. ­Neurosurgery 2012;70(4):900–910, discussion 910

451

47

  Rolandic Cortex Surgery Ibrahim Jalloh and James T. Rutka

Summary Rolandic epilepsy is defined by the presence of seizures ­arising from the precentral and/or postcentral gyrus with or without involvement of adjacent cortex. Successful resection of the seizure-generating epileptogenic zone produces good out­ comes, curing seizures in most patients. Sensory and motor deficits produced by rolandic surgery are often transient, and when permanent, they are usually well-tolerated. Accurate functional localization of primary motor and somatosensory cortex is particularly important in rolandic surgery. Most patients will therefore require invasive monitoring to precisely define the epileptogenic zone and to accurately map the somatotopic organization of the primary motor cortex. In this chapter, we review the clinical, imaging, and neurophysiological information, as well as the surgical principles, required to precisely localize and safely resect the epileptogenic zone in patients with rolandic epilepsy. Keywords:  rolandic epilepsy, epilepsy surgery, intracranial EEG, multiple subpial transections, corticectomy, c­ omplications

„„ Introduction Rolandic epilepsy is defined by the presence of seizures arising from the precentral and/or postcentral gyrus with or without involvement of adjacent cortex. It can be the result of focal lesions such as tumors and vascular malformations located in the rolandic cortex. In such cases, the surgical management is often dictated by management of the underlying lesion rather than by the epilepsy. Many patients with refractory rolandic epilepsy do not have a clear focal lesion responsible for seizure generation. The presurgical workup of these patients is critical to precisely define the epileptogenic zone. A period of intracranial electroencephalography (EEG) monitoring is usually essential. Successful resection of the seizure-generating epileptogenic zone produces good outcomes, curing seizures in most patients and mitigating the deleterious cognitive and psychosocial consequences of refractory epilepsy. Integral to patient outcome is the risk of permanent neurological deficit. Sensory and motor deficits produced by rolandic surgery are often transient, and when permanent, they are usually well-tolerated. However, loss of function does occur in some patients and this must be predicted as best as possible to allow appropriate counseling of

patients and their families. Consequently, accurate functional localization of primary motor and somatosensory cortex is particularly important in rolandic surgery. In this chapter, we review the clinical, imaging, and neurophysiological information, as well as the surgical principles, required to precisely localize and safely resect the epileptogenic zone in patients with rolandic epilepsy.

„„ Historical Background of Rolandic Surgery Surgery for epilepsy in the rolandic cortex is not new. Some of the earliest neurosurgical procedures of the modern era were performed in this region reflecting the reliance on clinical findings to localize intracerebral lesions rather than modern imaging techniques. In the 1880s, Victor Horsley was the first to describe the subpial resection of the rolandic cortex to treat seizures.1 His surgeries predate the use of antiepileptic medications, and many of Horsley’s patients were truly disabled by seizures, including one patient who suffered close to 3,000 seizures during his hospital stay.1 Horsley inspired larger and more systematic surgical series by others including Ernest Sachs, who spent a period working with Horsley, Leonard Furlow, and Cobb Pilcher. Their publications in the 1930s and 1940s described the subsequent development of rolandic surgery.2,​3,​4 In particular, they described the use of cortical stimulation for anatomical and epileptogenic zone localization and provided a better understanding of seizure and neurological outcomes. Importantly, they recognized that patients can have recurrence of seizures even after several years of apparent cure, and that limb weakness and function recovers at least in part in most patients after rolandic surgery.

„„ Clinical and Pathological Features of Rolandic Epilepsy: Idiopathic and Lesional Idiopathic epilepsy syndromes arising from the rolandic cortex are frequently benign in nature. It is important to distinguish benign rolandic epilepsy from the more malignant rolandic epilepsy syndromes suffered by those patients selected for surgery.

47  Rolandic Cortex Surgery Benign rolandic epilepsy, also termed benign partial epilepsy of childhood with centrotemporal spikes (BECCTS), accounts for approximately 15% of childhood epilepsies and has a good prognosis.5 There is no role for surgery. Typically, seizures start between the ages of 4 and 10 years and consist of simple partial seizures involving the face and tongue that occur during sleep with occasional secondary generalization. They are not associated with neurological or cognitive deficits and are easily controlled with medications. Seizures usually terminate during adolescence irrespective of treatment. This contrasts with a more malignant rolandic epilepsy syndrome in which sensorimotor seizures readily progress to generalized seizures and are refractory to antiepileptic drugs.6 These patients frequently have associated cognitive deficits. Both benign rolandic epilepsy and its more malignant counterpart are typically associated with normal structural imaging studies, and the ictal EEG morphology, observed in the centrotemporal region, may also not help to distinguish these entities.6,​7 Many patients undergoing surgery for rolandic epilepsy do not have an obvious structural lesion on magnetic resonance imaging (MRI), although microstructural abnormalities are frequently revealed on histopathological analysis of resected cortical tissue. These microstructural abnormalities include cortical dysplasia, cortical astrogliosis, microdysgenesis, and diffuse

astrocytic inclusions.8,​9,​10 With the increasing resolution of more powerful MRI scanners, the subtle imaging findings that characterize these lesions are increasingly recognized. The etiology of lesional rolandic epilepsy includes a diverse list of pathologies that can be classified as neoplastic, most commonly dysembryoplastic neuroepithelial tumor, developmental anomalies, vascular lesions, and other pathologies (Table 47.1; Fig. 47.1).8,​9,​10,​11,​12

Table 47.1  Etiology of lesional and nonlesional rolandic epilepsy

Lesional Neoplastic

DNET, ganglioglioma, glioma

Developmental anomalies

Cortical dysplasia, tubers

Vascular lesions

AVM, cavernoma

Other pathologies

Gliosis, infarct, Sturge–Weber ­syndrome

Nonlesional

Diffuse astrocytic inclusions, ­microdysgenesis Histiocytic ­infiltrates

Abbreviations: AVM , arteriovenous malformation; DNET, dysembryoplastic neuroepithelial tumor.

Fig. 47.1  The neuroradiological studies of lesional rolandic epilepsy. (a) Axial T2-weighted MR image showing a parasagittal lesion of mixed signal in the postcentral gyrus. The lesion was diagnosed as a cavernous malformation on MRI, which was confirmed at the final pathological examination. (b) Axial T1-weighted MR image showing abnormal focal thickened gray matter within the left posterior frontal cortex (denoted by bracket and asterisk) confirmed on pathological examination to be cortical dysplasia. (c) Axial T1 gadolinium-enhanced MR image showing a well-defined heterogeneously enhancing lesion in the right parietal lobe that anteriorly involves the right postcentral gyrus confirmed on pathological examination to be a dysembryoplastic neuroepithelial tumor (DNET). (d) Coronal T2/fluid-attenuated inversion recovery (FLAIR) MR image showing linear hyperintensity in the subcortical and deep white matter that tapers toward the ipsilateral right ventricle confirmed on pathological examination to be focal cortical dysplasia.

453

454

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 47.2  Management paradigm for rolandic epilepsy patients. As shown here, phase 1 consists of a detailed clinical assessment, video EEG monitoring, and structural imaging. This information is presented to the epilepsy surgery team who then decide on whether epilepsy surgery is an option. Phase 2 consists of neuropsychological testing and functional imaging. If there is a focal rolandic lesion with concordant clinical, EEG, and radiological data, then a period of extraoperative EEG monitoring (phase 3) may not be necessary, but for most rolandic surgery patients a period of extraoperative EEG monitoring and functional mapping is useful to precisely define the epileptogenic zone and functional cortex.

„„ Evaluation of Patients for Surgery Patients undergoing rolandic epilepsy surgery constitute a mix of lesional and nonlesional cases. The management paradigm may differ depending on whether the aim is to simply confirm a lesion as the source of ictal spread or whether the aim is to define the epileptogenic zone in the absence of a clear focal lesion. A vital part of the presurgical evaluation is to determine the anatomic localization of primary motor and somatosensory cortex (Fig. 47.2).

Presurgical Evaluation Patients undergo a careful clinical evaluation, prolonged scalp video-EEG monitoring, and structural MRI imaging. Video-EEG permits a detailed evaluation of seizure semiology, which is essential to help guided resection. Scalp EEG will at best localize an epileptogenic focus to a relatively wide region of cortex. Following this first phase of evaluation, the epilepsy team decides

what kind of seizures are present, whether the seizures are medically refractory, whether they are lateralized and/or localized on EEG, and whether surgery is a consideration. A second phase of investigations consists of neuropsychological assessment and functional radiographic studies. We routinely use magnetoencephalography (MEG) coregistered with structural MRI which permits more precise seizure foci localization than scalp EEG, to within a few millimeters, although it is most often restricted to interictal recordings. ­Additionally, MEG can be used for functional mapping (see below). Our methods of detection, localization, and analysis of MEG spikes have been previously published.13,​14 The use of MEG to localize seizure foci has been evaluated alongside invasive monitoring and demonstrated to be both accurate and reliable for localization of the seizure foci.15 Ancillary investigations used to help defining the epileptogenic zone include fludeoxyglucose-positron emission tomography (FDG-PET) and single-photon emission computed tomography (SPECT). These are used selectively where foci localization is challenging, for instance, due to the rapidity and

47  Rolandic Cortex Surgery

Fig. 47.3  Subdural grid with epileptogenic zone and functional mapping data. A subdural grid has been placed after a generous craniotomy, exposing much of the lateral hemisphere. The grid has been supplemented with frontal and temporal strip electrodes. Temporal depth electrodes have also been placed. The epileptogenic zones for different seizure types are traced in yellow, green, blue, and orange. The annotated letters, circle, and squares indicate functional mapping data from MEG (A = MEG motor, B = MEG somatosensory evoked fields), extraoperative somatosensory evoked potentials (green square), and extraoperative cortical stimulation (yellow squares = hand, yellow circle = foot).

extent of ictal spread. Our experience is that in these complex extratemporal epilepsy cases, SPECT and/or PET studies are not that helpful in accurately predicting the epileptogenic zone.8,​11

Locating the Epileptogenic Zone: The Case for Extraoperative Seizure Mapping In patients with a focal lesion and concordant clinical, EEG, and radiological data, a period of invasive EEG monitoring may not be absolutely necessary. However, in the context of rolandic epilepsy where the epileptogenic zone resides wholly or partly within eloquent cortex, a period of invasive EEG monitoring is usually recommended to ensure maximum precision in defining the epileptogenic zone. Invasive monitoring is also required for accurate speech and language functional localization in children where awake surgery may not be an option (see below).

„„ Subdural Grid Placement A subdural electrode array (grid) with up to 128 contacts is used for extraoperative EEG monitoring. Subdural grids are tailor-made for each patient (Ad-Tech, Racine, WI) with cortical coverage based on the semiology of the seizures, scalp EEG recordings, and MEG clusters. This is supplemented with strip and depth electrodes, as determined by seizure semiology and MEG clusters, to ensure coverage of potential epileptogenic zones not covered by the subdural grid (Fig. 47.3). To insert the grid, we perform a generous craniotomy with a large M-shaped dural opening based on the sagittal sinus. The central sulcus is identified visually, with the aid of neuronavigation, and confirmed using a strip electrode to determine the inversion polarity of a somatosensory evoked potential (SSEP) triggered by electrical stimulation of the median nerve (SSEP phase reversal method) (Fig. 47.4).16 The location of the primary motor gyrus is confirmed by direct cortical stimulation of hand area while monitoring evoked electromyography (EMG) ­responses.

Stimulation is applied either using bipolar biphasic pulses (50 Hz) or by direct monopolar stimulation using “trains of five” short-train high-frequency stimulation, which is the authors’ preferred method (Fig. 47.4). Ice-cold saline is on hand to terminate any seizures precipitated by direct cortical stimulation, although short-train high-frequency stimulation is associated with a low risk of triggering seizures.17 The edges of the grid are secured to the dura at its edges. Photographs are taken of the final operative field and then the dura is closed with dural grafts forming an expansile duroplasty. We hinge the bone flap at its superior edge. The cables and a subgaleal drain are tunneled through the scalp flap and secured with purse-string sutures. We routinely use prophylactic broad-spectrum antibiotics during the postoperative period while the subdural grid is in situ. Postoperative CT or MRI is undertaken on the first postoperative day to review the electrode positioning and to exclude postoperative swelling and/or hematoma. Antiepileptic drugs are usually weaned prior to grid insertion. During the period of extraoperative monitoring, seizures are captured and a seizure map created corresponding to the different seizure types.

„„ Localization of Primary Motor and Somatosensory Cortex Rolandic surgery requires a detailed understanding of the location and somatotopic organization of the primary motor and somatosensory cortex so as to evaluate and minimize the risk of neurological deficits. This is particularly important in cases where normal cortical anatomy has been distorted. Moreover, rolandic epilepsy patients may have reorganization of their motor homonculus.17 Functional localization can be achieved through a number of methods including presurgical functional MRI (fMRI; with or without diffusion tensor imaging for tracing

455

456

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 47.4  Methods for intraoperative localization and monitoring of primary motor cortex during rolandic surgery. (a) Somatosensory evoked potential (SSEP) phase reversal method. Fourcontact strip electrode placed across central sulcus showing the inversion of polarity of a SSEP measured from the sensory cortex (contacts 3 and 4) and from the motor cortex (contacts 1 and 2). The letters “A” and “B” indicate somatosensory and motor cortex, respectively, as identified using MEG preoperatively. (b) Train-of-five continuous monitoring. Four-contact strip electrodes have been placed on primary and premotor cortical areas. This allows continuous monitoring of the motor cortex and corticospinal tract during resective surgery.

corticospinal tracts), MEG, direct recording of cortical sensory evoked potentials, direct cortical stimulation in the awake patient, direct cortical stimulation in the asleep patient with monitoring of evoked EMG responses (as described above), and by a combination of these methods. Both fMRI and MEG can be used to localize the somatosensory cortex by monitoring the change in blood-oxygen-level-­ dependent signal or evoked magnetic field, with fMRI or MEG, respectively, in response to tactile or electrical stimulation.18,​19 Reliably localizing the primary motor cortex can, however, be challenging as it requires the patient to perform repeated voluntary movements. Both fMRI and MEG have important methodological limitations and interpretation of data is subject to various caveats. For patients with rolandic epilepsy, fMRI and MEG cannot be relied on solely to precisely localize functional cortex. Direct cortical stimulation in the awake patient, either intraor extraoperatively, is the gold standard for functional localization. The authors perform extraoperative cortical mapping via the implanted subdural grid usually during the third or fourth day after grid insertion. We perform mapping of motor, sensory, and language functions using trains of 50-Hz biphasic pulses for up to 25 seconds starting with an intensity of 2 mA working up to a maximum of 20 mA in increments of 1 to 2 mA. Short-acting benzodiazepines are used in the case of triggered seizures. The primary somatosensory cortex is mapped by evoked p ­ otentials using the grid electrodes.

„„ Resective Surgery The anatomical information obtained from analysis of seizure semiology, structural MRI, MEG, and EEG and from functional localization is used to create a map that outlines the epileptogenic zone and the primary motor and somatosensory cortex (Fig. 47.3). This is used to plan surgical resection and also importantly to inform the patient and their family of the expected outcomes in terms of seizure control and risk of neurological deficit. Photographs taken at the time of grid insertion and annotated with the planned resection margin and eloquent cortex are used to help identify the resection margin at the time of surgery. Sections of the grid outside the resection margin are cut away, leaving parts of the grid corresponding to the cortex outlined for resection. The location of the somatosensory cortex is reconfirmed using the SSEP phase reversal method. Neuronavigation and ultrasound are used when there is a lesion. Subpial resection of cortex is carried out deep enough, down to white matter, to ensure complete corticectomy while care is taken to preserve surface veins and passing arteries. Multiple subpial transections (MSTs), as described by ­Morrell and others, are an alternative to complete corticectomy and have been used in eloquent cortex with the aim of arresting ictal propagation while sparing neurological deficit.20 However, our experience is that although MST often produces an early response in terms of seizure control, the results are not durable

47  Rolandic Cortex Surgery over time.21 Consequently, our use of MST has become limited in recent years. Electrocorticography (ECoG) can be used at the time of resective surgery to map interictal epileptiform discharges in the vicinity of a lesion, defining an “irritative zone.” Resecting this “irritative zone” together with the lesion has been associated with a better seizure outcome than just resecting the lesion alone.15,​22 However, the utility of using ECoG to guide cortical resections in this manner is debated. Surgical irritation of the cortex may produce reactive epileptiform discharges leading to a false assessment of the epileptogenic zone.23 Hence, this strategy must be used with caution.

„„ Surgical Complications Recognized complications from subdural grid insertion include cerebrospinal fluid (CSF) leak, cerebral edema, and intracranial hemorrhage.8,​24 We minimize the risk of CSF leak by securing cables with purse-string sutures and by using a subgaleal drain. Subdural hemorrhage identified on post-grid insertion imaging occurs in approximately 15% of patients; however, very few of these patients are symptomatic or require additional surgery to evacuate the hematoma.24 Similarly, cerebral edema is identified on post-grid insertion imaging in approximately 15% of patients but this is rarely symptomatic.24 We minimize the risk of clinically significant cerebral edema by using large duroplasties and hinging the bone flap at its superior margin. We also take care to ensure that the edges of large grids do not compress and impede the outflow of major draining veins as they enter the dural venous sinuses. Complications seen after second-stage surgery for cortical excision are most commonly due to infection and wound healing problems. Wound infection, meningitis, osteomyelitis, and epidural abscess formation are all reported in several invasive monitoring series with an overall infection rate of 4 to 12%.8,​10,​11,​24 Prolonged wound healing and scar hypertrophy can also occur. Many of our patients require a blood transfusion during their inpatient stay, reflecting the large size of the craniotomies that are typically required for subdural grid insertion.

„„ Neurological Outcomes Neurological deficits after rolandic surgery are predictable as is functional recovery. Resection of the inferior rolandic region, inferior to the thumb area, is not usually associated with long-lasting deficits. Speech dysfunction occurs in some patients, most likely due to a neuropraxia associated with impairment of unilateral face and tongue motor pathways. These impairments tend to be mild and improve in all patients.8–​10,​12 Resection of hand and limb primary motor cortex in the superior rolandic region produces an immediate flaccid hemiparesis (or a worsening of an existing weakness). Recovery begins in the proximal muscles and is largely complete by 6 to 12 weeks, although more subtle improvements will be observable

even after several years.8,​9,​10,​11,​12 The majority of patients will be ambulatory without assistance albeit with increased tone. Approximately 25% of patients will exhibit some long-lasting impairment of hand function and this will leave a completely nonfunctioning hand in 10%.8,​11 Resection of primary somatosensory cortex is usually well-tolerated, although sometimes, particularly in children, a profound sensory neglect will ensue that resembles a primary motor deficit. This almost always fully recovers. Other sensory deficits such as agraphesthesia, lack of two-point discrimination, and impaired proprioception are sometimes demonstrable but not usually symptomatic.8

„„ Seizure Outcomes The literature reports good seizure outcomes in rolandic surgery patients, with most series including ours reporting worthwhile improvements (Engel Class I–III) of the order of 70%.8–12,​25 Engel Class I outcomes are achieved in 31 to 63% of patients. Factors that appear to influence seizure outcomes, albeit judged from univariate analyses of small numbers, include age of the patient and the postsurgical EEG. Older children fare better than younger children and those with persistent rolandic interictal EEG epileptiform activity are associated with worse seizure ­outcomes.8,​12 An Engel Class I outcome is arguably not the most appropriate way to judge the benefits of rolandic surgery. Many of the patients selected for surgery are considerably disabled by their epilepsy such that even a reduction in seizures, as opposed to complete seizure freedom, is a worthwhile aim. Moreover, the impact of postresection neurological deficits must be evaluated when considering the utility of rolandic surgery.

„„ Conclusion The choice to sacrifice eloquent brain for a chance of seizure freedom is a difficult decision for patients, their family, and the epilepsy team. Rolandic epilepsy surgery requires a detailed understanding of individual patients’ seizures and the likelihood of achieving a worthwhile improvement in their epilepsy with surgical resection. Furthermore, the risk of permanent neurological deficit must be carefully evaluated. Consequently, most patients will require invasive monitoring to precisely define the epileptogenic zone and to accurately map the somatotopic organization of the primary motor cortex. The surgical management of rolandic epilepsy can be carried out safely when undertaken in a specialist epilepsy center that has evolved its expertise in managing complex extratemporal lobe epilepsy patients. Good outcomes are achieved in our patients with acceptable risks. Surgery for rolandic epilepsy is a valuable treatment offsetting the sometimes catastrophic cognitive and psychosocial consequences of medically refractory epilepsy.

457

458

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

References 1. Horsley V. Remarks on ten consecutive cases of operations upon the brain and cranial cavity to illustrate the details and safety of the method employed. BMJ 1887;1(1373):863–865

14. Otsubo H, Ochi A, Elliott I, et al. MEG predicts epileptic zone in lesional extrahippocampal epilepsy: 12 pediatric surgery cases. Epilepsia 2001;42(12):1523–1530

2. Sachs E. The subpial resection of the cortex in the treatment of jacksonian epilepsy (Horsley operation) with observations on areas 4 and 6. Brain 1935;58:492–503

15. Minassian BA, Otsubo H, Weiss S, Elliott I, Rutka JT, Snead OC III. Magnetoencephalographic localization in pediatric epilepsy surgery: comparison with invasive intracranial electroencephalography. Ann Neurol 1999;46(4):627–633

3. Furlow LT. Subpial resection of the cortex for focal epilepsy: further observations. J Am Med Assoc 1938;111(23):2092–2095 4. Pilcher C, Meacham WF, Holbrook TJ. Partial excision of the motor cortex in treatment of jacksonian convulsions; results in 41 cases. Arch Surg 1947;54(6):633–643 5. Camfield P, Camfield C. Epileptic syndromes in childhood: clinical features, outcomes, and treatment. Epilepsia 2002;43(June, Suppl 3):27–32 6. Otsubo H, Chitoku S, Ochi A, et al. Malignant rolandic-sylvian epilepsy in children: diagnosis, treatment, and outcomes. Neurology 2001;57(4):590–596 7. Ong HT, Wyllie E. Benign childhood epilepsy with centrotemporal spikes: is it always benign? Neurology 2000;54(5):1182–1185 8. Benifla M, Sala F Jr, Jane J, et al. Neurosurgical management of intractable rolandic epilepsy in children: role of resection in ­ eloquent cortex. Clinical article. J Neurosurg Pediatr 2009;4 (3):199–216 9. Sarkis RA, Jehi LE, Bingaman WE, Najm IM. Surgical outcome following resection of rolandic focal cortical dysplasia. Epilepsy Res 2010;90(3):240–247 10. de Oliveira RS, Santos MV, Terra VC, Sakamoto AC, Machado HR. Tailored resections for intractable rolandic cortex epilepsy in children: a single-center experience with 48 consecutive cases. Childs Nerv Syst 2011;27(5):779–785 11. Behdad A, Limbrick DD Jr, Bertrand ME, Smyth MD. Epilepsy surgery in children with seizures arising from the rolandic cortex. Epilepsia 2009;50(6):1450–1461 12. Pondal-Sordo M, Diosy D, Téllez-Zenteno JF, Girvin JP, Wiebe S. Epilepsy surgery involving the sensory-motor cortex. Brain 2006;129(Pt 12):3307–3314 13. Otsubo H, Sharma R, Elliott I, Holowka S, Rutka JT, Snead OC III. Confirmation of two magnetoencephalographic epileptic foci by invasive monitoring from subdural electrodes in an adolescent with right frontocentral epilepsy. Epilepsia 1999;40(5):608–613

16. Romstöck J, Fahlbusch R, Ganslandt O, Nimsky C, Strauss C. Localisation of the sensorimotor cortex during surgery for brain tumours: feasibility and waveform patterns of somatosensory evoked potentials. J Neurol Neurosurg Psychiatry 2002;72(2):221–229 17. Ng WH, Ochi A, Rutka JT, Strantzas S, Holmes L, Otsubo H. Stimulation threshold potentials of intraoperative cortical motor mapping using monopolar trains of five in pediatric epilepsy surgery. Childs Nerv Syst 2010;26(5):675–679 18. Gallen CC, Schwartz BJ, Bucholz RD, et al. Presurgical localization of functional cortex using magnetic source imaging. J Neurosurg 1995;82(6):988–994 19. Tieleman A, Deblaere K, Van Roost D, Van Damme O, Achten E. Preoperative fMRI in tumour surgery. Eur Radiol 2009;19(10):2523–2534 20. Morrell F, Whisler WW, Bleck TP. Multiple subpial transection: a new approach to the surgical treatment of focal epilepsy. J Neurosurg 1989;70(2):231–239 21. Blount JP, Langburt W, Otsubo H, et al. Multiple subpial transections in the treatment of pediatric epilepsy. J Neurosurg 2004;100(2, Suppl Pediatrics):118–124 22. Palmini A, Gambardella A, Andermann F, et al. Intrinsic epileptogenicity of human dysplastic cortex as suggested by corticography and surgical results. Ann Neurol 1995;37(4):476–487 23. Schwartz TH, Bazil CW, Forgione M, Bruce JN, Goodman RR. Do reactive post-resection “injury” spikes exist? Epilepsia 2000;41(11):1463–1468 24. Onal C, Otsubo H, Araki T, et al. Complications of invasive subdural grid monitoring in children with epilepsy. J Neurosurg 2003;98(5):1017–1026 25. Devinsky O, Romanelli P, Orbach D, Pacia S, Doyle W. Surgical treatment of multifocal epilepsy involving eloquent cortex. Epilepsia 2003;44(5):718–723

48

  Anterior Peri-insular Quadrantotomy Giulia Cossu, Mahmoud Messerer, Sebastien Lebon, Etienne Pralong, Margitta Seeck, and Roy Thomas Daniel

Summary Disconnective surgery is increasingly practiced in children with pharmacoresistant epilepsy secondary to subhemispheric and hemispheric epileptic syndromes. A complete disconnection of the entire epileptogenic zone assures adequate seizure control. In patients with subhemispheric epilepsy and intact motor functions, the surgery should also ensure the preservation of motor functions. When the epileptogenic focus is localized to one frontal lobe anterior to the motor cortex and in presence of residual motor functions, the anterior peri-insular quadrantotomy with preservation of the primary motor cortex and of its efferent pathways represents a viable alternative to the traditional frontal lobectomy. The intraoperative functional mapping is a key point to guide the whole disconnective procedure and to preserve residual motor functions. From a technical perspective, the surgery may be summarized into four steps as follows: the suprainsular window, the anterior callosotomy, the intrafrontal disconnection, and the frontobasal disconnection. This chapter deals with the indications of this surgery and details the surgical steps of this anterior quadrant subhemispheric epilepsy surgery. Keywords:  subhemispheric epilepsy, frontal disconnection, disconnective surgery, quadrantotomy, peri-insular ­disconnection

„„ Introduction Surgery for pharmacoresistant epilepsy in children with subhemispheric and hemispheric epileptic syndromes has undergone a change with respect to most centers preferring to ­perform disconnective procedures over resective approaches. Adequate seizure control is assured by a complete disconnection of the entire epileptogenic zone. In subhemispheric epilepsy where motor function is intact, this disconnection should ensure the preservation of motor functions. When the epileptogenic focus is localized to one frontal lobe anterior to the motor cortex and in presence of residual motor functions, the anterior peri-insular quadrantotomy with preservation of the primary motor cortex and of its efferent pathways represents a viable alternative to the traditional frontal lobectomy. This chapter deals with the indications, surgical steps, and outcome of this form of subhemispheric epilepsy surgery.

„„ Indications Similar to all surgical procedures for resective or disconnective epilepsy surgery, a good concordance between clinical, radiological, and electrophysiological data is fundamental, all pointing toward an epileptogenic zone in one frontal lobe, anterior to the motor strip, in the presence of normal/near-normal motor functions. The etiological pathologies that present in this manner include cortical dysplasia (atrophic or hypertrophic variants), Sturge–Weber syndrome, prenatal ischemic insults, or hemorrhagic accidents.

An Illustrative Case A boy presented to us at the age of 6 years, with the diagnosis of pharmacoresistant epilepsy. He experienced his first seizure at the age of 3.5 years, which was generalized tonic–clonic and followed by a postictal left-sided hemiparesis. The seizures were refractory to carbamazepine, lamotrigine, and valproate. Electroencephalography (EEG) showed right rapid rhythmic activity and spikes over the right frontal region. The ictal and interictal EEG showed severe epileptogenic activity of the right frontal lobe (Fig. 48.1, Fig. 48.2, and Fig. 48.3). Magnetic resonance imaging (MRI) images revealed a large dysplasia that involved the right frontal cortex extending to the basal ganglia, upper insula, and corpus callosum (Fig. 48.4). 18-Fluoro2-deoxy-D-­glucose positron emission tomography (FDG-PET) showed a more extensive hypometabolism of the right hemisphere, maximal in the right frontal cortex, insula, and basal ganglia. High-density EEG with 256 electrodes suggested two right frontal foci. He was able to collaborate for the functional MRI and the hand and foot motor cortices were identified posterior to the dysplastic cortex. His seizure disorder was characterized by four to five generalized seizures per year, predominantly nocturnal, and starting with pain and tingling in the left leg. Possible dyscognitive seizures were also observed, during which the child responded only with a delay of several minutes, but correctly, indicating absence of loss of consciousness and aphasia during the seizure. He was also known for severe headaches (without lateralization) and vomiting, which occurred during the day but would wake him up sometimes from sleep. Sometimes these headaches were observed also during the postictal phase, and ­therefore a seizure-related origin was suspected.

460

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 48.1  Interictal EEG: quasiconstant discharges with predominance in the right frontal contacts (Fp2, F4, FC2, F4). Note that almost all discharges propagate to the left hemisphere, leading effectively to a bifrontal syndrome.

Fig. 48.2  Ictal EEG with right onset (arrow). No major changes except delayed reactivity was observed.

His neurological examination was normal, except for poor speech and a discrete right facial weakness. Examination of developmental milestones revealed poor language development, characterized by a reduced vocabulary, but had no deficiency of other expressive or comprehensive skills. He needed special schooling because he suffered from major inattentiveness. Due to this important attention deficiency, he scored low in most of the tasks. Following peri-insular anterior quadrantotomy, he has been seizure-free for the past 3 years, except one seizure that occurred when he forgot to take his antiepileptic drugs (Engel’s Class I). He now enjoys normal schooling and is developing well. Headaches have disappeared completely. Antiepileptic medications have been partially tapered off.

„„ Preoperative Assessment The neurological status may be often unremarkable. Delayed developmental milestones may be common. Mild to moderate speech disorders may be found when the epileptic focus

Fig. 48.3  Ictal EEG showing rhythmic spike–wave discharges with right frontal predominance.

is located in the left anterior quadrant. EEG recordings and ­video-telemetry are used to confirm that the epileptogenic foci are localized in the frontal lobe unilaterally with a secondary propagation of abnormal discharges to the rest of the hemisphere or to the contralateral lobe during the ictal phases. The brain MRI detects the presence of radiological abnormality in the frontal lobe. Radiological data should be concordant with electrophysiological findings.

„„ Anesthetic Considerations Specialized pediatric anesthesiologist should assist the surgical team. A perfect collaboration and communication is crucial to avoid complications. Patients are often fragile and may present multiples complications secondary to the chronic uptake of antiepileptic drugs. The collaboration may be difficult given the developmental delay. Large venous accesses are mandatory because the maintenance of normovolemia is imperative and massive bleeding may be expected because of the important vascularization of skull

48  Anterior Peri-insular Quadrantotomy

Fig. 48.4  (a) Axial fluid-attenuated inversion recovery (FLAIR) sequence showing the enlarged right frontal lobe with coarsened gyri and extensive white matter hyperintensities. (b) Axial T1-weighted sequence in the same slice as in (a). Dysplasia extends to the right insula and basal ganglia. The thalamus appears to be grossly symmetric. (c) Coronal view of the right frontal lobe confirming dysplasia in T2-weighted sequences. (d) Horizontal T1sequence focusing on the basal ganglia. Note that they are larger on the right than left side.

and brain. Invasive arterial pressure monitoring is also assured, as well as central venous pressure. Serial monitoring of oximetry, electrolytes, and coagulation parameters is fundamental. Ideally, anesthesia should not interfere with EEG recordings and provide neuroprotection while maintaining a stable intracranial pressure (ICP). The younger the patient, the more challenging the intervention is. Inhalatory gases, such as sevoflurane and isoflurane, are commonly used and coupled with moderate hypocapnia. Propofol may help in obtaining a reduction of ICP values. A normal cerebral perfusion should be guaranteed while avoiding hypertension. An adequate and opportune volume replacement is mandatory.

„„ Intraoperative Functional Mapping A careful analysis of the intraoperative anatomy is the first step to identify the primary motor and sensory cortices. The precentral, central, and postcentral sulci should be localized. The precentral sulcus is the most anterior vertical sulcus and in front of it the superior, middle, and inferior frontal gyrus (IFG) may normally be identified. The lateral sulcus and the opercular portion of the frontal lobe should be localized. The study of the preoperative MRI may help in identifying radiological anomalies, which may guide or confound the surgeon in the identification of perirolandic cortices. We have to remember that patients with subhemispheric epilepsy present structural abnormalities in the anterior quadrant and the anatomy may be distorted. Also, the vascularization may vary.

The safest way to identify the perirolandic structures is, however, through electrophysiology. A platinum electrode is positioned over the exposed cortex, where the pre- and postcentral gyri have been visually identified. The recording of a phase inversion at the stimulation of the contralateral median nerve allows the identification of the precentral (P22) and postcentral (N20) gyrus. We use the Ad-Tech electrodes and the parameters used to stimulate the median nerve are: 3.7 Hz, 10 mA, 200 microseconds.1 Stimulation parameters may vary according to the system used. The contralateral motor response is also used to verify the location of the motor cortex through a cortical anodic stimulation with trains of 5, 500 Hz, 200 micros, 5 mA. The same concept was used for the white matter stimulation to avoid injuries of the pyramidal bundle during the disconnection procedure. A continuous somatosensory evoked potential (SSEP) monitoring is also performed to guide the disconnection and avoid postoperative deficits.

„„ Operative Technique of Periinsular Anterior Quadrantotomy The patient is placed supine with the head fixed on clamps and turned toward the contralateral side. The incision is in the shape of “barn door” or can be through a large question mark incision. A large fronto-parieto-temporal craniotomy is performed and the dura is opened based inferiorly. This allows a wide exposure of the frontal lobe, the sylvian fissure, and temporal operculum. It is essential to have a good exposure of the perirolandic cortices for anatomical and electrophysiological identification.

461

462

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Step 1: Suprainsular Window The incision starts 5 mm above from the sylvian fissure at the IFG. The pia mater is coagulated and the operculum is removed in an anteroposterior fashion, through the pars orbitalis, pars triangularis, and pars opercularis (Fig. 48.5). The procedure is best performed in the subpial plane, to preserve the vessels in the sylvian fissure and to minimize bleeding. This allows the visualization of the superior half of the insula. The disconnection is then started in the depth of the circular cistern in the direction of the ventricle, interrupting the anterior limb of the internal capsule. This disconnection ends in the frontal horn of the lateral ventricle, just anterior to the head of the caudate nucleus (Fig. 48.6). During this step, the anterior thalamic radiation and frontopontine tract are disconnected (anterior limb of the internal capsule). The fibers of the superficial portion of the inferior fronto-occipital fasciculus (IFOF) and the arcuate fasciculus (AF) are also interrupted during the incision of the IFG. The IFOF starts from the posterior part of the temporobasal surface and from the occipital lobe, coursing along the lateral wall and roof of the atrium lining the inferior limiting sulcus of the insula and reaching the IFG, dorsolateral prefrontal, and premotor cortices. The deep IFOF portion, with its vertical temporofrontal course, is interrupted before reaching the prefrontal cortex, the middle frontal gyrus, and the frontobasal cortex. The AF encircles the insula connecting the superior temporal and the IFG. The fibers of the uncinate fasciculus are equally interrupted before their arrival in the orbitofrontal cortex.

junction, while the posterior limit corresponds to the callosal fibers originating from the premotor cortex. This ensures the complete section of the anterior part of the body, genu, and the rostrum of the corpus callosum (Fig. 48.7). The commissural fibers connecting the two frontal lobes at the rostrum and genu of the corpus callosum are thus disconnected, while sparing the posterior fibers arising from the perirolandic cortices.

Step 3: Intrafrontal Disconnection From the frontal opercular incision on the convexity, the pia is coagulated and incised from inferior to superior just anterior and parallel to the primary motor cortex (based on electrophysiological identification) (Fig. 48.5). The white matter is

Step 2: Anterior Callosotomy Once the homolateral ventricle is reached, the callosal fibers are cut in a parasagittal plane to reach the pericallosal cistern in the interhemispheric fissure (Fig. 48.6). The pericallosal c­ istern is identified using the arachnoidal anatomy and the anterior cerebral arteries. Following identification of the pericallosal arteries, the callosal disconnection is performed following the arteries in both directions. The anterior limit is the A2–A3

Fig. 48.5  Schematic representation of the disconnective procedure. The suprainsular window and the intradisconnection (just anterior to the primary motor cortex) is illustrated.

Fig. 48.6  The peri-insular incision is deepened from the frontal operculum till the circular cistern and then it proceeds in the direction of the ventricle, interrupting the anterior limb of the internal capsule. This part of the disconnection ends in the frontal horn of the lateral ventricle, just anterior to the head of the caudate nucleus. Also the anterior callosotomy is illustrated here, with a parasagittal cut to reach the pericallosal cistern.

48  Anterior Peri-insular Quadrantotomy

„„ General Considerations and Complications

Fig. 48.7  The medial part of the intrafrontal disconnection is here illustrated, along with the frontobasal disconnection with the cut crossing the rectus and orbital gyri. Also the anterior callosotomy is here illustrated, with the disconnection of the genu and rostrum of corpus callosum.

then disconnected deep to this pial incision till the ventricle is completely opened. Subsequently, the medial ventricular wall is incised to reach the interhemispheric fissure along its entire vertical dimension (Fig. 48.7). During this step, the frontal part of corona radiata is disrupted. The superior longitudinal fasciculus and the superior fronto-occipital fasciculus, with their horizontal parietofrontal courses, are traversed. The afferent inputs from the temporal, parietal, and occipital lobes to the frontal lobe are interrupted. Also, corticocortical intrafrontal connections between the precentral gyrus and the supplementary motor area, premotor areas, and prefrontal cortex are interrupted during this ­surgical step.

Step 4: Frontobasal Disconnection The pial incision is continued from the opercular cortex along the projection of the sphenoid ridge in the basal pia, crossing the orbital gyrus (Fig. 48.7). This disconnection is continued parallel to the olfactory tract along the olfactory sulcus and then medially across the gyrus rectus and the mesial pia. The disconnection continues in the depth till the ventricle is reached, while posteriorly it reaches the previous incision at the level of the rostrum. The frontal lobe is thus disconnected from the cingulate gyrus and the paraterminal and paraolfactory areas, thus limiting the afferences from the limbic system, principally from the amygdala.

Closure The ventricle is irrigated to ensure adequate hemostasis. A drain is left within the ventricle. The dura should be carefully closed and the wound is then closed in a multilayered fashion over a subgaleal drain. The wound drain is removed in 24 to 48 hours after surgery. The external ventricular drain is left in place for at least 4 to 5 days, until the cerebrospinal fluid becomes clear.

The complication rates in disconnective surgery are possibly less than anatomical resection cases due to several reasons. The surgical time and the blood loss are generally limited in disconnective procedure, thus diminishing the risk of postoperative hypovolemia and coagulopathy. Respecting pial planes and thus avoiding vascular injuries allows preservation of the viability of disconnected cortex. This should reduce the incidence of postoperative brain swelling and intracranial hypertension significantly.1 At least in hemispheric epilepsy surgery, disconnective procedures are known to reduce cavity complications like hydrocephalus due to the reduced cavity size and subsequent minor bleeds that occur in large cavity. Copious irrigation of the ventricular system during the procedure and postoperative ventricular drainage also aid this process. Subhemispheric epilepsy surgery relies on a good localization of the perirolandic cortices which forms the key to avoid a motor deficit after surgery. This identification is the key to judge the location of the intrafrontal disconnection and the extent of the callosotomy. The continuous intraoperative SSEP monitoring should remain stable during the whole procedure. Anticonvulsant drugs should be maintained at the same dosage as before surgery. This is kept at these levels for at least 3 months, following which these drugs can be tapered off based on clinical and EEG data. A prophylactic course of antibiotics is generally not necessary in the postoperative period.

„„ Outcomes The epileptogenic focus is localized in the frontal lobe in about 20% of cases of refractory epilepsy. The treatment of such type of epilepsy is not standardized and a frontal lobectomy is classically proposed. The surgery for frontal lobe epilepsy is more challenging than surgery for temporal lobe epilepsy.2,​3 A frontal lesionectomy is performed for small localized lesions when clearly identified.3 In many cases, however, the exact limits of the epileptogenic focus may not be determined and the resection of large areas of brain tissue may lead to higher rate of postoperative complication and neurological morbidities.4 When a large lesion is seen with concordant data limiting the electrophysiological abnormalities to this lesion, a disconnective procedure might allow to obtain the same rates of seizure control as large brain resections similar to as shown for posterior quadrantic ­disconnection.5,​6 This assumption is based on the fact that a complete isolation of the epileptogenic focus provokes an interruption of the pathways of abnormal epileptiform discharges. Our illustrative case, undergoing an anterior peri-insular quadrantotomy for a frontal cortical dysplasia, was seizure-free (Engel’s Class I) at 3 years of follow-up. In the postoperative period, the neurological functions are also expected to be preserved, because the motor strip has been carefully identified at the beginning of the procedure and an electrophysiological monitoring guides the whole disconnective process.

463

464

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

„„ Acknowlgement The authors thank Prof. Jean-Guy Villemure, former Head of our Department and pioneer of disconnective pediatric epilepsy surgery, for mentoring and guiding our epilepsy ­

surgery program. We would also like to express our gratitude to Mrs. Marion Brun, of the Department of Otorhinolaryngology at the University Hospital of Lausanne, for the creation of the diagrammatical images of the procedure.

References 1. Daniel RT, Villemure JG. Peri-insular hemispherotomy: potential pitfalls and avoidance of complications. Stereotact Funct ­Neurosurg 2003;80(1–4):22–27

4. Stone JJ, Reynolds MR, Leuthardt EC. Transient hemispatial neglect after surgical resection of a right frontal lobe mass. World Neurosurg 2011;76(3–4):361.e7–361.e10

2. Englot DJ, Wang DD, Rolston JD, Shih TT, Chang EF. Rates and predictors of long-term seizure freedom after frontal lobe epilepsy surgery: a systematic review and meta-analysis. J Neurosurg 2012;116(5):1042–1048

5. Daniel RT, Thomas SG, Thomas M. Role of surgery in pediatric epilepsy. Indian Pediatr 2007;44(4):263–273

3. Garcia PA, Laxer KD. Lateral frontal lobe epilepsies. In: Lüders HO, Comair YG, eds. Epilepsy Surgery. 2nd ed. Philadelphia, PA: Lippincott Williams and Wilkins; 2001:111–134

6. D’Agostino MD, Bastos A, Piras C, et al. Posterior quadrantic dysplasia or hemi-hemimegalencephaly: a characteristic brain ­malformation. Neurology 2004;62(12):2214–2220

49

  Posterior Peri-insular Quadrantotomy Giulia Cossu, Mahmoud Messerer, Sebastien Lebon, Etienne Pralong, Krothapalli Srinivasa Babu, Margitta Seeck, and Roy Thomas Daniel

Summary Surgery for refractory epilepsy has progressively evolved from techniques of resection to disconnection. When faced with a refractory epilepsy secondary to epileptogenic foci restricted to the posterior quadrant (temporo-parieto-occipital lobes) and when patients have residual motor functions, the posterior peri-insular quadrantotomy may represent the surgical strategy of choice. The objective of this surgery is to disconnect the posterior quadrant, while sparing key anatomical areas, namely the perirolandic cortices. The intraoperative and electrophysiological identification of pre- and postcentral gyrus is the prerequisite to perform the procedure safely. The surgery may be divided into two steps: (1) the infrainsular window and (2) the parieto-occipital disconnection (posterior perisylvian disconnection, intraparietal disconnection, posterior callosotomy, and posterior hippocampotomy). In this chapter we describe the indications, the functional correlates of each surgical step, and the surgical outcomes of posterior peri-insular quadrantotomy. Keywords:  posterior quadrant, epilepsy surgery, peri-insular disconnection, temporo-parieto-occipital disconnection

„„ Introduction Surgery for refractory epilepsy due to large brain lesions has progressively evolved from techniques of resection to disconnection. When the entire hemisphere is involved and hemiplegia is present, hemispherotomy is the treatment of choice. When the epileptogenic foci are restricted to the posterior quadrant (temporo-parieto-occipital lobes), the epilepsy is secondary to a static condition, and residual motor functions are present, then posterior peri-insular quadrantotomy may represent the preferred surgical strategy. The objective of this surgery is to disconnect the posterior quadrant, while sparing key anatomical areas, namely the perirolandic cortices.

„„ Indications The indication for a posterior peri-insular quadrantotomy relies on the concordance between clinical, radiological, and

electrophysiological data, which localize the e ­pileptogenic zone ­ unilaterally to the posterior quadrant. The pathological condition should be static in surgical candidates for this ­procedure. According to our experience and reports from literature, this may occur in cases of cortical dysplasia, Sturge–Weber syndrome, ischemic prenatal insults or sequelae of hemorrhagic/ ischemic accidents such as ruptured arteriovenous malformations/aneurysms, or following surgery for other conditions performed at a young age.

„„ Case Series Our experience with this cohort of posterior subhemispheric epilepsy includes 14 consecutive patients requiring surgery for refractory temporo-parieto-occipital epilepsy. This series (partly reported previously1) included eight males and six females with a mean age of 16.4 years. Infants presented with generalized clonic epilepsy or complex partial seizure with infantile spasms (two cases of 9 and 4 months of age, respectively). Older patients had partial seizures with a secondary generalization. The different etiologies are reported in Table 49.1. The responsible lesion was in the right posterior quadrant in eight patients. After a mean follow-up of 6.7 years, 11 patients presented with Engel seizure outcome Class IA, 1 patient with Engel Class IB, and 1 patient with Engel Class IIIA.

Table 49.1  The different pathologies provoking refractory epilepsy in the authors’ cohort of patients requiring a posterior quadrant disconnection

Etiology

Number

Percentage

Porencephaly

5

36

Cortical atrophy

3

21

Cortical dysplasia

3

21

Sturge–Weber syndrome

2

14

Hemorrhage after AVM rupture

1

7

Abbreviation: AVM, arteriovenous malformation.

466

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Illustrative Case This girl of 17 years of Serbian origin, right-handed, came to our attention after exacerbation of her seizures. The early history remained obscure, due to loss of medical documentations and conflicting information provided by family regarding her seizure onset. The family reported a previous head trauma, although no details were provided. Her magnetic resonance imaging (MRI), however, rather suggested an early insult during the postnatal period. The semiology included right parietal headaches followed by left hemibody clonic movements and dysesthesia. Altered respiration and oral automatisms were also present. Rare secondary generalizations were observed. Seizure frequency varied between three per day and three per month. Her neurological exam revealed a severe spastic hemiparesis and atrophy of the left side, including face and predominating upper extremity weakness. She also had a left superior quadrantanopia. She presented with mainly executive (“frontal”) and language-related deficits (semantic paraphasias, verbal memory impairment, object naming), which became somewhat accentuated during the postictal phase. The electroencephalography (EEG) showed electrographic findings suggestive for right parieto-temporo-occipital onset. After partial drug withdrawal, she developed right posterior

focal status epilepticus without loss of consciousness, allowing neuropsychological testing. No memory deficits, neglect, sensory extinction, or language alteration was noted during this status, confirming the transfer of most of right posterior functions to the contralateral hemisphere and, therefore, a very early insult, probably of vascular origin. The MRI showed a porencephalic cavity in the posteroinferior part of the right hemisphere. There was an atrophy of the parietal, temporal, and occipital lobes and a thinned-out corpus callosum with predominance to its posterior part, together with an overlying bone defect (Fig. 49.1). Positron emission tomography (PET) showed concordant hypometabolism of the atrophic right hemisphere predominantly in the temporal and occipital lobe (Fig. 49.2). Given the predominance of EEG findings over the posterior region, it was decided to proceed to a posterior quadrantic disconnection. Five years after the intervention, she was still seizure-free, with persistent mild cognitive impairment and some improvement compared to the preoperative period.

„„ Preoperative Assessment The examination of neurological status generally confirms the presence of homonymous hemianopia. Motor functions

Fig. 49.1  (a, b) Coronal views of a T1-weighted brain MRI showing a large porencephalic cavity in the posterior part of the right hemisphere. An atrophy and deformation of the temporal, parietal, and occipital lobes in the right hemisphere is evident, as well as a thinned-out corpus callosum, predominantly in its posterior part. An overlying bone defect was also present.

Fig. 49.2  Cerebral FDG-PET showing a concordant hypometabolism in the right posterior quadrant of the right hemisphere.

49  Posterior Peri-insular Quadrantotomy such as finger movements and foot tapping are preserved, while patients may show some degrees of parietocortical sensory disturbances. Preoperative assessment includes brain MRI and EEG recordings. The brain MRI reveals structural abnormalities restricted to the posterior quadrant. The EEG recording and telemetry confirm the localization of the epileptogenic focus in the temporo-parieto-occipital lobes with a secondary propagation of the abnormal activity to the contralateral hemisphere or anterior lobes during the ictal phases. In our surgical series, there was no need for invasive EEG monitoring. The pathology responsible for refractory seizures should be stable and nonprogressive, such as ­Rasmussen’s encephalitis. If the posterior peri-insular quadrantotomy is performed in the dominant hemisphere, then postoperative speech deficits might be a potential problem. According to our experience, lesions provoking refractory epilepsy are often congenital (12 out of 14 cases in our series) and a functional shift toward the nondominant hemisphere is generally present. A complementary functional MRI may be used to assess this issue ­preoperatively.

„„ Anesthetic Considerations The majority of patients with subhemispheric epilepsy are children; some are infants2,​3,​4 and anesthesia in these situations may be challenging. Most of these children also show developmental delay and are often uncooperative. They may have systemic complications secondary to chronic antiepileptic drug treatment, such as gum hypertrophy, loose teeth, enlarged adenoids, or cardiac problems. Some antiepileptic medications are also ­a ssociated with enzymatic induction and modified metabolism of narcotics and muscle relaxants. After orotracheal intubation, a central venous line is placed in the right internal jugular or subclavian vein. A periprocedural blood loss should be minimized and two peripheral venous accesses should be secured as well as an arterial access (generally radial artery) for close monitoring of the oxygenation, invasive arterial pressure, temperature, expired carbon dioxide and anesthetic gases, urine output (indwelling catheter), serial hematocrit, arterial blood gases, electrolytes, and coagulation parameters. Continuous intra-arterial blood pressure and central venous pressure monitoring is mandatory because these monitors help in optimizing volume replacement. Normovolemia is maintained through the combined use of crystalloids, colloids, and transfusions when required. According to the coagulation status, fresh frozen plasma and cryoprecipitates should be readily available. Isoflurane and propofol are often combined to obtain a general anesthesia with analgesia and neuroprotection. With younger patients, the management of blood losses, volume replacement, hemodynamics, and temperature is more difficult and working with a team with expertise in the management of young children has critical significance and should be the standard of care.

„„ Intraoperative Functional Mapping Anatomical Localization of Perirolandic Cortices Careful analysis of the intraoperative anatomy and preoperative MRI guides the surgeon in the primary identification of the central sulcus. The vascularization may also help, but in a previous article, we reported that a predominant vein was present in the central sulcus in only 68% of cases.5 Cortical dysplasia, structural abnormalities, or brain shift may be present and may modify the classical anatomic landmarks. The preoperative brain MRI should be carefully studied. Surgical navigation systems and intraoperative MRI may also help in the identification of the central sulcus but electrophysiological monitoring is still considered the gold standard.

Electrophysiological Localization of Perirolandic Cortices The principal intraoperative application of electrophysiological tests is the localization of the perirolandic cortices and of the pyramidal bundle. Direct cortical electrical stimulation with a monopolar cortical anodic stimulation is performed using trains of 5, 500 Hz, 200 microseconds, 5 mA to elicit a contralateral motor response and localize the motor cortex. On the contrary, a cathodic stimulation is used for the white matter to identify the various bundles and avoid injury to the pyramidal tract. A continuous somatosensory evoked potential (SSEP) monitoring is also performed: the contralateral median nerve is stimulated at the wrist and a quadripolar subdural platinum electrode (Ad-Tech, Racine, WI) is used for this purpose in our institution. Each stimulus is a constant bipolar electric pulse of 200 μs at a rate of 4.7/second. The stimulus strength should be settled to obtain a moderate thumb twitch. Each recording is an average of about 100 responses. The reference electrode is localized at Fpz according to the International 10–20 system, with the ground electrode at Erb’s point. The electrode is moved on the exposed cortex to localize the hand area, corresponding to the area where the maximal response is elicited at 20-ms latency. In 2 of the 20 electrodes, the phase inversion between the frontal (P22) and the parietal (N20) response to the contralateral median nerve stimulation (3.7 Hz, 10 mA, 200 micros) is recorded: this indicates that these two electrodes are correctly located over the precentral and postcentral gyri, respectively.6 The SSEP should remain stable during the whole disconnection procedure: this ensures that perirolandic cortices are not damaged.

Electrocorticographical Localization and Monitoring of Perirolandic Cortices Electrocorticography is used intraoperatively to determine the exact localization of the epileptogenic zone and to control the completeness of disconnection. This technique has a higher spatial resolution than scalp EEG. A grid

467

468

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches or strip electrode is placed on the cortical surface and the number of electrodes may vary from 4 to 256. Scalp EEG electrodes are also placed outside the operative area. Subdural electrodes may be slid underneath the dura to monitor cortical regions not exposed by the craniotomy. Depth electrodes may be implanted to record activity from the mesial temporal lobe. This allows monitoring the EEG during the stages of the d ­ isconnection. The absence of epileptiform discharge propagation ­­ to the ipsilateral frontal lobe and to the ­contralateral hemisphere attests to a complete disconnection.

„„ Operative Technique of Peri-insular Posterior Quadrantotomy The patient is placed supine and the head is fixed in a threepin Mayfield clamp and turned toward the opposite side in a slightly extended position. In younger children, the head may rest in a soft support. A cushion is put under the ipsilateral shoulder to facilitate head rotation without excessive neck stretching. A “barn-door” incision is performed, with the anterior limb starting at the superior edge of the zygoma, the medial part passing about 1.5 cm from the midline, and the posterior limb descending till the osseous projection of the transverse sinus. A large bone flap is removed and the dura is reflected inferiorly. A large exposure of the pre- and postcentral gyri is fundamental to identify and preserve them during the procedure (see the section Intraoperative Functional Mapping). This ensures that new postoperative deficits are avoided, thereby increasing the safety of surgery.

Step 1: Infrainsular Window The surgical procedure starts with the coagulation of the pia mater of the superior temporal gyrus, in an anteroposterior direction, 5 mm away from the sylvian fissure. The temporal opercular cortex is removed subpially and the disconnection is continued in the depth till the inferior half of the insula is reached, thus exposing the inferior half of the circular cistern (Fig. 49.3). From the inferior limiting sulcus, the disconnection is continued in an oblique fashion till the temporal horn of the lateral ventricle is reached. Once in the ventricle, the amygdala is identified in the anteromedial part of the ventricle and it is resected, with the uncus and the anterior part of the hippocampus, till the choroid fissure. The whole resection should be performed subpially and the vascularization should be carefully preserved.

Functional Neuroanatomy During the passage through the inferior limiting sulcus toward the roof of the temporal horn of the lateral ventricle, the white matter bundles of the temporal stem are incised.7 The insulo-opercular fibers are disconnected superficially, followed by the incision of the uncinate and inferior occipitofrontal fasciculi anteriorly (coursing in the ventral part of the extreme and external capsule, respectively) and of the middle longitudinal fasciculus posteriorly (coursing in the dorsal extreme capsule).7,​8,​9 The sublenticular and retrolenticular parts of the internal capsule are also traversed, with the interruption of the acoustic radiation, the corticotectal fibers, the temporopontine fibers, and the corticothalamic fibers, as well as the optic radiations forming the Meyer’s loop.10 The tail of the caudate ­nucleus is also incised during this step. It extends along the anterior wall of the atrium and curves superiorly. The connections with the limbic system are ­interrupted through the resection of the mesial temporal structures (uncus, hippocampus, and Fig. 49.3  Illustration of the main steps of the disconnection procedure. The temporal opercular cortex is removed and the disconnection is continued in the depth till the inferior half of the insula is reached. Then the disconnection is continued in an oblique fashion till the temporal horn of the lateral ventricle is reached (infrainsular window). The incision in the superior temporal gyrus is continued posteriorly while preserving the vein of Labbé and the peripheral branches of M4. The parietal opercular cortex is aspirated and the white matter is traversed to reach the ventricular atrium and the posterior part of the body of the homolateral ventricle (posterior perisylvian disconnection). The disconnection is then continued superiorly just posterior to the postcentral gyrus till the interhemispheric fissure (intraparietal disconnection) is reached. The black arrow illustrates the disconnection of the splenium. The whole resection should be performed subpially and the vascularization should be carefully preserved.

49  Posterior Peri-insular Quadrantotomy ­ mygdala). The insular cortex, with its afferents and efferents a to the limbic system (amygdala, ento- and perirhinal cortex, orbitofrontal cortex, and cingulate gyrus), dorsal thalamus, and sensory and auditory cortex, is also disconnected.

Step 2: Parieto-occipital Disconnection • Stage 1 (posterior perisylvian disconnection): The incision in the superior temporal gyrus is continued posteriorly while preserving the vein of Labbé and the peripheral branches of M4. The parietal opercular cortex is aspirated and the white matter is traversed to reach the ventricular atrium and the posterior part of the body of the homolateral ventricle (Fig. 49.3). The medial wall of the atrium, composed by the calcar avis and the bulb of the callosum, the tail of the hippocampus, the crus fornicis, the precuneus, and the cuneus, is disconnected behind the choroid plexus of the atrium in an ascending direction till the roof of the atrium, formed by the splenium of the callosum, is reached. The disconnection of the medial wall of the atrium should be performed posteriorly to the glomus to avoid damages to the thalamus. yy Stage 2 (intraparietal disconnection): The disconnection is continued superiorly from the parietal operculum and just posteriorly to the postcentral gyrus. The incision is deepened till reaching the interhemispheric fissure and the falx. The disconnection is done up to the pia along the falx to reach the sagittal sinus superiorly and the corpus callosum inferiorly (Fig. 49.4 and Fig. 49.5).

Fig. 49.4  Intraoperative photograph showing the infrainsular window and the intraparietal disconnection in a patient with Sturge–Weber syndrome that involved the posterior quadrant: the white arrow indicates that the temporal, parietal, and occipital lobes are completely disconnected from the frontal lobe. The small black arrow denotes the position of the central sulcus: the intraparietal disconnection is performed just posterior to the postcentral gyrus.

Fig. 49.5  A postoperative brain MRI in a patient with a large atrophic lesion of the posterior quadrant showing the infrainsular and intraparietal disconnections (white arrows) in the sagittal plane (a, T1-weighted MRI) and in the axial plane (b, fluid-attenuated inversion recovery [FLAIR] MRI).

469

470

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches yy Stage 3 (posterior callosotomy): An intraventricular posterior callosotomy (splenium) is performed in the parasagittal plane. This incision reaches the previously performed intraparietal disconnection (disconnection at the level of the medial part of the parietal lobe). yy Stage 4 (posterior hippocampotomy): After the splenial disconnection, the medial wall of the ipsilateral ventricle is incised to disconnect the fornix. The incision exposing the medial pia has to reach the choroidal fissure to ensure complete hippocampotomy.

Functional Neuroanatomy • Posterior perisylvian disconnection: The posterior limb of the internal capsule is interrupted at its posterior edge during this stage, thus interrupting the fibers coming from the posterior parietal lobe. yy Intraparietal disconnection: The cortico-cortical connections through the long and short arcuate fibers are here disconnected. Furthermore, the superior longitudinal fasciculus (SLF) is interrupted. This fasciculus is divided into four components: the SLF I starts from the superior parietal and frontal lobes to the dorsal premotor and dorsolateral prefrontal regions; the SLF II extends from the angular gyrus to the prefrontal regions; the SLF III extends from the supramarginal gyrus to the ventral premotor and prefrontal cortex. The fourth component is the arcuate fasciculus, which encircles the insula and connects the superior temporal gyrus with the inferior frontal gyrus.11 The cingulum is also interrupted in the depth of the intraparietal disconnection. yy Posterior callosotomy: The parieto-occipital commissural fibers crossing at the level of the posterior part of the body and splenium of the corpus callosum are interrupted (forceps major). yy Posterior hippocampotomy: The section of the fornix ensures the disconnection of the efferents from the hippocampus. At the end of this procedure, the temporo-parieto-­occipital cortices are isolated from the frontal and central cortex, basal ganglia, and contralateral hemisphere.

„„ Closure and Early Postoperative Management During the whole disconnective procedure, the large branches of the middle and posterior cerebral arteries supplying the temporal, parietal, and occipital lobes, as well as the most important draining veins, should be carefully preserved to avoid postoperative brain swelling. A continuous intraoperative electrophysiological monitoring is performed and the evoked potentials should remain stable during the whole procedure to ensure the absence of new postoperative deficits. The ventricle is irrigated and an intraventricular drain is left in place. The dura should be carefully closed and the wound is closed in a multilayered fashion over a subgaleal drain. The subgaleal drain is removed 24 to 48 hours postoperatively, while the intraventricular drain is left in place for at least 4 to 5 days, until the cerebrospinal fluid (CSF) becomes clear.

Anticonvulsant drugs should be maintained at the same dosages as in the preoperative period 2 to 3 months after surgery. A progressive taping of the treatment may be performed thereafter according to the clinical evolution. A prophylactic antibiotic therapy is generally not necessary in the postoperative period.

„„ Complications Several complications are described for epilepsy surgery in infants and children. The risk of complications is considered lower with disconnective techniques compared to resective procedures, especially for hemispheric/subhemispheric surgery. Operating time and intraoperative blood loss have lessened with disconnective techniques and, above all, the risk of cavity-related complication is expected to be lower. The preservation of the vascularization with disconnective procedures is the key to avoid brain swelling. CSF drainage should be carefully controlled to avoid overdrainage and resultant remote hemorrhages (distant from the operative site). The incidence of early and late hydrocephalus may be reduced through the performance of a good hemostasis and reducing the spillage of blood into the ventricles using cottonoids and irrigation.12 During the whole procedure, anesthesiologists should replace blood loss to avoid hypovolemia and ­coagulopathy. Epilepsy surgery may be defined as an “anatomy-based microsurgery” and a clear knowledge of the anatomy and a careful study of the preoperative MRI are fundamental factors to avoid complications and to achieve a complete disconnection of the epileptogenic focus.

„„ Outcomes Literature data reporting the outcome in patients with temporo-parieto-occipital disconnection are scarce, while more data are available for multilobar surgery. We previously reported1 a cohort of patients who underwent posterior quadrant epilepsy surgery: Engel Class I outcome was obtained in 92% of patients at 6 years of follow-up. No significant morbidity and mortality were reported. The quality of life was strongly improved and antiepileptic medications were stopped in the majority of patients. The hyperactive or aggressive behaviors also improved. Boesebeck et al reported 68.5 and 48% of their patients as having Engel Class I at 1 and 2 years after posterior quadrant surgery, respectively. Good prognostic indicators included lateralizing auras and lateralizing clinical seizures, tumor­ al etiology, and absence of epileptiform discharges in the ­postoperative EEG.13,​14 Jehi et al also reported Engel Class I postoperative outcome in 73% of patients at 6 months, 68.5% at 1 year, 65.8% between 2 and 5 years, and 54.8% at 6 years and beyond.15 According to their study, parietal resections fared worse outcome than occipital or parieto-occipital resections (52% seizure freedom vs. 89 and 93%, respectively, at 5 years). Most recurrences (75%) in this series occurred within the first 6 postoperative months.15

49  Posterior Peri-insular Quadrantotomy According to Dorfer et al, 9 of 10 patients undergoing posterior quadrantic disconnection were seizure-free in the postoperative period (Wieser Class 1a) at a follow-up of 2 years.3 Mohamed et al also reported a cohort of 16 patients with a posterior quadrant disconnection: 56% of patients were seizure-free and seizure reduction of greater than 50% was reported in 31% of patients after a mean follow-up of 52 months.16 Fifty percent of children showed developmental progress. None of the children developed new motor deficits postoperatively. In the postoperative period, generally parents report dramatic catch-up, developmental progress after surgery with rapid acquisition of new skills. Even if evidence is limited, we believe that the posterior quadrant disconnection may achieve the same results, in

terms of seizure control, with resective procedures. The success rate of posterior quadrant epilepsy surgery (in terms of Engel Class I or II) is variable in literature but our personal experience reports excellent seizure outcome in over 90% of patients.

„„ Acknowledgment The authors thank Prof. Jean-Guy Villemure, former Head of our Department and pioneer of disconnective pediatric epilepsy surgery, for mentoring and guiding our epilepsy surgery program.

References 1. Daniel RT, Meagher-Villemure K, Farmer JP, Andermann F, Villemure JG. Posterior quadrantic epilepsy surgery: technical variants, surgical anatomy, and case series. Epilepsia 2007;48(8):1429–1437 2. Daniel RT, Meagher-Villemure K, Roulet E, Villemure JG. Surgical treatment of temporoparietooccipital cortical dysplasia in infants: report of two cases. Epilepsia 2004;45(7):872–876 3. Dorfer C, Czech T, Mühlebner-Fahrngruber A, et al. Disconnective surgery in posterior quadrantic epilepsy: experience in a consecutive series of 10 patients. Neurosurg Focus 2013;34(6):E10 4. Thomas SG, Chacko AG, Thomas MM, Babu KS, Russell PS, Daniel RT. Outcomes of disconnective surgery in intractable pediatric hemispheric and subhemispheric epilepsy. Int J Pediatr 2012;2012:527891 5. Chandy MJ, Babu KS, Srinivasa BK. Surgery of perirolandic mass lesions with central sulcus mapping. Neurol India 1997;45(1):14–19 6. Allison T, McCarthy G, Wood CC, Jones SJ. Potentials evoked in human and monkey cerebral cortex by stimulation of the median nerve. A review of scalp and intracranial recordings. Brain 1991;114(Pt 6):2465–2503

9. Saur D, Kreher BW, Schnell S, et al. Ventral and dorsal pathways for language. Proc Natl Acad Sci U S A 2008;105(46):18035–18040 10. Rubino PA, Rhoton AL Jr, Tong X, Oliveira Ed. Three-dimensional relationships of the optic radiation. Neurosurgery 2005;57(4, Suppl):219–227, discussion 219–227 11. Makris N, Kennedy DN, McInerney S, et al. Segmentation of subcomponents within the superior longitudinal fascicle in humans: a quantitative, in vivo, DT-MRI study. Cereb Cortex 2005;15(6):854–869 12. Daniel RT, Villemure JG. Peri-insular hemispherotomy: potential pitfalls and avoidance of complications. Stereotact Funct Neurosurg 2003;80(1–4):22–27 13. Boesebeck F, Janszky J, Kellinghaus C, May T, Ebner A. Presurgical seizure frequency and tumoral etiology predict the outcome after extratemporal epilepsy surgery. J Neurol 2007;254(8):996–999 14. Boesebeck F, Schulz R, May T, Ebner A. Lateralizing semiology predicts the seizure outcome after epilepsy surgery in the posterior cortex. Brain 2002;125(Pt 10):2320–2331

7. Kucukyuruk B, Yagmurlu K, Tanriover N, Uzan M, Rhoton AL Jr. Microsurgical anatomy of the white matter tracts in hemispherotomy. Neurosurgery 2014;10(Suppl 2):305–324, discussion 324

15. Jehi LE, O’Dwyer R, Najm I, Alexopoulos A, Bingaman W. A longitudinal study of surgical outcome and its determinants following posterior cortex epilepsy surgery. Epilepsia 2009;50(9):2040–2052

8. Fernández-Miranda JC, Rhoton AL Jr, Kakizawa Y, Choi C, Alvarez-Linera J. The claustrum and its projection system in ­ the human brain: a microsurgical and tractographic anatomical study. J Neurosurg 2008;108(4):764–774

16. Mohamed AR, Freeman JL, Maixner W, Bailey CA, Wrennall JA, ­ Harvey AS. Temporoparietooccipital disconnection in children with intractable epilepsy. J Neurosurg Pediatr 2011;7(6): 660–670

471

50

  Tailored Extratemporal Resection in Children with Epilepsy Alessandro Consales and Massimo Cossu

Summary Extratemporal focal epilepsy is more frequent in children than in adults, and the rate of extratemporal resections usually exceeds 50% of cases in pediatric series of epilepsy surgery. More often than in temporal epilepsy, surgery for extratemporal epilepsy requires tailored, rather than standard, resections. This is explained by several reasons, including a prevalence of cortical malformations as etiology, involvement of highly eloquent areas, and significant proportion of magnetic resonance imaging (MRI)-negative cases. In many instances, invasive electroencephalography (EEG) evaluation is needed for the definition of the epileptogenic zone, employing either subdural electrodes or stereoelectroencephalography (SEEG). P ­ resurgical planning with advanced multimodal imaging allows integration of all the available structural and functional information for refining the strategy of invasive EEG evaluation and choosing the more appropriate and safer surgical approaches. Under the guidance of an accurate and individualized clinical, electrophysiological, and anatomo-functional presurgical workup, a broad array of tailored resection may be performed, from small-size lesionectomy to large multilobar resections or disconnections. ­Stereotactic neuronavigation and intraoperative neurophysiological monitoring allows for safer resections in eloquent cortex. Seizure outcome is less favorable than in temporal lobe resections, but seizure freedom is achieved in approximately 60% of children. Neuroplasticity enhances the potential of functional recovery in younger patients, and allows for resections in critical regions. Keywords:  drug-resistant epilepsy, children, epilepsy surgery, extratemporal epilepsy, epileptogenic zone, tailored resection, stereoelectroencephalography, seizure outcome

„„ Introduction The frequency of extratemporal epilepsy in children is higher than that of temporal lobe epilepsy, which typically prevails in surgical series of adult population. Indeed, the reported rate of extratemporal resections for childhood epilepsies usually exceeds 50% of cases.1,​2 There are certain characteristics of this patient group, including pathological substrates, duration of seizures, and surgical outcome, which are different than adults. Different pathological substrates distinguish extratemporal childhood epilepsies, such as ­ malformative

brain anomalies and low-grade cortical tumors.3,​4,​5 Postoperative outcome in extratemporal epilepsy is also better in children than in adults,6,​7 especially in patients with a shorter duration of epilepsy.8 Furthermore, the potential for functional recovery after extratemporal epilepsy surgery in eloquent areas is much higher in children as compared to adults.4,​9

„„ Patient Selection and Presurgical Evaluation Patient selection and presurgical evaluation include a tailored diagnostic workup, which aims to identify the epileptogenic zone (EZ), which is the cortical region of seizure onset and early spread of the ictal discharge.10,​11 The resection of the EZ is expected to result in seizure freedom.12

Noninvasive Evaluation Clinical assessment, video-electroencephalography (video-EEG) monitoring of ictal events, and high-resolution magnetic resonance imaging (MRI) remain the cornerstones of the preoperative evaluation. Age at seizure onset must be assessed, as well as seizure frequency. Detailed information including age of seizure onset and seizure frequency and information regarding ictal semiology should be obtained from the patient and family. Detailed seizure history about the presence of subjective manifestations (often underestimated in infants and in uncooperative children), type and chronology of ictal symptoms and signs, occurrence of loss of consciousness, presence of postictal deficits, factors that may trigger seizures, including sleep, etc., should be obtained. This initial step may provide crucial clues for possible lateralization and/or gross localization of ictal onset.13 Interictal EEG is helpful in determining the side and site of epileptiform activity. Long-term monitoring with scalp video-EEG recording is mandatory when electroclinical correlates are needed.14 High-resolution MRI provides essential localizing information in these patients. The MRI protocol for children is different than our adult protocol, and it has been previously described.15 The neuroradiologist must be aware of the electroclinical features of the patient in order to tailor the study to the specific requirements of each case. The most frequently encountered substrates in children are malformations of cortical development, including focal cortical dysplasia (FCD), and developmental tumors.

50  Tailored Extratemporal Resection in Children with Epilepsy Detection of FCDs may be challenging, as the findings are often subtle, and may require particular experience to be recognized. The patient’s age may also be relevant because incomplete myelination may render some cortical malformations occult to MRI, and therefore specific MRI protocols may be required in infancy to disclose epileptogenic lesions.16 Extratemporal epilepsy often involves highly functional areas and may alter the anatomical patterns. In these instances, functional MRI (fMRI) might be helpful to define the borders of the functionally critical areas for a safer resection. Interictal 2-deoxy-2[18F] fluoro-D-glucose positron emission tomography (FDG-PET) has a reported sensitivity of 60 to 80% to identify the areas of hypometabolism in patients with intractable extratemporal epilepsy and normal MRI.17 However, interictal FDG-PET may lack the specificity required for precise planning of cortical resection. It still provides valuable information for optimizing the strategy of invasive EEG evaluations. Single-photon emission computed tomography (SPECT) identifies the area of increased flow when obtained during a seizure and correlates very well with the cortical areas participating in the onset of ictal discharges.18 Subtraction ictal SPECT co-registered to MRI imaging (SISCOM) increases the sensitivity of ictal SPECT by about 70% and it has been reported to represent a good predictor of seizure-free outcome after surgery.19

Invasive Evaluation When the noninvasive phase does not provide sufficiently concordant findings, invasive EEG evaluation is indicated to localize the EZ and to define the resection area. An additional goal of invasive monitoring is functional mapping of eloquent areas. Two different types of electrodes are currently available: subdural strip/grid electrodes20 and depth electrodes for stereoelectroencephalography (SEEG).21 Some centers use combination of both subdural and depth electrodes. Every recording technique has certain advantages and disadvantages. Subdural electrodes provide a wide surface coverage of targeted cortical areas. Its disadvantages are recording of deep-seated structures, poor localization accuracy of implanted electrode contacts, and need for an additional craniotomy to remove the electrode even for nonresective surgical patients. The main disadvantages of SEEG are a sort of “tunnel vision” provided by single electrodes and limited spatial sampling of cortical areas. On the other hand, SEEG electrodes may target every cortical (lateral, mesial, intrasulcal) and subcortical region, and provide information on the progression of the ictal discharge within a three-dimensional space. In particular, insular cortex, where sampling may be especially important for the evaluation of some extratemporal epilepsies, is easily accessible by intracerebral electrodes through several trajectories (Fig. 50.1). Fig. 50.1  SEEG exploration. In this case, arrangement of intracerebral electrodes was designed to verify the hypothesis of a right temporoperisylvian origin of the patient’s seizures. (a) Electrode entry points on the pial surface (left), and intracerebral trajectories through the transparent pial surface reconstruction (right) are highlighted. (b, c) Electrodes imaged by postimplant CT scan co-registered to preimplant volumetric T1-weighted MRI: the exact position of electrodes and of single contacts may be accurately defined. Note that the insular cortex is targeted by several contacts of different electrodes (right-sided images).

473

474

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches Furthermore, functional mapping may be extended to subcortical regions with depth electrodes, and documenting the precise localization of electrode contacts is possible by co-registration of pre- and postimplantation neuroimages.22 Regarding the safety of invasive monitoring electrode implantation procedures, it has been reported that SEEG has lower risks than other methods of invasive recording.23 Both of these recording techniques are widely used in children and in infants.24 At Niguarda Hospital (Table 50.1), we have operated on 321 pediatric patients for extratemporal epilepsies; 25 patients (22%) were MRI-negative and 116 patients (36%) underwent SEEG evaluation, which allowed tailoring the resection based on electroclinical and functional information.

„„ Surgery Presurgical Planning Compared to standard temporal lobe resections (anterior temporal lobectomy, selective amygdalohippocampectomy), planning a tailored extratemporal resection may be a challenging task, especially when facing with an EZ adjacent

to or overlapping with eloquent areas. Availability of different morphological and functional information by image co-registration in a three-dimensional anatomical space provides the surgeon with an extremely useful tool for surgical planning. Several diagnostic modalities may be integrated in the same multimodal scene (Fig. 50.2), including cortical morphology, lesions, vasculature, white matter tracts, areas of activation at fMRI, intracranial electrodes, and PET scan findings. Multimodality co-registration imaging techniques provide valuable data for planning of SEEG implantations25 as well as surgical resection. Evaluation of the extent of the presumed EZ, targeted resection area, and their relation to functionally critical structures is defined with details to minimize surgical risks. The identification of anatomical landmarks, including the cortical vascular anatomy, allows for the best surgical approach. Appropriate postprocessing allows importing co-registered multimodal imaging to the neuronavigation system for intraoperative guidance (Fig. 50.3). Finally, it has been recently reported that the use of multimodal co-­registration in the presurgical planning has reduced the need for invasive EEG recording and improved the 2-year seizure outcome in children operated on for focal epilepsy.26

Table 50.1  Pediatric tailored extratemporal resections (Niguarda Hospital series, 1996–2016)

SEEG-guided resections (116 patients), N (%)

No SEEG (205 patients), N (%)

pa

Age at surgery (y)

10.8 ± 4.5

9.0 ± 4.8

n.s.

Age at seizure onset (y)

3.1 ± 3.0

3.7 ± 4.2

n.s.

Lesion

91 (78%)

205 (100%)

0.0001

No lesion

25 (22%)

0 (0%)

Frontal

43 (37%)

78 (38%)

n.s.

Posterior quadrant

33 (28%)

82 (40%)

0.04

Perisylvian/insular

17 (15%)

9 (4%)

0.002

Rolandic/perirolandic

12 (10%)

30 (15%)

n.s.

Other multilobar

11 (10%)

6 (3%)

0.017

Tumor (± FCD)

9 (8%)

52 (25%)

0.0001

FCD II

51 (44%)

64 (31%)

0.029

FCD I

21 (18%)

18 (9%)

0.020

Other MCDs

7 (6%)

13 (7%)

n.s.

Tuberous sclerosis

2 (2%)

31 (15%)

0.0001

Other pathologies

14 (12%)

15 (7%)

n.s.

Unremarkable

12 (10%)

12 (6%)

n.s.

Engel Class I

68 (60%)

150 (79%)

0.0009

Engel Class II–IV

45 (40%)

41 (21%)

MRI

Site of resection

Histology

Seizure outcomeb

Abbreviations: FCD, focal cortical dysplasia; MCDs, malformations of cortical development; MRI, magnetic resonance imaging; n.s., not significant; SEEG, stereoelectroencephalography. a Statistical analysis: Student’s t-test for continuous variables; Fisher’s exact test for categorical variables. b Only for 304 patients with follow-up > 12 months.

50  Tailored Extratemporal Resection in Children with Epilepsy

Fig. 50.2  Example of multimodal imaging postprocessing for surgical planning. (a) Patient with a type II, right postrolandic focal cortical dysplasia (FCD) (arrows). (b) Reconstructed profiles of the pial surface (left) and of white matter (right), with superimposition of structures of interest: lesion (red), corticospinal tract reconstructed from DTI (light blue), BOLD activation at functional MRI for the left-hand finger tapping (yellow). (c) Venous landmarks (phase-contrast sequences of angio-MRI) complete the scene.

Types of Resection The main goal of epilepsy surgery is to make patients seizure-free by removing the EZ. Favorable results may be obtained by tailoring the specific requirements of each patient among a broad array of surgical options, from a small-sized lesionectomy to extended lobar and multilobar cortical resections (Fig. 50.4). The term “lesionectomy” refers to resection of a presumably epileptogenic, structural abnormality. It can be indicated in patients with a recent onset of epilepsy and/ or low-frequency seizures with a lesion documented on MRI

(e.g., well-defined, small-size type II FCD). However, it must be emphasized that the decision to perform a lesionectomy in epilepsy surgery is made based on clinical, neurophysiological, and anatomical characteristics, and decision-making process is different than typical lesion-oriented neurosurgical approach. Lesionectomy is not necessarily synonymous with small-size resection. Lesionectomy for a large cortical malformation may be tailored as a multilobar resection to obtain seizure freedom. The extension of the resection may be determined by several variables, including epileptological, functional, and oncological

475

476

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 50.3  Plan of neuronavigation for a left perisylvian resection in a patient previously evaluated by SEEG recording. (a–c) The position of the SEEG electrodes (yellow) and the tractography of the arcuate and corticospinal bundles (white) are implemented. (d) The pial surface, the vascular anatomy (reconstructed from rotational angiography), and the position of SEEG electrodes are integrated into the same three-dimensional scene.

Fig. 50.4  Some examples of surgical resections for extratemporal epilepsy in children. (a) Pure lesionectomy. (b) Frontal lobectomy. (c) Parietal lobectomy. (d) Insulo-opercular resection. (e) Occipital lobe disconnection. (f) Combination of an all-but-central resection (frontal-temporal lobes) and disconnection (parietal-occipital lobes).

50  Tailored Extratemporal Resection in Children with Epilepsy considerations. The surgical plan is tailored based on the individual patient characteristics and goal of the procedure. The resection may be strictly limited to the epileptogenic portion of a large malformative lesion, such as a large size polymicrogyria27 or a subtotal removal may be planned as a best option when an epileptogenic lesion is partly embedded in highly eloquent area. On the other hand, if the lesion is a tumor, then total removal of epilepsy-associated tumor should be aimed whenever it is feasible. When the imaging studies do not show any structural abnormality, the surgical strategy is defined based on electroclinical information such as invasive EEG evaluation. In selected cases, tailored excision of EZ may extend beyond the lesion, and into the apparently normal cortex, when electroclinical data imply the presence of an epileptogenic area surrounding the lesion.

Technical Aspects Subpial dissection is the preferred technique for removal of epileptogenic cortex. The gyral-sulcal anatomy should be followed whenever possible, and major vessels crossing the resection zone should be respected to avoid ischemic injury to cortical areas not included in the resection plan. The ultrasonic aspirator may also be helpful, but its use should be limited in order to provide the pathologist a sufficient amount of tissue for histological processing. Microsurgical disconnection of the EZ, combined with or without tissue removal, may be a valuable alternative to resection. Disconnective procedures such as anatomical or functional separation of epileptogenic tissue, while sparing the vascular structures supplying blood from the disconnected areas, can be performed to avoid postischemic swelling. This was initially developed for hemispheric surgery, to reduce the hemispherectomy-related complications. It is based on the concept that disconnecting a structure is functionally equivalent to its removal. Likewise, lobar or multilobar tailored disconnections can be considered when a serious risk of brain shift exists after resective surgery (Fig. 50.4). This is recommended in the case of extended disconnections of the posterior quadrant28 in patients with enlarged ventricular system and potential risk of developing postoperative complications, including displacement of residual brain tissue causing extracerebral hematomas or fluid collections, hydrocephalus, and hemosiderosis. Disconnective epilepsy surgery is not indicated in the case of presumed neoplastic etiology of epilepsy. Neuronavigation is useful to identify both anatomically and functionally critical landmarks, including gyro-sulcal patterns, vascular structures, subcortical fascicles, and regions sampled by previously implanted intracerebral electrodes. When the removal of epileptogenic tissue is adjacent to the rolandic region and/or to the motor pathways, neurophysiological mapping and monitoring by cortical and subcortical electrical stimulations helps preserving these critical structures and minimizing the risk of neurological deficits.29 Neurophysiological mapping of language areas requires awake surgery, which may not be feasible in younger children.

Sites of Resections There is striking heterogeneity of anatomical resection sites across published series, particularly in patients with extratemporal epilepsy. For instance, an international survey on

pediatric epilepsy surgery documented that frontal, parietal, occipital, and multilobar resections represented, respectively, 50, 8, 5, and 37% of extratemporal cases.1 Nevertheless, resections including both parietal and occipital lobes were included in multilobar resections, making posterior quadrant resections strongly underestimated. Furthermore, both perisylvian and rolandic resections are often embedded either in frontal or in parietal cases. By analyzing the Niguarda experience, five localization patterns have been identified, which included frontal lobe (38% of cases), posterior quadrant (36%), perisylvian/ insular (8%), rolandic/perirolandic (10%), and other multilobar (8%), which seem to be more homogeneous in terms of clinical, anatomo-functional, and surgical peculiarities. Furthermore and interestingly, sites of tailored SEEG-guided resections significantly differ from cases not submitted to SEEG: there were more perisylvian/insular and multilobar cases in the SEEG group, whereas posterior quadrantic resections prevailed among non-SEEG patients (Table 50.1).

Histological Findings As stated before, the most frequent histological substrates of pediatric extratemporal focal epilepsies are represented by malformations of cortical development and epilepsy-associated tumors.30,​ 31,​32 The histological findings of the Niguarda series confirm this figure (Table 50.1). Interestingly, the rates of FCDs were significantly higher in cases operated on after SEEG, whereas tumors and tuberous sclerosis prevailed among patients submitted only to noninvasive evaluation.

„„ Outcomes Seizure Outcome The seizure outcome of patients receiving surgery for extratemporal epilepsies is less favorable compared to patients undergoing temporal lobe resections, the rate of seizure-free cases ranging from 54 to 66%.33 Furthermore, data from series with long-term follow-up document a progressive drop in the rate of seizure-free cases, from roughly 70% at 1 year to 45 to 50% at 5 years.34,​35,​36,​37,​38 Among the factors associated with seizure freedom, the employment of invasive EEG deserves attention: in our series, patients undergoing tailored, SEEG-guided extratemporal resections had fewer chances of seizure freedom (60%) than those operated on after noninvasive evaluation (79%) (Table 50.1).8 This is most likely related to complex nature of the cases requiring invasive investigations.

Functional Outcome Compared to adult patients, brain plasticity provides children with a higher potential for recovery from postoperative neurological deficits. There is therefore a strong rationale for offering surgery to children as early as possible, especially in epilepsies involving highly eloquent areas. Epilepsy surgery may improve cognition in children with drug-resistant epilepsy.39,​40 Quality-of-life issues are related to the degree of improvement after epilepsy surgery.41 Improvement of behavior, socialization, and higher cognitive functions is frequently observed after extra-

477

478

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches temporal cortical resection, but whether it should be ascribed to seizure control or to weaning of drugs is still unclear.

„„ Complications The improvement of anesthesiological and microsurgical techniques as well as advancement of functional preoperative planning and intraoperative mapping techniques has resulted in a decreased complication rate.42,​43,​44,​45 The reported mortality is less than 0.5%,43 and morbidity ranges between 5 and 10%.43,​46,​47 Surgical complications include surgical site infections, cerebral infarction, intracranial hemorrhage, cortical injury, and extracerebral fluid collections. Procedures for intracranial electrode implantation may also be complicated by adverse events, which include hemorrhage, infection, edema, and cerebrospinal fluid leakage for subdural electrodes, and hemorrhage for SEEG electrodes. Incidence of complicated invasive procedures seems to be lower (~1%) for SEEG compared to subdural (5–8%) implants.38

„„ Two Illustrative Cases Case One A 9-year-old right-handed girl had begun to experience seizures during sleep at 2 years of age. Her physical and cognitive development profiles were normal. At referral, she presented with several seizures every night. Seizures were characterized by awakening, agitated behavior, fearful look, wandering, and no clear speech impairment. Scalp video-EEG recording showed interictal and ictal activity over the left anterior leads. Brain MRI was initially reported as normal but higher resolution scan identified a region of subtle cortical thickening, with ill-defined borders, in the left orbital and anteromesial frontal regions (Fig. 50.5a), suggestive for FCD. SEEG exploration was carried out to identify the EZ (Fig. 50.5b), with multilead electrodes placed in the anterior portions of the left frontal and temporal lobes. Ictal onset was seen to involve the orbital cortex and Fig. 50.5  (a–d) Illustrative Case #1: For details see the text.

50  Tailored Extratemporal Resection in Children with Epilepsy Fig. 50.6  (a–d) Illustrative Case #2: For details see the text.

the anteromesial portion of the frontal lobe, with subsequent propagation to neighboring areas (Fig. 50.5c). A tailored frontal orbitopolar resection, including the genu of cingulate gyrus, was carried out (Fig. 50.5d). Histology was positive for a type IIb FCD. The patient is seizure-free 5 years after surgery.

Case Two A 3-year-old, adopted, normally developed young girl had presented her first seizures at 2 months of age. These occurred mostly during sleep: the patient woke up crying, scared as if she saw something frightening, with head and eyes deviated to the right and with a dystonic posture with her right limbs. Both

interictal and ictal EEG showed epileptogenic activity on the left posterior leads. Brain MRI was not informative, showing only subtle abnormalities of the gyral pattern in the posterior portion of the left hemisphere. A SEEG exploration was performed, by implanting intracerebral electrodes in the left occipital, temporal, and parietal lobes (Fig. 50.6a,b). Ictal electrical activity originated in the lateral cortex of the occipital lobe, propagating to the contiguous lateral portions of the temporal and parietal lobes (Fig. 50.6c). Resection of the lateral cortex of the left occipital lobe, with limited resection to the lateral temporoparietal cortices (Fig. 50.6d), was followed by a significant improvement of seizure frequency 6 years following surgery (Engel Class IIa). The histology was unremarkable.

References 1. Harvey AS, Cross JH, Shinnar S, Mathern GW; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epilepsia 2008;49(1):146–155

2. Cossu M, Lo Russo G, Francione S, et al. Epilepsy surgery in children: results and predictors of outcome on seizures. Epilepsia 2008;49(1):65–72

479

480

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches 3. Farrell MA, DeRosa MJ, Curran JG, et al. Neuropathologic findings in cortical resections (including hemispherectomies) performed for the treatment of intractable childhood epilepsy. Acta Neuropathol 1992;83(3):246–259

26. Perry MS, Bailey L, Freedman D, et al. Coregistration of multimodal imaging is associated with favourable two-year seizure outcome after paediatric epilepsy surgery. Epileptic Disord 2017; 19(1):40–48

4. Wyllie E, Comair YG, Kotagal P, Bulacio J, Bingaman W, Ruggieri P. Seizure outcome after epilepsy surgery in children and adolescents. Ann Neurol 1998;44(5):740–748

27. Cossu M, Pelliccia V, Gozzo F, et al. Surgical treatment of polymicrogyria-related epilepsy. Epilepsia 2016;57(12):2001–2010

5. Paolicchi JM, Jayakar P, Dean P, et al. Predictors of outcome in pediatric epilepsy surgery. Neurology 2000;54(3):642–647 6. Cascino GD. Surgical treatment for extratemporal epilepsy. Curr Treat Options Neurol 2004;6(3):257–262 7. Zentner J, Hufnagel A, Ostertun B, et al. Surgical treatment of extratemporal epilepsy: clinical, radiologic, and histopathologic findings in 60 patients. Epilepsia 1996;37(11): 1072–1080

28. Daniel RT, Meagher-Villemure K, Farmer JP, Andermann F, Villemure JG. Posterior quadrantic epilepsy surgery: t­echnical variants, surgical anatomy, and case series. Epilepsia 2007;48 (8):1429–1437 29. Sala F, Lanteri P. Brain surgery in motor areas: the invaluable assistance of intraoperative neurophysiological monitoring. J Neurosurg Sci 2003;47(2):79–88 30. Sinclair DB, Aronyk K, Snyder T, et al. Extratemporal resection for childhood epilepsy. Pediatr Neurol 2004;30(3):177–185

8. Liava A, Francione S, Tassi L, et al. Individually tailored extratemporal epilepsy surgery in children: anatomo-electro-clinical features and outcome predictors in a population of 53 cases. Epilepsy Behav 2012;25(1):68–80

31. Dagar A, Chandra PS, Chaudhary K, et al. Epilepsy surgery in a pediatric population: a retrospective study of 129 children from a tertiary care hospital in a developing country along with assessment of quality of life. Pediatr Neurosurg 2011;47(3):186–193

9. Centeno RS, Yacubian EM, Sakamoto AC, Ferraz AF, Junior HC, Cavalheiro S. Pre-surgical evaluation and surgical treatment in children with extratemporal epilepsy. Childs Nerv Syst 2006;22(8):945–959

32. Obeid M, Wyllie E, Rahi AC, Mikati MA. Approach to pediatric epilepsy surgery: state of the art, part II: approach tospecific epilepsy syndromes and etiologies. Eur J Paediatr Neurol 2009;13(2):115–127

10. Talairach J, Bancaud J, Szikla G, Bonis A, Geier S, Vedrenne C. Approche nouvelle de la neurochirurgie de l’épilepsie. Méthodologie stéréotaxique et résultats thérapeutiques. 1. Introduction et historique. Neurochirurgie 1974;20(Suppl 1):1–240

33. Spencer S, Huh L. Outcomes of epilepsy surgery in adults and children. Lancet Neurol 2008;7(6):525–537

11. Kahane P, Landré E, Minotti L, Francione S, Ryvlin P. The Bancaud and Talairach view on the epileptogenic zone: a working hypothesis. Epileptic Disord 2006;8(Suppl 2):S16–S26

34. Englot DJ, Breshears JD, Sun PP, Chang EF, Auguste KI. Seizure ­outcomes after resective surgery for extra-temporal lobe epilepsy in pediatric patients. J Neurosurg Pediatr 2013;12(2): 126–133

12. Lüders HO, Najm I, Nair D, Widdess-Walsh P, Bingman W. The epileptogenic zone: general principles. Epileptic Disord 2006;8 (Suppl 2):S1–S9

35. Englot DJ, Wang DD, Rolston JD, Shih TT, Chang EF. Rates and predictors of long-term seizure freedom after frontal lobe epilepsy surgery: a systematic review and meta-analysis. J Neurosurg 2012;116(5):1042–1048

13. Wieser HG, Williamson PD. Ictal semiology. In: Engel J Jr, ed. Surgical Treatment of the Epilepsies. 2nd ed. New York, NY: Raven Press; 1993:161–171

36. D’Argenzio L, Colonnelli MC, Harrison S, et al. Seizure outcome after extratemporal epilepsy surgery in childhood. Dev Med Child Neurol 2012;54(11):995–1000

14. Munari C, Kahane P. Traitement neurochirurgical de l’épilepsie. Encycl Méd Chir Neurole 1998

37. Hauptman JS, Pedram K, Sison CA, et al. Pediatric epilepsy surgery: long-term 5-year seizure remission and medication use. Neurosurgery 2012;71(5):985–993

15. Martinez-Rios C, McAndrews MP, Logan W, Krings T, Lee D, Widjaja E. MRI in the evaluation of localization-related epilepsy. J Magn Reson Imaging 2016;44(1):12–22

38. Blount JP. Extratemporal resections in pediatric epilepsy surgery-an overview. Epilepsia 2017;58(Suppl 1):19–27

16. Gaillard WD. Structural and functional imaging in children with partial epilepsy. Ment Retard Dev Disabil Res Rev 2000;6(3):220–226

39. Helmstaedter C. Neuropsychological aspects of epilepsy surgery. Epilepsy Behav 2004;5(Suppl 1):S45–S55

17. Juhász C, Chugani HT. Imaging the epileptic brain with positron emission tomography. Neuroimaging Clin N Am 2003;13(4): 705–716, viii

40. Freitag H, Tuxhorn I. Cognitive function in preschool children after epilepsy surgery: rationale for early intervention. Epilepsia 2005;46(4):561–567

18. Kumar A, Chugani HT. The role of radionuclide imaging in epilepsy, part 1: sporadic temporal and extratemporal lobe epilepsy. J Nucl Med Technol 2017;45(1):14–21

41. Hallböök T, Tideman P, Rosén I, Lundgren J, Tideman E. Epilepsy surgery in children with drug-resistant epilepsy, a long-term follow-up. Acta Neurol Scand 2013;128(6):414–421

19. Chiron C. SPECT (single photon emission computed tomography) in pediatrics. Handb Clin Neurol 2013;111:759–765

42. Ventureyra EC, Higgins MJ. Complications of epilepsy surgery in children and adolescents. Pediatr Neurosurg 1993; 19(1):40–56

20. Bruce DA, Bizzi JWJ. Surgical technique for the insertion of grids and strips for invasive monitoring in children with intractable epilepsy. Childs Nerv Syst 2000;16(10–11):724–730 21. Cossu M, Cardinale F, Castana L, Nobili L, Sartori I, Lo Russo G. Stereo-EEG in children. Childs Nerv Syst 2006;22(8):766–778 22. Cardinale F, Cossu M, Castana L, et al. Stereoelectroencephalography: surgical methodology, safety, and stereotactic application accuracy in 500 procedures. Neurosurgery 2013;72(3):353–366, discussion 366 23. Mullin JP, Shriver M, Alomar S, et al. Is SEEG safe? A systematic review and meta-analysis of stereo-electroencephalographyrelated complications. Epilepsia 2016;57(3):386–401 24. Cossu M, Schiariti M, Francione S, et al. Stereoelectroencephalography in the presurgical evaluation of focal epilepsy in infancy and early childhood. J Neurosurg Pediatr 2012;9(3):290–300 25. Nowell M, Rodionov R, Zombori G, et al. Utility of 3D multimodality imaging in the implantation of intracranial electrodes in epilepsy. Epilepsia 2015;56(3):403–413

43. Behrens E, Schramm J, Zentner J, König R. Surgical and neurological complications in a series of 708 epilepsy surgery procedures. Neurosurgery 1997;41(1):1–9, discussion 9–10 44. Tebo CC, Evins AI, Christos PJ, Kwon J, Schwartz TH. Evolution of cranial epilepsy surgery complication rates: a 32-year systematic review and meta-analysis. J Neurosurg 2014;120(6): 1415–1427 45. Bjellvi J, Flink R, Rydenhag B, Malmgren K. Complications of epilepsy surgery in Sweden 1996–2010: a prospective, population-based study. J Neurosurg 2015;122(3):519–525 46. Duchowny M, Harvey AS. Pediatric epilepsy syndromes: an update and critical review. Epilepsia 1996;37(Suppl 1): S26–S40 47. Goldring S. Surgical management of epilepsy in children. In: Engel J Jr, ed. Surgical Treatment of Epilepsies. New York, NY: Raven; 1987:445–464

51

  Surgical Management of MRINegative Extratemporal Lobe Epilepsy Jarod L. Roland and Matthew D. Smyth

Summary Surgically treated epilepsy most commonly involves the ­temporal lobe and is often associated with structural abnormalities on magnetic resonance imaging (MRI). Cases where structural MRI does not reveal a lesion are particularly challenging in the preoperative assessment. In pediatric epilepsy, such MRI-negative cases more often involve areas other than the temporal lobe, termed “extratemporal lobe epilepsy” (ETLE). MRI-negative ETLE presents a particular challenge to pediatric neurosurgeons. Here, we discuss the preoperative assessment of ­pediatric patients with medically refractory epilepsy that localizes outside of the temporal lobe and is not associated with an identifiable lesion on structural MRI. We frame this topic by the relevant and representative literature pertaining to the unique aspects of MRI-negative ETLE in pediatric patients. The discussion follows a typical workflow of evaluating seizure semiology and standard imaging followed by more advanced functional imaging studies (i.e., positron emission tomogra­ phy and single photon emission computed tomography) and invasive electrophysiological diagnostics (i.e., electrocorticography and stereo-electroencephalography). Despite these challenges, the published literature concerning MRI-negative ETLE supports high rates of surgical success, albeit lower seizure-­ freedom rates than comparable cases involving the temporal lobe or associated with a lesion on MRI. Keywords:  MRI negative, nonlesional, extratemporal, epilepsy, pediatric

„„ Introduction Epilepsy is a disease of the brain characterized by recurrent unprovoked seizures.1 The underlying etiology of such a ­predisposition may range broadly from focal cerebral lesions, such as cortical dysplasia, to systemic disease processes, such as L­ennox-Gastaut syndrome. A precise diagnosis is often ­challenging and requires a battery of clinical tests. For a neurosurgeon considering surgical intervention for medically refractory epilepsy (MRE), differentiating between various unique forms of epilepsy is an important factor. In particular, identifying focal from nonfocal seizure onset is an early dichotomy in determining appropriate treatment options. In the case of focal-onset seizures, the principle of seizure localization is to obtain sufficient concordant data to strongly implicate a single unilateral region in the genesis of a patient’s

typical seizures. Such a region is termed the “epileptogenic zone (EZ)” if its complete surgical resection or disconnection ends the occurrence of seizures.2,​3 By this definition, the EZ can only be confirmed after successful surgical disconnection results in seizure freedom. The seizure-onset zone (SOZ) is an electrographically defined area, as opposed to the EZ that is functionally defined. The SOZ is the region where seizure onset is first recorded, such as by electrocorticography (ECoG) or ­ stereo-electroencephalography. The EZ and SOZ typically overlap, but may not be identical. For example, an EZ may be an area of cortical dysplasia at the depth of a sulcus, which may or may not be evident on imaging, while the SOZ may localize to the crowns of neighboring gyri as measured by ECoG. The difference has functional importance for interpreting and applying the data at hand to enact the best surgical procedure in the treatment of a patient’s seizures. In the majority of surgically treated focal epilepsy, the EZ is associated with a lesion identified on standard preoperative diagnostic imaging. Advances in magnetic resonance imaging (MRI) have ­provided better noninvasive methods to identify brain abnormalities that may be associated with seizure generation. Higher magnetic field strength (e.g., 3T vs. 1.5T) provides a greater ­signal-to-noise ratio (SNR), which enhances the ability to detect ­subtle findings. Higher field strength and other advances to MRI sequences, such as FLAIR and diffusion-weighted imaging, have been shown to improve the detection of cortical dysplasia that are frequently associated with focal epilepsy.4,​5,​6,​7,​8 Reports of lesions identified on 3T MRI that were previously missed on 1.5T MRI range widely from 5 to 65%.9,​10 Rubinger et al studied this effect specifically in pediatric patients and found a nonsignificant trend of more studies detecting lesions after switching to 3T (92%) MRI compared to their prior experience with 1.5T (86%) MRI.11 A comparison with 7T MRI, which is not ­widely available to clinicians at the time of this writing, similarly showed detection of lesions not seen on lower field imaging in 23% of patients in a combined adult and pediatric cohort.12 These reports suggest that MRI-negative epilepsy represents a combination of cases where tissue is abnormal on histopathology but appears normal on standard imaging, along with other cases where the histopathology and imaging both appear ­normal.13 Lack of precise definitions for such cases can cause ambiguity in the literature. For example, a patient’s epilepsy may be labeled as cryptogenic if an underlying pathology is not evident, such as when the MRI appears normal. ­However, histopathology from surgical resection may identify cortical dysplasia that was not apparent on preoperative imaging. In this manner, a case termed “MRI-negative cryptogenic epilepsy”

482

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches before surgery may be more appropriately identified as l­esional secondary to focal cortical dysplasia (FCD), which could only be recognized after successful intervention. On the other hand, an otherwise similar case without abnormality on 3T MRI may also be normal on histopathology, and therefore termed “MRI-­ negative and nonlesional.” Precise definitions for such cases are imperative for accurate reporting and predicting outcomes. Surgical resection for MRI-negative and nonlesional histopathology is associated with a lower rate of seizure freedom. Reports on the proportion of patients undergoing surgical evaluation for MRI-negative epilepsy range from 16 to 32%.14,​15,​16 A meta-analysis by Téllez-Zenteno et al in 2010 found odds of seizure freedom to be 2.5 times higher when an abnormality was present on MRI or histopathology17 as compared to MRI-negative and nonlesional cases. They also found normal MRI and histopathology to be more common in extratemporal lobe epilepsy (45% in ETLE vs. 24% involving temporal lobe) and in pediatric patients (31% in children vs. 21% in adults). Therefore, despite continued advances in technology, MRI-negative cases will remain a significant factor in the neurosurgical treatment of pediatric epilepsy. In the preoperative workup of MRI-negative epilepsy, the challenge remains to find sufficient concordant information to accurately localize the EZ. The most common location of focal seizure onset in adults and children is the temporal lobe, but children have a relatively higher rate of ETLE.18 The International League Against Epilepsy (ILAE) organized a pediatric epilepsy task force to conduct an international survey of medical centers and reported the rates of focal cortical resections (45% of all 543 interventions) for pediatric epilepsy to be 23.2% in temporal, 17.5% in frontal, 2.8% in parietal, and 1.7% in occipital locations.15 Outcomes are significantly different between extratemporal and temporal lobe epilepsy (TLE).19 MRI-negative ETLE has lower success rates compared to TLE.17 Two meta-analysis reports by Englot et al address this topic by studying seizure outcomes in pediatric patients after treatment for ETLE and TLE. The TLE group included 1,318 pediatric patients treated by temporal lobectomy and found seizure freedom (Engel class I) in 76%.20 The ETLE group included 1,259 pediatric patients and found seizure freedom in only 56%.21 A similar meta-analysis was performed by Ansari et al in 2010 that focused on nonlesional cases defined by histopathology and found seizure freedom in 34% of the 95 pediatric patients included in their study.18 These data illustrate the difficulty in successfully treating ETLE. Given this challenge, we devote the remainder of this chapter to discussing the surgical considerations and nuances of treating MRI-negative ETLE. We begin by reviewing characteristic semiology of seizure onset with relation to anatomic regions. We then discuss other advanced imaging and diagnostic modalities that provide adjunct data for seizure localization when MRI is unrevealing. In the absence of abnormal findings on conventional structural MRI, termed “MRI-negative,” further workup involves various adjunct methods for additional imaging and diagnostic studies to define the EZ. Once sufficient concordant data are accumulated, the appropriate surgical intervention can be determined and offered to the patient.

„„ Seizure Semiology Semiology is a field of linguistics concerning the study of signs and symbols.22 In the study of epilepsy, seizure semiology

refers to the overt signs and symptoms of a seizure. Separate from the electrophysiological traits of a seizure, a description of the semiology helps characterize the observed physical manifestations of a seizure. Understanding the relationship between semiology and cerebral function is helpful in localizing the seizure to a brain region. However, the brain region associated with the semiology, known as the symptomatogenic zone (SZ), may be different from the EZ.23 For example, the EZ may be in a relatively quiescent area of the brain, such as the anterior frontal lobe, and only once the seizure has spread to neighboring cortex, such as the precentral gyrus, does a physical manifestation occur, which may be contralateral clonic movements of an extremity in this scenario. The challenge is that a SZ, such as the precentral gyrus, may similarly relate to an EZ in temporal lobe or the frontal lobe, depending on how the seizure spreads. Careful attention to the timing of onset for an observed action with other seizure-related signs, such as loss of consciousness, can be helpful in discriminating semiology related to the EZ (closer to the start) from the SZ (occurring later in the course). For this reason, many such observations are not specific in localizing the EZ. Yet, characterization of seizure semiology remains a cost-effective and nonintrusive diagnostic that is particularly useful in MRI-negative epilepsy.23 Here, we will discuss some of the most common and/or specific signs and symptoms that may prove helpful to localization by seizure semiology.

Hemispheric Lateralization A clinical sign that strongly implicates one hemisphere in seizure onset may be observed in the postictal state as a palsy of the contralateral affected limb or hemi-body.24 The paresis is temporary and typically recovers without intervention. The phenomenon of postictal palsy (PP) was described by Todd in 1849 and now carries the eponym Todd’s palsy or Todd’s paresis.25 More recent studies have combined video-EEG data with the observation of either postictal focal palsy or hemiparesis and found such a PP to be highly concordant with contralateral seizure onset.26,​27 While this sign is suggestive of involving the motor cortex of one hemisphere, it is not reliable in localizing to the frontal lobe or peri-Rolandic area, as spread from the temporal lobe and other areas is common. Additionally, an individual with bilateral seizure onset may have a single seizure that is strongly lateralized and associated with PP, while additional seizures may arise from the opposite hemisphere and not associate with the same postictal paresis. Therefore, in the MRI-negative setting, PP is not sufficient to define a single EZ.25 Beyond hemispheric lateralization, some ictal signs and preictal auras may help further localize a seizure to a single lobe. The usual caution in differentiating a SZ from the EZ remains as a given seizure semiology may be specific for involvement of a given brain region that may neighbor two distinct lobes. In this manner, the same SZ may be affected by an EZ in different lobes.

Frontal Lobe Seizures The frontal lobe is the most common location for MRI-­negative ETLE.28,​29 However, frontal lobe seizures are often difficult to localize by semiology alone. For a review of the complexities in surgical evaluation of frontal lobe epilepsy, see the article by McGonigal et al.30 Activation of the motor cortex by seizures involving the posterior frontal lobe is a classic semiology.

51  Surgical Management of MRI-Negative Extratemporal Lobe Epilepsy The primary motor strip is the posterior border of the frontal lobe. Seizures originating in the precentral gyrus are often characterized by onset with focal clonic movements in the ­distal extremity. As the ictal activity spreads to neighboring cortex, the clonic movements march along the proximal ­extremity and to the rest of the body before generalizing. The effect was described by John Hughlings Jackson and now carries the eponym “Jacksonian march.”31 Hypermotor, or hyperkinetic, seizures involve large and complex motor patterns, such as appearing to run or ride a bike.23 These seizures may involve the ventromedial frontal lobe. Seizures involving limbic areas of the ventromedial prefrontal cortex also have an association with an intense fear, or other emotional response and manifest in facial expression, screaming, and fighting or agitation.23,​32 The dorsolateral frontal lobe includes the frontal eye fields (FEF). Seizures involved the FEFs are often manifested with gaze deviation to the contralateral side. Versive seizures are a type of seizure involving a stereotyped head and eye turning. When this stereotyped movement, termed “version,” is present early in the seizure course, the EZ is often in the frontal lobes and reliably lateralizes seizure onset contralateral to the direction of eye gaze. However, version may occur later in the seizure, after loss of consciousness, when involvement of the FEF occurs secondary to ictal spread originating in the ­temporal lobe.33 Nocturnal frontal lobe epilepsy (NFLE) is characterized by seizures arising from the frontal lobes and occur during sleep.34 Onset is most often during childhood and can be difficult to differentiate from other parasomnias that are also common in children.35 Patients may experience simple stereotyped movements or complex motor patterns, sudden brief arousals, and vocalization or other expressions of fear. Nobili and colleagues reviewed their experience with tailored frontal lobe resection for NFLE in 21 individuals and report a 75% seizure freedom rate, suggesting a high rate of success in surgically treating NFLE.36 A meta-analysis by Englot and colleagues identified 1,199 patients in 21 studies surgically treated for frontal lobe epilepsy and reported a seizure freedom rate of 45%. When divided by preoperative imaging findings, the seizure freedom rates were lower in MRI-negative cases (39.2%) compared to MRI-positive (60.7%) cases.37

Parietal-Occipital Onset Primary sensory cortex resides in the postcentral gyrus as the anterior border of the parietal lobe. Seizures originating from this area may involve distinct somatosensory auras that are typically distal in the contralateral extremity.23 These auras may be described as painful or a less well-qualified paresthesia. Cephalic sensations, which are described as light-­headedness or similar vague sensations related to the head, are a related sensory aura that occurs in parietal lobe ­seizures,38 although not with great specificity.39 Similar auras can also be misinterpreted as vertigo.38 Visual auras are common in parietal and occipital lobe epilepsy, likely due to their close anatomic location.39 Seizures originating in the occipital lobe are often associated with visual auras.39 The visual auras from occipital lobe ­seizures tend to be more elementary (e.g., contralateral multicolored circles) when compared to parietal or temporal onset.

Ictal amaurosis or focal visual loss is a unique aura of occipital-­ onset seizures.40 Furthermore, many individuals with occipital lobe epilepsy also have preoperative visual field defects.41,​42 In addition to field defects, abnormal eye movements and sensations of abnormal eye movement are also suggestive of occipital ­epilepsy.43 Much of the literature evaluating parietal and occipital lobe epilepsy combines the two lobes, referred to as posterior cortex, owing to the difficulty in localizing the EZ to one or the other lobe. This may be due to rapid spread of seizures between the two lobes or due to difficulty studying the interhemispheric cortex of the mesial occipital lobe.43 For this reason, accurate investigation of the mesial surface and differentiation from the lateral surface of the occipital lobe often necessitates invasive electrophysiological monitoring.42 Despite these challenges, successful surgical treatment by way of resection or disconnection procedures can be successful in posterior cortex epilepsy with reported rates of 60 to 75% seizure freedom. 38,​43,​44,​45,​46 In pediatric case series of occipital lobe surgeries, seizure freedom rates range from 33 to 86%. 47,​48,​49,​50,​51 A meta-analysis by Harward et al in 2017, including adult and pediatric case series, reported an overall seizure freedom rate of 65% with age less than 18 years being a statistically significant predictor of seizure freedom.52

Insular Seizures Seizures arising from insular cortex can be challenging to localize and frequently mimic temporal lobe seizures.53 False localization of an insular EZ may account for some cases of failed temporal lobe surgery.53,​54 Indeed, some of the earliest insights into the function of the insula resulted from observations during epilepsy surgery performed for seizures that initially localized to the temporal lobe.55 Therefore, an insular EZ should be considered when electrodiagnostics suggest temporal onset but imaging and semiology are not consistent with TLE.2 Insular seizures have been described as starting with perioral and laryngeal sensations culminating in focal motor activity.54 A seizure semiology with combined somatosensory, visceral, and motor signs and symptoms should alert the physician to investigate the insula. On the contrary, loss of consciousness is not consistent with purely insular­ epilepsy.2 Invasive monitoring of insular cortex can be challenging due to its deep location. ECoG does not permit direct measurements along the depths of the Sylvian fissure or insular surface. Electrophysiological monitoring via penetrating depth electrodes is often required to adequately study the insula. These can be combined with surface electrodes or by the method of ­stereo-electroencephalography.56,​57

Confirming Localization Localization by clinical assessment of seizure semiology is a first step and can be a useful data element when combined with additional concordant information to localize the EZ. Howe­ver, semiology alone is insufficient in the absence of a lesion on MRI. For this reason, the presurgical evaluation of MRI-negative ETLE is highly dependent on advanced imaging and invasive electrophysiology diagnostics.

483

484

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

„„ Advanced Imaging and Electrophysiology Positron Emission Tomography Positron emission tomography or PET is a functional imaging modality that provides an indirect measure of cerebral metabolism. First a radionuclide, also known as a tracer, is administered by peripheral venous injection. The tracer emits positrons that then result in gamma ray emission. The tracer is preferentially taken up in areas of the brain undergoing active metabolism. When the gamma rays are detected by the scanner, the spatial location where the signal originated can be reconstructed by computed tomography to create maps of cerebral metabolism. PET imaging in clinical pediatric epilepsy typically utilizes the glucose analog 18 F-fluorodeoxyglucose, or FDG, as a tracer to visualize areas of relative hypometabolism, which corresponds to the EZ. In cases where structural MRI is unrevealing, PET may show functional abnormalities that are helpful in localizing the EZ. In 2014, Perissinotti et al studied 54 pediatric patients undergoing presurgical workup for refractory epilepsy and found that PET imaging was able to localize the presumptive EZ in 57% of cases compared to 39% with structural MRI.58 Among only the MRI-negative cases, which was 61% of their cohort, PET was able to localize the presumptive EZ in 67% of those cases. A larger study of 194 patients (mostly adult) investigating PET in MRI-negative (n = 158) or discordant (n = 24) cases reports similar localization results with PET finding unilateral focal hypometabolism in 51% of cases, of which 54% were entirely extratemporal.59 Of cases that where MRI-negative, PET revealed areas of hypometabolism in 62%, similar to that of Perissinotti et al’s pediatric cohort.58,​59 The combination of PET and MRI helps identify imaging abnormalities that are difficult to detect in a single modality alone. By coregistering the two studies, the functional PET results may be color coded and overlaid on the structural MRI (Fig. 51.1). The combined imaging results provide improved identification of subtle lesions, such as some cortical dysplasia, that may have previously been unrecognized.60

Fig. 51.1  PET fused to MRI. Anatomic imaging of a patient represented by the T1-weighted MRI (a) did not identify a structural abnormality. PET imaging was coregistered and overlaid on the anatomic imaging (b) to localize an area of hypometabolism near the right posterior temporal–parietal–occipital junction.

Single Photon Emission Computed Tomography Single photon emission computed tomography or SPECT is another functional imaging modality that is complimentary to PET. The radioactive isotope tracer used in SPECT, commonly technetium-99m, directly emits gamma rays that are detected and reconstructed by computed tomography. Where PET uses a glucose analog to reveal hypometabolism during the interictal state, SPECT imaging typically begins with injection of the ­tracer during a seizure and is followed shortly by imaging to reveal regions of hyperperfusion associated with the ictal state. Alternatively, the tracer may be injected in the interictal period to obtain interictal SPECT, similar to the way PET is obtained. When both ictal and interictal SPECT are obtained, the two images can be algebraically subtracted to emphasize the differences in ictal and interictal perfusion. The difference image can then be aligned and overlaid on a structural MRI in a process known as subtraction ictal SPECT coregistered to MRI, or ­SISCOM (Fig. 51.2). In 2014, Perissinotti et al found SISCOM localizing in 67% of pediatric epilepsy cases compared to 39% for MRI alone.58 When PET was combined with SISCOM, the cumulative results localized 76% of all 54 cases, and 100% of the 18 cases in their series that proceeded to resective surgery.58 In 2013, Kudr et al reported their experience in 14 patients (mostly pediatric) with MRI-negative ETLE that were subsequently found to have focal cortical dysplasia after surgical resection and found localizing results in 13 (93%) cases with SISCOM and 5 (36%) with PET.61 A similar report by Kim et al from 2011 studied a mostly pediatric cohort with FCD and found localizing results in 63% of 48 who underwent structural MRI, 83% of 36 who underwent PET, and 89% of 27 who underwent ictal SPECT.62 These studies, and similar results in adult populations,63 demonstrate how SPECT and PET can be used independently and in combination to aid seizure localization. These studies are particularly applicable to MRI-negative epilepsy where the absence

Fig. 51.2  SISCOM. The difference image of interictal SPECT subtracted from ictal SPECT is overlaid on anatomic imaging (a) to localize an area in the left frontal cortex. This corresponds to a similar area of hypometabolism identified by interictal PET (b). This patient subsequently underwent invasive electrocorticography monitoring, which confirmed the seizure-onset zone in correspondence with functional imaging.

51  Surgical Management of MRI-Negative Extratemporal Lobe Epilepsy of a lesion on structural imaging leaves a dearth of concordant data. While PET and SPECT are valuable, false-negative cases are not uncommon and should not necessarily preclude an individual from further evaluation with invasive electrophysiology.

Magnetoencephalography Magnetoencephalography (MEG) is another technology with applications to localizing seizures and functional cortex. It has the advantage of not using ionizing radiation and is noninvasive. MEG is often combined with MRI structural imaging in a process known as magnetic source imaging (MSI). MSI is useful to the neurosurgeon for mapping the functional data provided by MEG to the patient’s anatomy for use in surgical decision making. Despite these benefits, however, MEG has relatively limited availability to many clinicians. For those who do have access to MEG, it has shown good results in various forms of epilepsy for localizing seizures and functional cortex in concordance with invasive studies such as ECoG and stereo-electroencphalography.64,​65,​66,​67 Interictal epileptiform discharges (IEDs) can be detected by examining dipoles recorded by MEG between seizures and has demonstrated value in identifying surgical targets and predicting seizure freedom. Englot et al examined their case series of 132 patients (age range: 3–68 years) who were studied with MEG prior to resective surgery.67 They noted that 43% of patients who had an extratemporal resection first underwent MEG for seizure localization. In their series, of patients whose MEG findings were concordant with area of resection, ECoG findings, or MRI abnormalities, 85% went on to seizure freedom.

Electrocorticography Electrocorticography may also be referred to as intracranial EEG or iEEG owing to its similarity with the more common, noninvasive, extracranial scalp EEG. Invasive monitoring with ECoG is typically performed via staged surgeries in a single admission, where electrodes are implanted in the first stage followed by explant and therapeutic intervention (e.g., resection) in the second. The advantages of ECoG over scalp EEG are greater spatial resolution afforded by placing electrodes directly on the cortical surface in the subdural space, better spectral resolution resulting partly from the absence of calvarium and scalp that impart a low-pass filter effect, and fewer artifacts from temporalis and extraocular muscles. However, these advantages result from a tradeoff with the risks of craniotomy and potential for infection during long-term monitoring. Overall, these risks are low and invasive monitoring can be well tolerated in children. In 2006, Johnston et al reviewed their experience over a 10+year history with ECoG in 112 mostly pediatric patients finding wound infection in 2.4% and a 7% overall rate of reoperation for a complication other than placing additional electrodes.68 Furthermore, transient deficits were observed in 3.3% of cases and no permanent deficits or death occurred in this series. A systematic literature review analyzed intracranial monitoring cases in children and adults and found higher rates of medical complications (5.7 vs. 4.3% minor and 4.5 vs. 1.7% major, respectively) and neurological deficits (11.2 vs. 5.5% transient and 5.1 vs. 3.3% persistent, respectively) in children than in adults.69

Complication rates may be lower in select cases where s­ ingle-stage monitoring is feasible. Bansal et al in 2017 reviewed their series of 130 pediatric patients undergoing s­ ingle-stage intraoperative ECoG-guided resections for epilepsy and report a 6.9% complication rate with no permanent deficits.70 Single-stage surgery is less feasible in the MRI-­negative setting; however, the study of Bansal et al suggests it is an option if ancillary data are concordant with PET (or potentially other functional imaging). Undertaking invasive monitoring for MRI-negative epilepsy carries the risk of a failed invasive diagnostic electrophysiology study resulting from inability to localize a focal SOZ. This may be due to absence of seizure during the monitoring period or multifocal onset that was not apparent on noninvasive studies. In a series by Dorward et al in 2011, forty-three children underwent ECoG monitoring for MRI-negative ETLE and 10 (23%) were unable to identify a suitable SOZ for resection or multiple subpial transection.71 A study by Brna et al in 2015 found invasive monitoring to be useful in 87% of cases in their series of 102 children with a combination of MRI-positive and MRI-negative epilepsy involving any lobe.72 One possible explanation for cases of failed ECoG localization may relate to the intrinsic nature of recording from electrodes placed over the pial surface. A large percentage of the cortex is located in the sulcal invaginations and surface electrodes record from the gyral crowns. This is particularly challenging for an EZ in the insula, where ECoG electrode grids are prohibitive. Intraparenchymal depth electrodes can be placed solely or in combination with surface electrodes to maximize the chance of successful localization when concern for insular or depth of sulcal seizure onset is suggested a priori.57

Stereotactic EEG Stereotactic EEG (SEEG) is an alternative to surface electrodes for intracranial monitoring that relies on neuronavigation for the precise placement of intraparenchymal electrodes via a percutaneous procedure. It has the advantage of being less invasive as each electrode only requires a stab skin incision and small twist-drill-hole through which an electrode is passed to a prespecified depth (Fig. 51.3). In addition, a three-dimensional volume sampling can be achieved with SEEG by strategically choosing entry points systematically distributed along the surface of the brain, where each electrode has multiple contacts along its length.73 In this manner, the putative seizure network can be studied in three dimensions as opposed to only the exposed surface, as with ECoG. Understanding the progression of a seizure throughout the complex three-dimensional geometry of the cerebrum may be critical in cases where a two-­ dimensional study may identify the SOZ at the crown of a gyral surface, potentially distant from the true EZ. SEEG was described over 50 years ago by Talairach et al74 and has been used in some areas of Europe for much longer than the recent revival in North America.75 The increased use of SEEG may be related to a greater recognition of surgical management in MRI-negative and extratemporal cases.76 In addition, recent advances in stereotaxy make accurate placement of intraparenchymal electrodes safer and more efficient. Early SEEG methods made use of frame-based ste-

485

486

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

Fig. 51.3  SEEG. Electrode placement is planned preoperatively with an application-specific software suite (a). SEEG electrodes are placed percutaneously in the operating room (b) using an image-guided robotic arm. Postoperative CT (c) is used to rule out hemorrhagic complication and confirm accurate electrode placement. A 3D image reconstruction (d) facilitates identification of percutaneous leads that are labeled with bedside electrophysiology monitoring for localization of the seizure-onset zone.

reotaxy, which has largely been replaced by frameless stereotactic navigation. More recently, robotic navigation and assisted electrode placement has further advanced the SEEG procedure. The ROSA robot (Medtech) and Neuromate (Renishaw) are two such assistive devices, which consist of a robotic arm combined with frameless stereotaxy that navigates to predefined insertion sites to facilitate accurate and efficient electrode placement. A meta-analysis attempted to compare the accuracy of SEEG placement by frameless, frame-based, and robotic techniques, but was unable to provide meaningful comparison due to significant heterogeneity in the reported methods. However, all three techniques had a mean error of less than 3 mm.77 Despite placement of penetrating electrodes without direct visualization, SEEG is a safe procedure that is well tolerated in children.78 To this end, many centers incorporate vascular imaging to avoid blood vessels in planning electrode trajectories. A meta-analysis evaluating the safety of SEEG found a 1.0% rate of hemorrhagic complications (0.18% per electrode) and an ­overall complicate rate of 1.3%.79 In 2017, Bourdillon et al reviewed their experience with insular SEEG, where the candelabra of middle cerebral artery vessels pose a greater risk, and reported no hematomas associated with insular electrodes and a 0.08% per electrode rate in all other electrodes.56 These data represent a very favorable safety profile when compared to procedures requiring open craniotomy.

„„ Seizure Outcomes Although MRI-negative ETLE is particularly challenging, good outcomes are achievable with neurosurgical treatment. A ­ nsari et al reported a pair of meta-analyses in 2010 investigating outcomes for children and adults after surgery for nonlesional ETLE.18 They found seizure freedom in 33.7% and at least worthwhile reduction of seizures (Engel class I, II, or III) in 73.7% of 95 pediatric ETLE cases.18 More recently, several additional case series of MRI-negative ETLE have been published that report seizure freedom rates of 28 to 46%.28,​29,​71 These reported seizure freedom rates in children are generally lower than those for adults with similar MRI-negative ETLE, which was 45.8% by meta-analysis.80 An additional consideration when interpreting outcomes reported in the literature is the proportion of MRI-negative cases that do not reach surgical intervention. A study highlighting this effect reviewed 85 patients with MRI-negative MRE, of which 31 (36%) underwent invasive monitoring and 24 (28%) went on to resective surgery. Of those who underwent resection, nine (38%) achieved excellent outcomes (Engel Classes I–IIA). However, this represents only 29% of those undergoing invasive monitoring and 11% of the total cohort.81 On the one hand, these humbling numbers emphasize the challenges of MRI-negative epilepsy, while on the other hand, these demonstrate the utility of resective surgery in appropriately selected patients.

51  Surgical Management of MRI-Negative Extratemporal Lobe Epilepsy

References 1. Fisher RS, Acevedo C, Arzimanoglou A, et al. ILAE official report: a practical clinical definition of epilepsy. Epilepsia 2014;55(4): 475–482

21. Englot DJ, Breshears JD, Sun PP, Chang EF, Auguste KI. Seizure outcomes after resective surgery for extra-temporal lobe epilepsy in pediatric patients. J Neurosurg Pediatr 2013;12(2):126–133

2. Foldvary-Schaefer N, Unnwongse K. Localizing and lateralizing features of auras and seizures. Epilepsy Behav 2011;20(2):160–166

22. Blume WT, Lüders HO, Mizrahi E, Tassinari C, van Emde Boas W, Engel J Jr. Glossary of descriptive terminology for ictal semiology: report of the ILAE task force on classification and terminology. Epilepsia 2001;42(9):1212–1218

3. Lüders HO, Najm I, Nair D, Widdess-Walsh P, Bingman W. The epileptogenic zone: general principles. Epileptic Disord 2006; 8(2, Suppl 2):S1–S9 4. Goyal M, Bangert BA, Lewin JS, Cohen ML, Robinson S. High-­ resolution MRI enhances identification of lesions amenable to surgical therapy in children with intractable epilepsy. Epilepsia 2004;45(8):954–959 5. Phal PM, Usmanov A, Nesbit GM, et al. Qualitative comparison of 3-T and 1.5-T MRI in the evaluation of epilepsy. AJR Am J ­Roentgenol 2008;191(3):890–895 6. Zijlmans M, de Kort GA, Witkamp TD, et al. 3T versus 1.5T phased-array MRI in the presurgical work-up of patients with partial epilepsy of uncertain focus. J Magn Reson Imaging 2009;30(2):256–262 7. Mellerio C, Labeyrie MA, Chassoux F, et al. 3T MRI improves the detection of transmantle sign in type 2 focal cortical dysplasia. Epilepsia 2014;55(1):117–122 8. Martinez-Rios C, McAndrews MP, Logan W, Krings T, Lee D, Widjaja E. MRI in the evaluation of localization-related epilepsy. J Magn Reson Imaging 2016;44(1):12–22 9. Knake S, Triantafyllou C, Wald LL, et al. 3T phased array MRI improves the presurgical evaluation in focal epilepsies: a prospective study. Neurology 2005;65(7):1026–1031 10. Nguyen DK, Rochette E, Leroux J-M, et al. Value of 3.0 T MR imaging in refractory partial epilepsy and negative 1.5 T MRI. Seizure 2010;19(8):475–478 11. Rubinger L, Chan C, D’Arco F, et al. Change in presurgical diagnostic imaging evaluation affects subsequent pediatric epilepsy surgery outcome. Epilepsia 2016;57(1):32–40 12. Veersema TJ, Ferrier CH, van Eijsden P, et al. Seven tesla MRI improves detection of focal cortical dysplasia in patients with refractory focal epilepsy. Epilepsia Open 2017;2(2):162–171 13. Bast T. Outcome after epilepsy surgery in children with MRI-negative non-idiopathic focal epilepsies. Epileptic Disord 2013;15(2):105–113 14. Berg AT, Vickrey BG, Langfitt JT, et al; Multicenter Study of Epilepsy Surgery. The multicenter study of epilepsy surgery: recruitment and selection for surgery. Epilepsia 2003;44(11):1425–1433 15. Harvey AS, Cross JH, Shinnar S, Mathern GW; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epilepsia 2008;49(1):146–155 16. Bien CG, Szinay M, Wagner J, Clusmann H, Becker AJ, Urbach H. Characteristics and surgical outcomes of patients with refractory magnetic resonance imaging-negative epilepsies. Arch Neurol 2009;66(12):1491–1499 17. Téllez-Zenteno JF, Hernández Ronquillo L, Moien-Afshari F, Wiebe S. Surgical outcomes in lesional and non-lesional epilepsy: a systematic review and meta-analysis. Epilepsy Res 2010;89(2–3):310–318 18. Ansari SF, Maher CO, Tubbs RS, Terry CL, Cohen-Gadol AA. ­Surgery for extratemporal nonlesional epilepsy in children: a meta-analysis. Childs Nerv Syst 2010;26(7):945–951 19. Pomata HB, González R, Bartuluchi M, et al. Extratemporal epilepsy in children: candidate selection and surgical treatment. Childs Nerv Syst 2000;16(12):842–850 20. Englot DJ, Rolston JD, Wang DD, Sun PP, Chang EF, Auguste KI. ­Seizure outcomes after temporal lobectomy in pediatric patients. J Neurosurg Pediatr 2013;12(2):134–141

23. Tufenkjian K, Lüders HO. Seizure semiology: its value and limitations in localizing the epileptogenic zone. J Clin Neurol 2012;8(4):243–250 24. Werhahn KJ. Weakness and focal sensory deficits in the postictal state. Epilepsy Behav 2010;19(2):138–139 25. Todd RB. On the pathology and treatment of convulsive diseases. London Med Gaz 1849;8:822 26. Kellinghaus C, Kotagal P. Lateralizing value of Todd’s palsy in patients with epilepsy. Neurology 2004;62(2):289–291 27. Gallmetzer P, Leutmezer F, Serles W, Assem-Hilger E, Spatt J, Baumgartner C. Postictal paresis in focal epilepsies—incidence, duration, and causes: a video-EEG monitoring study. Neurology 2004;62(12):2160–2164 28. Shi J, Lacuey N, Lhatoo S. Surgical outcome of MRI-negative refractory extratemporal lobe epilepsy. Epilepsy Res 2017;133:103–108 29. Arya R, Leach JL, Horn PS, et al. Clinical factors predict surgical outcomes in pediatric MRI-negative drug-resistant epilepsy. Seizure 2016;41:56–61 30. McGonigal A, Chauvel P. Frontal lobe epilepsy: seizure semiology and presurgical evaluation. Pract Neurol 2004;4(5):260–273 31. Jackson J. Selected Writings of John Hughlings Jackson on Epilepsy and Epileptiform Convulsions. London: Hodder and Stoughton. Arch NeurPsych Volume 1. 1932;27(3):757 32. Bonini F, McGonigal A, Trébuchon A, et al. Frontal lobe seizures: from clinical semiology to localization. Epilepsia 2014;55(2):264–277 33. Stoyke C, Bilgin O, Noachtar S. Video atlas of lateralising and localising seizure phenomena. Epileptic Disord 2011;13(2):113–124 34. Nobili L, Proserpio P, Combi R, et al. Nocturnal frontal lobe epilepsy. Curr Neurol Neurosci Rep 2014;14(2):424 35. Miano S, Peraita-Adrados R. Nocturnal frontal lobe epilepsy is often misdiagnosed as sleep disorders in children: a case series [in Spanish]. Rev Neurol 2013;56(5):257–267 36. Nobili L, Francione S, Mai R, et al. Surgical treatment of drug-­ resistant nocturnal frontal lobe epilepsy. Brain 2007;130(Pt 2):561–573 37. Englot DJ, Wang DD, Rolston JD, Shih TT, Chang EF. Rates and predictors of long-term seizure freedom after frontal lobe epilepsy surgery: a systematic review and meta-analysis. J Neurosurg 2012;116(5):1042–1048 38. Francione S, Liava A, Mai R, et al. Drug-resistant parietal epilepsy: polymorphic ictal semiology does not preclude good post-surgical outcome. Epileptic Disord 2015;17(1):32–46, quiz 46 39. Ye BS, Cho Y-J, Jang SH, Lee MK, Lee BI, Heo K. The localizing and lateralizing value of auras in lesional partial epilepsy patients. Yonsei Med J 2012;53(3):477–485 40. So E, Ryvlin P. MRI-Negative Epilepsy: Evaluation and Surgical Management. Cambridge University Press; 2015 41. Yang P-F, Jia Y-Z, Lin Q, et al. Intractable occipital lobe epilepsy: clinical characteristics, surgical treatment, and a systematic review of the literature. Acta Neurochir (Wien) 2015;157(1):63–75 42. Blume WT, Wiebe S, Tapsell LM. Occipital epilepsy: lateral versus mesial. Brain 2005;128(Pt 5):1209–1225 43. Liava A, Mai R, Cardinale F, et al. Epilepsy surgery in the posterior part of the brain. Epilepsy Behav 2016;64(Pt A):273–282

487

488

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches 44. Ramantani G, Stathi A, Brandt A, et al. Posterior cortex epilepsy surgery in childhood and adolescence: predictors of long-term seizure outcome. Epilepsia 2017;58(3):412–419

63. Fernández S, Donaire A, Serès E, et al. PET/MRI and PET/MRI/ SISCOM coregistration in the presurgical evaluation of refractory focal epilepsy. Epilepsy Res 2015;111:1–9

45. Jehi LE, O’Dwyer R, Najm I, Alexopoulos A, Bingaman W. A longitudinal study of surgical outcome and its d ­ eterminants following posterior cortex epilepsy surgery. Epilepsia 2009; 50(9):2040–2052

64. Castillo EM, Kleineschay T, Baumgartner JE. Contributions of magnetoencephalography to characterizing brain function in pediatric epilepsy: evidences of validity and added value. J Pediatr Epilepsy 2015;4(4):207–215

46. Dalmagro CL, Bianchin MM, Velasco TR, et al. Clinical features of patients with posterior cortex epilepsies and predictors of surgical outcome. Epilepsia 2005;46(9):1442–1449

65. Minassian BA, Otsubo H, Weiss S, Elliott I, Rutka JT, Snead OC III. Magnetoencephalographic localization in pediatric epilepsy surgery: comparison with invasive intracranial electroencephalography. Ann Neurol 1999;46(4):627–633

47. Battaglia D, Chieffo D, Tamburrini G, et al. Posterior resection for childhood lesional epilepsy: neuropsychological evolution. Epilepsy Behav 2012;23(2):131–137 48. Ibrahim GM, Fallah A, Albert GW, et al. Occipital lobe epilepsy in children: characterization, evaluation and surgical outcomes. Epilepsy Res 2012;99(3):335–345 49. Kuzniecky R, Gilliam F, Morawetz R, Faught E, Palmer C, Black L. Occipital lobe developmental malformations and epilepsy: clinical spectrum, treatment, and outcome. Epilepsia 1997;38(2):175–181 50. Liava A, Mai R, Tassi L, et al. Paediatric epilepsy surgery in the posterior cortex: a study of 62 cases. Epileptic Disord 2014;16(2):141–164 51. Sinclair DB, Wheatley M, Snyder T, Gross D, Ahmed N. Posterior resection for childhood epilepsy. Pediatr Neurol ­ 2005;32(4):257–263 52. Harward SC, Chen WC, Rolston JD, Haglund MM, Englot DJ. Seizure outcomes in occipital lobe and posterior quadrant epilepsy surgery: a systematic review and meta-analysis. Neurosurgery 2018;82(3):350–358 53. Isnard J, Guénot M, Ostrowsky K, Sindou M, Mauguière F. The role of the insular cortex in temporal lobe epilepsy. Ann Neurol 2000;48(4):614–623 54. Laoprasert P, Ojemann JG, Handler MH. Insular epilepsy surgery. Epilepsia 2017;58(Suppl 1):35–45 55. Penfield W, Faulk ME Jr. The insula; further observations on its function. Brain 1955;78(4):445–470 56. Bourdillon P, Ryvlin P, Isnard J, et al. Stereotactic electroencephalography is a safe procedure, including for insular implantations. World Neurosurg 2017;99:353–361 57. Weil AG, Fallah A, Lewis EC, Bhatia S. Medically resistant pediatric insular-opercular/perisylvian epilepsy. Part 1: invasive monitoring using the parasagittal transinsular apex depth electrode. J Neurosurg Pediatr 2016;18(5):511–522 58. Perissinotti A, Setoain X, Aparicio J, et al. Clinical role of subtraction ictal SPECT coregistered to MR imaging and ­ 18F-FDG PET in pediatric epilepsy. J Nucl Med 2014;55(7): 1099–1105 59. Rathore C, Dickson JC, Teotónio R, Ell P, Duncan JS. The utility of 18F-fluorodeoxyglucose PET (FDG PET) in epilepsy surgery. ­Epilepsy Res 2014;108(8):1306–1314 60. Salamon N, Kung J, Shaw SJ, et al. FDG-PET/MRI coregistration improves detection of cortical dysplasia in patients with epilepsy. Neurology 2008;71(20):1594–1601 61. Kudr M, Krsek P, Marusic P, et al. SISCOM and FDG-PET in patients with non-lesional extratemporal epilepsy: correlation with intracranial EEG, histology, and seizure outcome. Epileptic Disord 2013;15(1):3–13 62. Kim YH, Kang HC, Kim DS, et al. Neuroimaging in identifying focal cortical dysplasia and prognostic factors in pediatric and adolescent epilepsy surgery. Epilepsia 2011;52(4): 722–727

66. Rozendaal YJ, van Luijtelaar G, Ossenblok PP. Spatiotemporal mapping of interictal epileptiform discharges in human absence epilepsy: a MEG study. Epilepsy Res 2016;119:67–76 67. Englot DJ, Nagarajan SS, Imber BS, et al. Epileptogenic zone localization using magnetoencephalography predicts seizure freedom in epilepsy surgery. Epilepsia 2015;56(6):949–958 68. Johnston JM Jr, Mangano FT, Ojemann JG, Park TS, Trevathan E, Smyth MD. Complications of invasive subdural electrode monitoring at St. Louis Children’s Hospital, 1994–2005. J Neurosurg 2006;105(5, Suppl):343–347 69. Hader WJ, Téllez-Zenteno J, Metcalfe A, et al. Complications of epilepsy surgery: a systematic review of focal surgical resections and invasive EEG monitoring. Epilepsia 2013;54(5):840–847 70. Bansal S, Kim AJ, Berg AT, et al. Seizure outcomes in children following electrocorticography-guided single-stage surgical resection. Pediatr Neurol 2017;71:35–42 71. Dorward IG, Titus JB, Limbrick DD, Johnston JM, Bertrand ME, Smyth MD. Extratemporal, nonlesional epilepsy in children: postsurgical clinical and neurocognitive outcomes. J Neurosurg Pediatr 2011;7(2):179–188 72. Brna P, Duchowny M, Resnick T, Dunoyer C, Bhatia S, Jayakar P. The diagnostic utility of intracranial EEG monitoring for epilepsy surgery in children. Epilepsia 2015;56(7):1065–1070 73. Kalamangalam GP, Tandon N. Stereo-EEG implantation strategy. J Clin Neurophysiol 2016;33(6):483–489 74. Talairach J, Bancaud J, Bonis A, Tournoux P, Szikla G, Morel P. Functional stereotaxic investigations in epilepsy. Methodological remarks concerning a case. [in French]. Rev Neurol (Paris) 1961;105:119–130 75. Bulacio JC, Chauvel P, McGonigal A. Stereoelectroencephalography: Interpretation. J Clin Neurophysiol 2016;33(6):503–510 76. Jehi L, Friedman D, Carlson C, et al. The evolution of epilepsy surgery between 1991 and 2011 in nine major epilepsy centers across the United States, Germany, and Australia. Epilepsia 2015;56(10):1526–1533 77. Vakharia VN, Sparks R, O’Keeffe AG, et al. Accuracy of intracranial electrode placement for stereoencephalography: a systematic review and meta-analysis. Epilepsia 2017;58(6):921–932 78. Taussig D, Chipaux M, Lebas A, et al. Stereo-electroencephalography (SEEG) in 65 children: an effective and safe diagnostic method for pre-surgical diagnosis, independent of age. Epileptic Disord 2014;16(3):280–295 79. Mullin JP, Shriver M, Alomar S, et al. Is SEEG safe? A systematic review and meta-analysis of stereo-electroencephalography-­ related complications. Epilepsia 2016;57(3):386–401 80. Ansari SF, Tubbs RS, Terry CL, Cohen-Gadol AA. Surgery for extratemporal nonlesional epilepsy in adults: an outcome meta-­ analysis. Acta Neurochir (Wien) 2010;152(8):1299–1305 81. Noe K, Sulc V, Wong-Kisiel L, et al. Long-term outcomes after nonlesional extratemporal lobe epilepsy surgery. JAMA Neurol 2013;70(8):1003–1008

52

  Surgical Management of Hypothalamic Hamartomas Neena I. Marupudi and Sandeep Mittal

Summary Hypothalamic hamartomas (HHs) are rare and aberrant nodules of developmental brain tissue that can result in drug-­ resistant epilepsy, classically presenting with isolated fits of gelastic seizures, dacrystic seizures, or other refractory seizure types. Although these lesions do not have neoplastic features, the tissue develops in a disorganized fashion and grows at the same rate as the derivative tissue. Keywords:  Gelastic seizures, epilepsy syndrome, precocious puberty, intrinsic epileptogenesis, microsurgical resection, transcallosal approach, endoscopic disconnection, stereotactic radiosurgery, laser ablation

„„ Anatomic Features The hypothalamus is an integrative basal forebrain structure made of gray matter and consisting of symmetric halves divided by the third ventricle. It coordinates various autonomic, somatic, endocrine, and behavioral activities through its abundant reciprocal afferent and efferent connections. The hypothalamus is notably associated with laughter, a behavioral manifestation of mirth, as well as emotional conditions such as sadness.1,​2,​3 The boundaries of the hypothalamus are defined by the hypothalamic sulcus superiorly, lamina terminalis anteriorly, posteriorly by a line between the caudal aspect of the mammillary body and posterior commissure, and ventrally by the tuber cinereum (gray matter protuberance along the floor of the third ventricle).2 The lateral borders of the hypothalamus are roughly defined by the optic tract, internal capsule, cerebral peduncle, and subthalamus. Hypothalamic hamartoma (HH) is a congenital heterotopic lesion of hypothalamic neurons and glial cells.4 The abnormal tissue is attached by a pedicle to the hypothalamus, lying between the tuber cinereum and the mammillary bodies and extending into the basal cistern. Incidental and asymptomatic hamartomas may be identified in up to 20% of autopsies. ­Disruption of hypothalamic function is rare, but results when these lesions enlarge and compress adjacent tissue. Sessile lesions are associated with epilepsy and closely connected to the mammillary bodies.5,​6 They may be associated with precocious puberty if in contact with the tuber cinereum or the infundibulum.7 Anatomically, HHs can be classified into two subtypes: sessile and pedunculated.8 Sessile, or intrahypothalamic, hamartomas

have a partial or complete base of attachment within the third ventricle.9 They vary in size and commonly distort surrounding neural structures, including the fornix and mammillary bodies.10 Although these lesions do not typically present with precocious puberty initially, up to 40% of patients with sessile HHs can develop central precocious puberty at some point in the disease course. On the other hand, a strong association has been shown between sessile HHs and neurological manifestation such as gelastic seizures.6,​9,​11,​12,​13 On the other hand, ­pedunculated, or parahypothalamic, hamartomas are attached only to the floor of the third ventricle or suspended from it by a peduncle.9 Pedunculated lesions are less commonly associated with childhood epilepsy syndromes or severe neurodevelopmental disorders; instead, they usually present with central precocious puberty in early childhood.

„„ Molecular and Genetic Characteristics HHs can occur as isolated lesions, in association with other brain lesions, or be present as part of a genetic syndrome. They can be a key feature of several developmental disorders, including Pallister-Hall syndrome, a pleiotropic autosomal dominant syndrome resulting from GLI3 gene mutations on chromosome 7p13.14 The condition is classically characterized by the presence of a HH and a spectrum of multiorgan malformations, including central postaxial polydactyl, pituitary hypoplasia, bifid epiglottis, dysplastic nails, and imperforate anus. Other features include cardiac anomalies, renal abnormalities, and mental retardation. The GLI proteins regulate target gene expression by acting downstream of the sonic hedgehog (SHH) signaling pathway, which is critical in directing dorsoventral patterning for central nervous system development.15,​16 In children with P ­ allister-Hall syndrome, frameshift mutations in the GLI3 gene result in the formation of a truncated protein that is functionally identical to the shorter, processed, repressor form of the GLI3 protein.17 Thus, the pathogenetic mutations in patients with ­Pallister-Hall syndrome eliminate the capacity of SHH to switch GLI3 between the repressor and activator state. Somatic GLI3 mutations can lead to spontaneous nonsyndromic HHs and gelastic seizures.18,​19 FOXC1 (chromosome 6) is another potential gene implicated in sporadic HHs; like GLI3, it expresses a DNA-binding transcription factor.20

490

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches

„„ Clinical Manifestations Epilepsy The pathognomonic presenting seizure type, gelastic seizures, is characterized by spontaneous mirthless laughter. Patients with epilepsy classically start to present with laughing seizures during the first years of life, often in the neonatal period. Gelastic seizures were first described by Gascon and Lombroso in 1971 as “gelastic attacks,” characterized by repeated short-lasting seizures with initial emotionless laughter or grimacing.21 Older children and adults may report unpleasant sensations or epigastric discomfort or simply a pressure to laugh.22 Dacrystic, or “crying,” seizures may also be seen. In addition to ictal laughter or crying, other seizure types, often more disabling, develop later in the disease course.23,​24 In addition to the classically described gelastic seizures, patients with HHs have other seizure types, including generalized tonic-clonic seizures, complex partial seizures, drop attacks, and atypical absences. In 85% of patients with HHs, gelastic seizures were the first observed seizure type.25 On the other hand, more than 75% of patients have other intractable seizure types, including complex partial seizures with or without secondary generalization in 35.5%, “falling” seizures in 33.3%, tonic seizures in 17.7%, and tonic-clonic seizures in 15.1%.26 Late-onset epilepsy associated with HHs may have the tendency for a milder epileptic syndrome, while gelastic seizures tend to be the dominant component of the semiology in children.27 On the other hand, gelastic seizures are not necessarily specific to or pathognomonic for HHs.25 They have been described in patients with frontal or temporal lobe epilepsies and in patients with a variety of pathologies, including cortical dysplasia, tuberous sclerosis, pituitary tumors, gliomas, meningiomas, and basilar artery aneurysms.23,​26,​28,​29,​30,​31,​32,​33 While seizures are the most common clinical presenting feature (61%), a significant proportion of patients are also found to have precocious puberty (66%), and 25% present with both.34 In most cases, continued seizures and drug resistance lead to epileptic encephalopathy, various generalized seizures, drop attacks, cognitive decline, and psychiatric comorbidities.9,​24,​30,​34,​35,​36 Although electrographic findings, even during a gelastic or dacrystic episode, may be normal, electroencephalography (EEG) is still recommended as subtle abnormalities may be apparent.26 Over time, changes in the EEG such as diffuse attenuation may suggest developmental impact on the individual. Seizure semiology often suggests the involvement of temporal or frontal lobe structures, supporting secondary ­epileptogenesis in these patients.10,​24 Progression to catastrophic symptomatic generalized epilepsy or epileptic encephalopathy is ­associated with significant cognitive impairment and b ­ ehavioral disorders, sometime with autistic features.24,​37

Precocious Puberty Central precocious puberty is defined as the onset of puberty before the age of 8 years in females and 9 years in males.38 Pedunculated HHs, which occur below the third ventricle, are commonly associated with precocious puberty and rarely present with seizures. Central precocious puberty associated with pedunculated hamartomas occurs at a significantly earlier age than idiopathic central precocious puberty, often occurring before 2 years of age in 80% of such cases.39 HHs causing

precocious puberty tend to be larger and maintain contact with the tuber cinereum or infundibulum than the ones not associated with precocious puberty.7 While the precise mechanism of how HHs cause central precocious puberty remains incompletely understood, HHs have been shown to express gonadotropin-releasing hormone (GnRH) and transforming ­ growth factor alpha, which stimulates the release of GnRH.7

Behavioral, Cognitive, and Psychiatric Disorders Patients with pharmacoresistant seizures from HHs often progress to a severe childhood epilepsy syndrome and epileptic encephalopathy characterized by profound behavioral difficulties, hyperactivity, rage, and aggressive behavior.11,​36,​40,​41 Delinquency and aggressive outbursts seen with HH-­associated ictal behavior can sometimes be misdiagnosed as a primary psychiatric disorder.42 Nevertheless, children with HH and seizures have a significantly higher incidence of psychiatric comorbidities, such as oppositional defiant disorder (83.3%), attention-deficit hyperactivity disorder (75%), conduct disorder (33.3%), and affective disorders (17.6%).35 A high prevalence of major depressive disorder and social anxiety disorder also make up the gamut of comorbid psychiatric disorders.43 With exacerbations and progression of the HH-related seizure disorder, a progressive and parallel decline in cognitive function is observed in these patients.24,​28,​30,​44,​45 Cognitive performance testing has demonstrated impairment in more than 50% of patients with gelastic seizures and HHs.36 In HH patients with refractory epilepsy, the number of antiepileptic drugs and the neuroanatomical features of the HH lesion correlate with impairment on intellectual test performance.46 While cognitive outcomes may be a research measure for investigating the efficacy of surgical treatments, health-related quality-of-life ­ measures may be equally important in optimizing treatment recommendations. HH patients not only have a significantly lower score in School Function Assessment (SFA) tests than children with benign epilepsy, but comorbid psychomotor retardation was predictive of lower quality of life.47

„„ Neuroimaging Findings For diagnosis confirmation and potential surgical planning, adequate imaging must be obtained. A high-resolution brain magnetic resonance imaging (MRI) protocol with contrast would demonstrate a small, nonenhancing lesion adjacent to the hypothalamus. Sequences to aid in the diagnosis include an axial/coronal T2, axial/coronal/sagittal T1, and a spin echo postcontrast. Volumetric fluid-attenuated inversion recovery (FLAIR) with sequences reformatted for visualization in three orthogonal planes (axial, coronal, and sagittal) is beneficial.5

Computed Tomography On computed tomographic (CT) scans, HHs appear as isodense, nonenhancing masses in the interpeduncular and/or suprasellar cistern.48 Although, these lesions are typically small, they may partially or completely obliterate the suprasellar cistern or the anterior portion of the third ventricle.48 Before the popularity of and increased access to high-resolution MRI scans,

52  Surgical Management of Hypothalamic Hamartomas CT c­isternography with metrizamide, a nonionic radiopaque contrast agent, could be used to improve diagnostic yield for smaller occult hypothalamic-hypophyseal lesions; the size of the mass and its relationship with surrounding neurovascular structures could be better defined.49

Magnetic Resonance Imaging and Magnetic Resonance Spectroscopy With the advent of high-resolution MRI, all other diagnostic imaging modalities have fallen out of favor. MRI has dramatically improved surgical management, particularly aiding in minimally invasive and stereotactic treatment approaches. HHs in patients with refractory epilepsy were found to be typically hyperintense on T2-weighted images (93%), while 74% of these lesions were found to be hypointense on T1-weighted images.10 Contrast enhancement is rarely seen in these lesions, as they do not typically disrupt the blood–brain barrier. Magnetic resonance spectroscopy (MRS) can be used to detect neuronal dysfunction in the temporal lobe and within the hamartomas of patients with HHs and gelastic seizures.50 The NAA/Cr ratio is significantly decreased in patients with hamartomas compared with hypothalamic healthy individuals.50 Significant reduction in NAA and an increase in myoinositol (mI) content within the hamartoma are reported.10 In addition to the decrease in NAA/Cr, an increase in mI/Cr ratios is seen in the HH compared to normal gray matter and amygdala; Cho/Cr ratios are also elevated.51 Overall, MRS studies suggest that HHs have decreased neuronal density and relative gliosis compared with normal gray matter.

„„ Electrographic Findings Interictal and ictal EEG can be normal or nonspecifically abnormal in patients with gelastic seizures. With progression to generalized seizures, increased in spike–wave activity is seen with enhanced bilateral wave synchrony.26 With scalp EEG testing, certain

seizures in patients with HHs can resemble complex partial seizures that originate from the frontal or temporal lobe.52 Overall, high variability is noted on interictal EEG between HH patients, with findings including focal, multifocal, and generalized spike and waves.53 On ictal surface EEG recordings, even with intracranial grid electrodes, seizures can appear to arise from the frontal or temporal cortex.53 It is well known that neocortical resection is largely ineffective in abolishing HH-related gelastic seizures. While scalp EEG frequently fails to demonstrate the origin of gelastic seizures in the HH patients, depth electrode placement and recording directly from within the HH has provided evidence that gelastic seizures and associated ictal events originate directly from the hamartoma itself.54​—​58 Complex partial seizures and generalized seizures seen in this patient population are often indicative of discharge spread from the HH to the frontal and temporal regions. Although these seizures are not directly associated with discharged from the HH itself, lateralization of motor features of seizures and of interictal EEG often corresponds to the side of attachment of the hamartoma.26

„„ Surgical Decision Making With progressive understanding of the intrinsic epileptogenesis from HHs, early surgical intervention has been increasingly advocated. With untreated HHs, the risk of progressive cognitive and behavioral problems remains high, especially with increases in seizure frequency and development of multiple seizure types. In circumstances where behavior and neuropsychological assessment is concerning, early referral for surgical options is beneficial. A precise knowledge of the relationship of the HH to the surrounding neurovascular structures is critical to choosing the best surgical approach. Several authors have classified HHs based on various morphological characteristics (Fig. 52.1). Delalande and Fohlen categorized HHs based on optimal surgical approaches (types I–IV).59 Type I hamartomas are horizontally

Fig. 52.1  Topological classification of hypothalamic hamartomas based on MRI findings.

491

492

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches attached to the hypothalamus and can be surgically accessed via the pterional approach. Type II hamartomas are intraventricularly located, attached to the side of the hypothalamus, and can be accessed endoscopically. Type III lesions are a blend of types I and II and, therefore, are treated with an e ­ ndoscopic approach for the intraventricular component and pterional approach for the extraventricular portion. Giant hamartomas are classified as type IV, and do not have a specific ­surgical approach recommendation. Another classification by Régis et al suggests guidelines for identifying lesions amenable to Gamma Knife radiosurgery.60,​61 In this classification, type I (intrahypothalamic lesions) and type II (intraventricular lesions) are amenable to stereotactic ­radiosurgery. Type III lesions arise from the floor of the third ventricle, and Gamma Knife radiosurgery is the recommended treatment. Type IV lesions are sessile tumors of the interpeduncular fossa and best treated with combination of surgery and radiosurgery. Type V HHs are pedunculated lesions that are attached to the hypothalamus inferiorly via a stalk; these are best treated with stereotactic radiosurgery or disconnective surgery. Lastly, type VI lesions are giant hamartomas requiring open surgical resections but may benefit from Gamma Knife radiosurgery following subtotal surgical resection. Preoperative workup includes neuropsychological, psychiatric, endocrinologic, visual field, and visual acuity examinations.

„„ Surgical Techniques Surgical treatment of HHs has evolved considerably over the past 50 years. In 1967, Northfield and Russell performed the first successful removal of a HH causing central precocious puberty.62 While several other investigators described microsurgical excision of HHs, surgical treatment of these lesions expanded significantly with the advent of high-resolution MRI being used to identify the lesions.63,​64,​65,​66 Surgical technique in the management of HHs and associated gelastic seizures include craniotomy with microsurgery, endoscopic disconnection, stereotactic radiosurgery, radiofrequency thermocoagulation, and MRI-guided laser ablation. A variety of microsurgical approaches have been described to successfully treat drug-resistant epilepsy in patients with HHs: transcallosal, anterior interforniceal, or pterional approaches for surgical resection.67,​68 Endoscopic techniques described include transcortical and transventricular approaches for disconnection and resection.59,​69 Minimally invasive procedures include radiosurgery, brachytherapy,70–​75 and thulium laser for coagulation69 and stimulation of the ­mamillothalamic tract.71 More recent technological advancement in minimally invasive techniques has further improved outcomes by minimizing surgical morbidity. The surgical approach is selected and tailored based on the patient’s age, size of the hamartoma, its anatomical relationship with surrounding neurovascular structures and the hypothalamic, and the surgeon’s comfort and expertise with specific techniques.

Microsurgical Approaches Frontotemporal Approach Microsurgical approaches for resection of HHs can be divided into approaches that access the lesions from above and

those that approach from below.76 The natural first step in the evolution of microsurgical techniques began with the use of ­frontotemporal and pterional approaches in initial clinical series.63,​77 Depending on the size and specific location of the lesion, the addition of subfrontal, subtemporal, transsylvian, and other skull base approaches can facilitate the complete excision of a hypothalamic lesion.78,​79 Frontotemporal approaches certainly support a direct route with the shortest distance. However, the approach poses other challenges, with the trajectory requiring passage between the internal carotid artery, optic nerve and chiasm, third cranial nerve, and infundibulum to access the third ventricle and the hamartoma. From this approach, delineating the margins of a hamartoma can be challenging, particularly if it is widely involving the hypothalamus and mammillary bodies. Complications reported with the pterional approach include transient and permanent third nerve palsies, visual field deficits, thalamocapsular infarcts, diabetes insipidus, and hyperphagia.67,​80 The orbitozygomatic approach benefits HHs attached below the third ventricle; on the other hand, most other patients with HHs would benefit from a superior approach or endoscopic approach.80 The pterional approach is best suited for targeting pedunculated HHs that cause central precocious puberty. Even a partial resection has been shown to relieve endocrinologic disturbance.81Pedunculated HHs are more amenable to complete resection that sessile ones. Palmini et al reported a 90 to 100% reduction in seizure frequency in a series of 13 patients undergoing a pterional approach for HH resection. All patients demonstrated improved behavior and cognition postsurgery. Of patients undergoing a frontotemporal approach for resection of a HH, 23 to 40% become seizure free.67,​80 A significant reduction in seizure frequency, as much as more than 90%, is reported in the remaining patients.67,​80 Despite the excellent seizure control outcomes, the frontotemporal approach is associated with major neurological complications, resulting in the need for alternative surgical approaches and strategies to limit surgery-related morbidity while achieving excellent seizure freedom rates.

Transcallosal Approach The transcallosal approach to remove HHs through the third ventricle offers a great intraventricular view of the hamartoma and also allows for adequate debulking and disconnection of the hamartoma (Fig. 52.2). This approach reduces risk of injury to structures such as the mammillary bodies, optic chiasm, and pituitary stalk. The original description of this approach was an anterior transcallosal, transseptal, interforniceal approach through the third ventricle for removal of HHs.68 Although there is inherent risk of forniceal injury that could result in short-term memory problems, the incidence of cerebral infarction and oculomotor nerve palsy is significantly reduced as the approach avoids manipulation of the neurovascular structures in the suprasellar cistern and interpeduncular fossa. The approach consists of a 1.5- to 2-cm postgenual callosotomy, midline transseptal dissection, and splitting of the fornix.68,​82 The third ventricle is then visualized and entered to remove or disconnect the HH. Alternatively, using a subchoroidal approach, as opposed to interforniceal approach, can reduce risk of postsurgical memory deficits.83 Via the transcallosal approach, seizure freedom can be achieved in more than 50% of patients and at least 90% decrease in frequency of seizures is seen in 24%.82 Younger patients demonstrate steady improvements in

52  Surgical Management of Hypothalamic Hamartomas

Fig. 52.2  (a–c) Transcallosal transseptal interforniceal approach for microsurgical resection of a small hypothalamic hamartoma.

multidomain cognition after a transcallosal procedure.84 Older adolescents and adults have not demonstrated such improvements, and, therefore, may be better candidates for stereotactic radiosurgery or other minimally invasive procedures.84 These older patients may also have increased risk of forniceal injury because the leaves of the septum pellucidum are not as easily separated as in younger patients.85 A direct relationship exists between the extent of HH excision and seizure outcome.85 Furthermore, an inverse correlation was also found between the duration of epilepsy and probability of seizure freedom.85 No single surgical technique is ideal for all HH resections. Some large lesions may require a multistep approach, including endoscopic biopsy, disconnection, and/or microsurgical resection. The approaches must be selected based on the HH location, type (sessile vs. pedunculated), and extensions or attachments of the hamartoma. Among the described microsurgical techniques, the transcallosal, interforniceal approach is the preferred approach at many epilepsy centers for best seizure freedom outcomes.

Endoscopic Surgery for Hypothalamic Hamartomas The first endoscopic procedure described for treatment of a hamartoma was for a biopsy followed by endoscopic laser coagulation.86 Not all patients are good candidates for an endoscopic approach for HH resection. The endoscopic transventricular approach is best for small hamartomas with a ­unilateral attachment to the hypothalamic wall.87 In addition, the presence of a space between the bottom of the hamartoma and pial surface of the interpeduncular cistern is ideal for the endoscopic approach.87 Having at least 6 mm of working space between the top of the HH and the roof of the third ventricle can be helpful.

Disconnection Procedures for Hypothalamic Hamartomas Isolating an intrinsically epileptogenic lesion by disconnection alone without excision can also result in good seizure outcomes.59 The rate of seizure freedom is about 50% after open or endoscopic disconnection.59,​88 Intraoperatively, a depth electrode can be placed in the hamartoma prior to and after endoscopic disconnection to assess for improvement in epileptic

activity via simple disconnection of hamartoma.56,​88 Smaller hamartomas have a better overall outcome with disconnection than larger more pedunculated lesions.56,​88 Successful disconnection depends on a number of factors, including the plane and extent of attachment of the hamartoma to the hypothalamus.89 Completeness of the disconnection correlates with improved rates of seizure freedom.

Staged Approaches and Repeat Surgery In patients with catastrophic intractable epilepsy as a result of a HH, neurosurgical treatment offers clear benefits. If after a surgical intervention, disabling seizures persist or behavioral and cognitive status continues to decline, further medical management options are not only limited but often futile. Reevaluation of patients with suboptimal initial surgical results is necessary to determine if additional interventions may be of benefit. A significant reduction in seizures can be achieved with reoperation with minimal additional morbidity.90

Stereotactic Gamma Knife Radiosurgery Selectively targeting HHs with radiosurgery to treat gelastic seizures has been achieved with brachytherapy, LINAC-based radiosurgery, and Gamma Knife radiosurgery. The vast majority of centers treated HHs with radiosurgery utilize stereotactic Gamma Knife radiosurgery. The first reported patient with a HH successfully treated by Gamma Knife radiosurgery was a 25-year-old man with a small hamartoma, long-standing medically refractory gelastic, and tonic-clonic seizures. He was treated with a dose of 18 Gy to the 50% isodose line, with complete seizure freedom from 3 months postprocedure and complete disappearance of the hamartoma at 12 months.91 The median prescribed marginal dose for Gamma Knife treatment of HHs is 17 Gy (range: 13–26 Gy).60 Beam blocking strategies are utilized to reduce the dose of radiation delivered to surrounding critical structures (e.g., mammillary bodies, tuber cinereum, fornix, infundibulum, and optic nerves and chiasm).60 Many studies have demonstrated that Gamma Knife radiosurgery is an effective treatment modality for select patients with HH-­associated epilepsy.60,​80,​88,​92,​93,​94,​95,​96 Prospective trials with stereotactic radiosurgery suggest an association of outcome with lesion size and topology, that is, small, intrahypothalamic lesions have the

493

494

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches best outcome.73 There appears to be a dose-dependent response in which improved seizure control rates are achieved with marginal doses greater than 16 Gy.97 One of the main disadvantages is the slow clinical response following radiosurgery. Beyond seizure reduction, the improvements of psychiatric and cognitive comorbidities along with better school performance and social functioning are major benefits of treatment with Gamma Knife radiosurgery, even groups with frequently catastrophic epilepsy.73 Although the outcomes can be comparable to surgical techniques, repeated treatments may be needed to partial results in up to 60% patients.73 On the other hand, transient or permanent amnesic or endocrinologic complications from surgical techniques (e.g., transcallosal, anterior interforniceal, endoscopy, and brachytherapy) are not sequelae of Gamma Knife treatment. Treatment of HHs with Gamma Knife radiosurgery is relatively new and longer follow-up studies are underway. In ­addition, other minimally invasive techniques, such as laser interstitial ablation, may eliminate the need for repeat procedures and become standard of therapy as more data are collected evaluating the long-term outcomes from the therapy.

Stereotactic Laser Ablation Stereotactic radiofrequency thermocoagulation was previously used for lesioning of HHs as an alternative to direct microsurgical approaches.98,​99 However, because the extent of the ablation was more estimated and could not be monitored real-time, the procedure carried risk of injury to adjacent critical structures (e.g., hypothalamus, vessels, visual pathways, mammillary bodies, ­fornix). This technique has largely been replaced by a more controlled MRI-guided stereotactic laser-based ablation technology. On the other hand, stereotactic laser ablation is being investigated and utilized in the management of all types of HHs, as

it can be used as a disconnective and ablative modality, encompassing the best of both worlds. MRI-guided laser interstitial thermal ablation involves the stereotactic cannulation of typically deep-seated structures to produce a precise lesion of soft tissue. Temperature changes in lesional and surrounding tissue are monitored real time, minimizing injury to surrounding brain. Several case series have reported seizure freedom, even in cases that have previously failed other surgical ­treatments.100,​101,​102,​103 Of the 25 patients reported in these case series, 21 patients achieved seizure freedom, although the duration of follow-up is relatively short with the recent advent of the technology. MRI-guided stereotactic laser ablation therapy is becoming the treatment of choice for HHs because of their deep location, proximity to critical neural and vascular structures, and because it allows for a more precise and controlled targeting for ablating the lesion.

„„ Conclusion The management and treatment of HHs and associated epilepsy has evolved from advanced open microsurgical techniques to minimally invasive and stereotactic techniques. All techniques, if appropriately selected based on lesion characteristics, have potential to achieve seizure freedom, significant reduction in seizure frequency, or even reversing epileptic encephalopathy. Although the debate continues as to whether complete resection versus disconnection is the most beneficial to the patient, it is imminently evident that early recognition of the hamartoma, a relatively small lesion size, shorter duration of seizures, younger age, and gross-­total resection are all predictors for achieving seizure freedom. Beyond seizure control, successful treatment of HHs improves neurocognitive and behavioral ­outcomes.

References 1. Arroyo S, Lesser RP, Gordon B, et al. Mirth, laughter and gelastic seizures. Brain 1993;116(Pt 4):757–780 2. Freeman JL. The anatomy and embryology of the hypothalamus in relation to hypothalamic hamartomas. Epileptic Disord 2003;5(4):177–186 3. Shumake J, Edwards E, Gonzalez-Lima F. Hypermetabolism of paraventricular hypothalamus in the congenitally helpless rat. Neurosci Lett 2001;311(1):45–48 4. Paillas JE, Roger J, Toga M, et al. [Hamartoma of the hypothalamus. Clinical, radiological and histological study. Results of excision] Rev Neurol (Paris) 1969;120(3):177–194 5. Parvizi J, Le S, Foster BL, et al. Gelastic epilepsy and hypothalamic hamartomas: neuroanatomical analysis of brain lesions in 100 patients. Brain 2011;134(Pt 10):2960–2968 6. Valdueza JM, Cristante L, Dammann O, et al. Hypothalamic hamartomas: with special reference to gelastic epilepsy and surgery. Neurosurgery 1994;34(6):949–958, discussion 958 7. Chan YM, Fenoglio-Simeone KA, Paraschos S, et al. Central precocious puberty due to hypothalamic hamartomas correlates with anatomic features but not with expression of GnRH, TGFalpha, or KISS1. Horm Res Paediatr 2010;73(5):312–319

10. Freeman JL, Coleman LT, Wellard RM, et al. MR imaging and spectroscopic study of epileptogenic hypothalamic hamartomas: analysis of 72 cases. AJNR Am J Neuroradiol 2004;25(3):450–462 11. Arita K, Ikawa F, Kurisu K, et al. The relationship between magnetic resonance imaging findings and clinical manifestations of hypothalamic hamartoma. J Neurosurg 1999;91(2):212–220 12. Boyko OB, Curnes JT, Oakes WJ, Burger PC. Hamartomas of the tuber cinereum: CT, MR, and pathologic findings. AJNR Am J Neuroradiol 1991;12(2):309–314 13. Jung H, Neumaier Probst E, Hauffa BP, Partsch CJ, Dammann O. Association of morphological characteristics with precocious puberty and/or gelastic seizures in hypothalamic hamartoma. J Clin Endocrinol Metab 2003;88(10):4590–4595 14. Démurger F, Ichkou A, Mougou-Zerelli S, et al. New insights into genotype-phenotype correlation for GLI3 mutations. Eur J Hum Genet 2015;23(1):92–102 15. Bertrand N, Dahmane N. Sonic hedgehog signaling in forebrain development and its interactions with pathways that modify its effects. Trends Cell Biol 2006;16(11):597–605 16. Pleasure SJ, Guerrini R. Hypothalamic hamartomas and hedgehogs: not a laughing matter. Neurology 2008;70(8):588–589

8. Mittal S, Mittal M, Montes JL, Farmer JP, Andermann F. Hypothalamic hamartomas. Part 1. Clinical, neuroimaging, and neurophysiological characteristics. Neurosurg Focus 2013;34(6):E6

17. Lai K, Kaspar BK, Gage FH, Schaffer DV. Sonic hedgehog regulates adult neural progenitor proliferation in vitro and in vivo. Nat Neurosci 2003;6(1):21–27

9. Kerrigan JF, Ng YT, Chung S, Rekate HL. The hypothalamic hamartoma: a model of subcortical epileptogenesis and encephalopathy. Semin Pediatr Neurol 2005;12(2):119–131

18. Craig DW, Itty A, Panganiban C, et al. Identification of somatic chromosomal abnormalities in hypothalamic hamartoma tissue at the GLI3 locus. Am J Hum Genet 2008;82(2):366–374

52  Surgical Management of Hypothalamic Hamartomas 19. Wallace RH, Freeman JL, Shouri MR, et al. Somatic mutations in GLI3 can cause hypothalamic hamartoma and gelastic seizures. Neurology 2008;70(8):653–655

41. Waldau B, McLendon RE, Fuchs HE, George TM, Grant GA. Few isolated neurons in hypothalamic hamartomas may cause gelastic seizures. Pediatr Neurosurg 2009;45(3):225–229

20. Kerrigan JF, Kruer MC, Corneveaux J, et al. Chromosomal abnormality at 6p25.1–25.3 identifies a susceptibility locus for hypothalamic hamartoma associated with epilepsy. Epilepsy Res 2007;75(1):70–73

42. Al-Hail HJ, Sokrab TE, Al-Moslamani NJ, Miyares FR. Hypothalamic hamartoma presenting with gelastic seizures, generalized convulsions, and ictal psychosis. Neurosciences (Riyadh) 2010;15(1):43–45

21. Gascon GG, Lombroso CT. Epileptic (gelastic) laughter. Epilepsia 1971;12(1):63–76

43. Ali S, Moriarty J, Mullatti N, David A. Psychiatric comorbidity in adult patients with hypothalamic hamartoma. Epilepsy Behav 2006;9(1):111–118

22. Helen Cross J, Spoudeas H. Medical management and antiepileptic drugs in hypothalamic hamartoma. Epilepsia 2017;58 (Suppl 2):16–21 23. Castro LH, Ferreira LK, Teles LR, et al. Epilepsy syndromes associated with hypothalamic hamartomas. Seizure 2007;16(1):50–58 24. Mullatti N, Selway R, Nashef L, et al. The clinical spectrum of epilepsy in children and adults with hypothalamic hamartoma. Epilepsia 2003;44(10):1310–1319 25. Tassinari CA, Riguzzi P, Rizzi R. Gelastic seizures. In: Tuxhorn I, Holthausen H, Boenigk H, eds. Paediatric Epilepsy Syndromes and Their Surgical Treatment. London: John Libbey; 1997 26. Harvey AS, Freeman JL. Epilepsy in hypothalamic hamartoma: clinical and EEG features. Semin Pediatr Neurol 2007; 14(2):60–64 27. Oehl B, Brandt A, Fauser S, Bast T, Trippel M, Schulze-Bonhage A. Semiologic aspects of epileptic seizures in 31 patients with hypothalamic hamartoma. Epilepsia 2010;51(10):2116–2123 28. Arzimanoglou AA, Hirsch E, Aicardi J. Hypothalamic hamartoma and epilepsy in children: illustrative cases of possible evolutions. Epileptic Disord 2003;5(4):187–199 29. Cercy SP, Kuluva JE. Gelastic epilepsy and dysprosodia in a case of late-onset right frontal seizures. Epilepsy Behav 2009;16(2):360–365 30. Frattali CM, Liow K, Craig GH, et al. Cognitive deficits in children with gelastic seizures and hypothalamic hamartoma. Neurology 2001;57(1):43–46 31. Kovac S, Deppe M, Mohammadi S, et al. Gelastic seizures: a case of lateral frontal lobe epilepsy and review of the literature. Epilepsy Behav 2009;15(2):249–253 32. Leal AJ, Monteiro JP, Secca MF, Jordão C. Functional brain mapping of ictal activity in gelastic epilepsy associated with hypothalamic hamartoma: a case report. Epilepsia 2009;50(6):1624–1631 33. Oehl B, Biethahn S, Schulze-Bonhage A. Mirthful gelastic seizures with ictal involvement of temporobasal regions. Epileptic Disord 2009;11(1):82–86 34. Nguyen D, Singh S, Zaatreh M, et al. Hypothalamic hamartomas: seven cases and review of the literature. Epilepsy Behav 2003;4(3):246–258 35. Weissenberger AA, Dell ML, Liow K, et al. Aggression and psychiatric comorbidity in children with hypothalamic hamartomas and their unaffected siblings. J Am Acad Child Adolesc Psychiatry 2001;40(6):696–703 36. Quiske A, Frings L, Wagner K, Unterrainer J, Schulze-Bonhage A. Cognitive functions in juvenile and adult patients with gelastic epilepsy due to hypothalamic hamartoma. Epilepsia 2006;47(1):153–158 37. Striano S, Santulli L, Ianniciello M, Ferretti M, Romanelli P, Striano P. The gelastic seizures-hypothalamic hamartoma syndrome: facts, hypotheses, and perspectives. Epilepsy Behav 2012;24(1):7–13 38. Berberoğlu M. Precocious puberty and normal variant puberty: definition, etiology, diagnosis and current management. J Clin Res Pediatr Endocrinol 2009;1(4):164–174 39. Cassio A, Cacciari E, Zucchini S, Balsamo A, Diegoli M, Orsini F. Central precocious puberty: clinical and imaging aspects. J Pediatr Endocrinol Metab 2000;13(Suppl 1):703–708 40. Veendrick-Meekes MJ, Verhoeven WM, van Erp MG, van Blarikom W, Tuinier S. Neuropsychiatric aspects of patients ­ with hypothalamic hamartomas. Epilepsy Behav 2007;11(2): 218–221

44. Andermann F, Arzimanoglou A, Berkovic SF. Hypothalamic hamartoma and epilepsy: the pathway of discovery. Epileptic Disord 2003;5(4):173–175 45. Berkovic SF, Kuzniecky RI, Andermann F. Human epileptogenesis and hypothalamic hamartomas: new lessons from an experiment of nature. Epilepsia 1997;38(1):1–3 46. Prigatano GP, Wethe JV, Gray JA, et al. Intellectual functioning in presurgical patients with hypothalamic hamartoma and refractory epilepsy. Epilepsy Behav 2008;13(1):149–155 47. Park C, Wethe JV, Kerrigan JF. Decreased quality of life in children with hypothalamic hamartoma and treatment-resistant epilepsy. J Child Neurol 2013;28(1):50–55 48. Mori K, Handa H, Takeuchi J, Hanakita J, Nakano Y. Hypothalamic hamartoma. J Comput Assist Tomogr 1981;5(4):519–521 49. Pang D, Rosenbaum AE, Wilberger JE Jr, Gutai JP. Metrizamide computed tomographic cisternography for the diagnosis of occult lesions of the hypothalamic-hypophyseal axis in children. Neurosurgery 1981;8(5):531–541 50. Tasch E, Cendes F, Li LM, et al. Hypothalamic hamartomas and gelastic epilepsy: a spectroscopic study. Neurology 1998;51(4): 1046–1050 51. Amstutz DR, Coons SW, Kerrigan JF, Rekate HL, Heiserman JE. Hypothalamic hamartomas: correlation of MR imaging and spectroscopic findings with tumor glial content. AJNR Am J Neuroradiol 2006;27(4):794–798 52. Troester M, Haine-Schlagel R, Ng YT, et al. EEG and video-EEG seizure monitoring has limited utility in patients with hypothalamic hamartoma and epilepsy. Epilepsia 2011;52(6):1137–1143 53. Cascino GD, Andermann F, Berkovic SF, et al. Gelastic seizures and hypothalamic hamartomas: evaluation of patients undergoing chronic intracranial EEG monitoring and outcome of surgical treatment. Neurology 1993;43(4):747–750 54. Munari C, Kahane P, Francione S, et al. Role of the hypothalamic hamartoma in the genesis of gelastic fits (a videostereo-EEG study). Electroencephalogr Clin Neurophysiol 1995;95(3):154–160 55. Palmini A, Chandler C, Andermann F, et al. Resection of the lesion in patients with hypothalamic hamartomas and catastrophic epilepsy. Neurology 2002;58(9):1338–1347 56. Choi JU, Yang KH, Kim TG, et al. Endoscopic disconnection for hypothalamic hamartoma with intractable seizure. Report of four cases. J Neurosurg 2004;100(5, Suppl Pediatrics): 506–511 57. Kahane P, Ryvlin P, Hoffmann D, Minotti L, Benabid AL. From hypothalamic hamartoma to cortex: what can be learnt from depth recordings and stimulation? Epileptic Disord 2003;5(4):205–217 58. Kuzniecky R, Guthrie B, Mountz J, et al. Intrinsic epileptogenesis of hypothalamic hamartomas in gelastic epilepsy. Ann Neurol 1997;42(1):60–67 59. Delalande O, Fohlen M. Disconnecting surgical treatment of hypothalamic hamartoma in children and adults with refractory epilepsy and proposal of a new classification. Neurol Med Chir (Tokyo) 2003;43(2):61–68 60. Régis J, Scavarda D, Tamura M, et al. Epilepsy related to hypothalamic hamartomas: surgical management with special reference to gamma knife surgery. Childs Nerv Syst 2006;22(8):881–895 61. Régis J, Scavarda D, Tamura M, et al. Gamma knife surgery for epilepsy related to hypothalamic hamartomas. Semin Pediatr Neurol 2007;14(2):73–79

495

496

IVd  Extratemporal Lobe Epilepsy and Surgical Approaches 62. Northfield DW, Russell DS. Pubertas praecox due to hypothalamic hamartoma: report of two cases surviving surgical removal of the tumour. J Neurol Neurosurg Psychiatry 1967;30(2): 166–173

83. Bunyaratavej K, Locharernkul C, Tepmongkol S, Lerdlum S, Shuangshoti S, Khaoroptham S. Successful resection of h ­ypothalamic hamartoma with intractable gelastic seizures—by transcallosal subchoroidal approach. J Med Assoc Thai 2006;89(8):1269–1276

63. Nishio S, Fujiwara S, Aiko Y, Takeshita I, Fukui M. Hypothalamic hamartoma. Report of two cases. J Neurosurg 1989;70(4): 640–645

84. Anderson JF, Rosenfeld JV. Long-term cognitive outcome after transcallosal resection of hypothalamic hamartoma in older adolescents and adults with gelastic seizures. Epilepsy Behav 2010;18(1–2):81–87

64. Ponsot G, Diebler C, Plouin P, et al. [Hypothalamic hamartoma and gelastic crises. Apropos of 7 cases] Arch Fr Pediatr 1983;40(10):757–761 65. Sato M, Ushio Y, Arita N, Mogami H. Hypothalamic hamartoma: report of two cases. Neurosurgery 1985;16(2):198–206 66. Mittal S, Mittal M, Montes JL, Farmer JP, Andermann F. Hypothalamic hamartomas. Part 2. Surgical considerations and outcome. Neurosurg Focus 2013;34(6):E7 67. Palmini A, Paglioli-Neto E, Montes J, Farmer JP. The treatment of patients with hypothalamic hamartomas, epilepsy and behavioural abnormalities: facts and hypotheses. Epileptic Disord 2003;5(4):249–255 68. Rosenfeld JV, Harvey AS, Wrennall J, Zacharin M, Berkovic SF. Transcallosal resection of hypothalamic hamartomas, with control of seizures, in children with gelastic epilepsy. Neurosurgery 2001;48(1):108–118 69. Calisto A, Dorfmüller G, Fohlen M, Bulteau C, Conti A, Delalande O. Endoscopic disconnection of hypothalamic hamartomas: safety and feasibility of robot-assisted, thulium laser-based procedures. J Neurosurg Pediatr 2014;14(6): 563–572 70. Homma J, Kameyama S, Masuda H, et al. Stereotactic radiofrequency thermocoagulation for hypothalamic hamartoma with intractable gelastic seizures. Epilepsy Res 2007;76(1):15–21 71. Khan S, Wright I, Javed S, et al. High frequency stimulation of the mamillothalamic tract for the treatment of resistant seizures associated with hypothalamic hamartoma. Epilepsia 2009;50(6):1608–1611 72. Régis J, Bartolomei F, de Toffol B, et al. Gamma knife surgery for epilepsy related to hypothalamic hamartomas. Neurosurgery 2000;47(6):1343–1351, discussion 1351–1352 73. Régis J, Lagmari M, Carron R, et al. Safety and efficacy of Gamma Knife radiosurgery in hypothalamic hamartomas with severe epilepsies: a prospective trial in 48 patients and review of the literature. Epilepsia 2017;58(Suppl 2):60–71 74. Schulze-Bonhage A, Trippel M, Wagner K, et al. Outcome and predictors of interstitial radiosurgery in the treatment of gelastic epilepsy. Neurology 2008;71(4):277–282 75. Wagner K, Buschmann F, Zentner J, Trippel M, Schulze-Bonhage A. Memory outcome one year after stereotactic interstitial radiosurgery in patients with epilepsy due to hypothalamic hamartomas. Epilepsy Behav 2014;37:204–209

85. Ng YT, Rekate HL, Prenger EC, et al. Transcallosal resection of hypothalamic hamartoma for intractable epilepsy. Epilepsia 2006;47(7):1192–1202 86. Akai T, Okamoto K, Iizuka H, Kakinuma H, Nojima T. Treatments of hamartoma with neuroendoscopic surgery and stereotactic radiosurgery: a case report. Minim Invasive Neurosurg 2002;45(4):235–239 87. Cappabianca P, Cinalli G, Gangemi M, et al. Application of neuroendoscopy to intraventricular lesions. Neurosurgery 2008;62(Suppl 2):575–597, discussion 597–598 88. Shim KW, Chang JH, Park YG, Kim HD, Choi JU, Kim DS. Treatment modality for intractable epilepsy in hypothalamic hamartomatous lesions. Neurosurgery 2008;62(4):847–856, discussion 856 89. De Ribaupierre S, Delalande O. Hemispherotomy and other disconnective techniques. Neurosurg Focus 2008;25(3):E14 90. Pati S, Abla AA, Rekate HL, Ng YT. Repeat surgery for hypothalamic hamartoma in refractory epilepsy. Neurosurg Focus 2011;30(2):E3 91. Arita K, Kurisu K, Iida K, et al. Subsidence of seizure induced by stereotactic radiation in a patient with hypothalamic hamartoma. Case report. J Neurosurg 1998;89(4):645–648 92. Barajas MA, Ramírez-Guzman MG, Rodríguez-Vázquez C, Toledo-Buenrostro V, Cuevas-Solórzano A, RodríguezHernández G. Gamma knife surgery for hypothalamic hamartomas accompanied by medically intractable epilepsy and precocious puberty: experience in Mexico. J Neurosurg 2005; 102(Suppl):53–55 93. Drees C, Chapman K, Prenger E, et al. Seizure outcome and complications following hypothalamic hamartoma treatment in adults: endoscopic, open, and Gamma Knife procedures. J ­Neurosurg 2012;117(2):255–261 94. Dunoyer C, Ragheb J, Resnick T, et al. The use of stereotactic radiosurgery to treat intractable childhood partial epilepsy. Epilepsia 2002;43(3):292–300 95. Mathieu D, Kondziolka D, Niranjan A, Flickinger J, Lunsford LD. Gamma knife radiosurgery for refractory epilepsy caused by hypothalamic hamartomas. Stereotact Funct Neurosurg 2006;84(2–3):82–87 96. Unger F, Schröttner O, Feichtinger M, Bone G, Haselsberger K, Sutter B. Stereotactic radiosurgery for hypothalamic hamartomas. Acta Neurochir Suppl (Wien) 2002;84:57–63

76. Rosenfeld JV. The evolution of treatment for hypothalamic hamartoma: a personal odyssey. Neurosurg Focus 2011;30(2):E1

97. Quigg M, Barbaro NM. Stereotactic radiosurgery for treatment of epilepsy. Arch Neurol 2008;65(2):177–183

77. Machado HR, Hoffman HJ, Hwang PA. Gelastic seizures treated by resection of a hypothalamic hamartoma. Childs Nerv Syst 1991;7(8):462–465

98. Fujimoto Y, Kato A, Saitoh Y, et al. Stereotactic radiofrequency ablation for sessile hypothalamic hamartoma with an image fusion technique. Acta Neurochir (Wien) 2003;145(8):697–700, discussion 700–701

78. Miller ML, Kaufman BA, Lew SM. Modified osteoplastic orbitozygomatic craniotomy in the pediatric population. Childs Nerv Syst 2008;24(7):845–850 79. Miranda P, Esparza J, Cabrera A, Hinojosa J. Giant hypothalamic hamartoma operated through subfrontal approach with orbitary rim osteotomy. Pediatr Neurosurg 2006;42(4): 254–257 80. Abla AA, Rekate HL, Wilson DA, et al. Orbitozygomatic resection for hypothalamic hamartoma and epilepsy: patient selection and outcome. Childs Nerv Syst 2011;27(2):265–277 81. Ghanta RK, Koti K, Kongara S, Meher GE. Surgical excision of hypothalamic hamartoma in a twenty months old boy with precocious puberty. Indian J Endocrinol Metab 2011;15 (Suppl 3):S255–S258 82. Harvey AS, Freeman JL, Berkovic SF, Rosenfeld JV. Transcallosal resection of hypothalamic hamartomas in patients with intractable epilepsy. Epileptic Disord 2003;5(4):257–265

99. Parrent AG. Stereotactic radiofrequency ablation for the treatment of gelastic seizures associated with hypothalamic hamartoma. Case report. J Neurosurg 1999;91(5):881–884 100. Burrows AM, Marsh WR, Worrell G, et al. Magnetic resonance imaging-guided laser interstitial thermal therapy for previously treated hypothalamic hamartomas. Neurosurg Focus 2016;41(4):E8 101. Buckley RT, Wang AC, Miller JW, Novotny EJ, Ojemann JG. Stereotactic laser ablation for hypothalamic and deep intraventricular lesions. Neurosurg Focus 2016;41(4):E10 102. Rolston JD, Chang EF. Stereotactic laser ablation for ­hypothalamic hamartoma. Neurosurg Clin N Am 2016;27(1): 59–67 103. Tovar-Spinoza Z, Carter D, Ferrone D, Eksioglu Y, Huckins S. The use of MRI-guided laser-induced thermal ablation for epilepsy. Childs Nerv Syst 2013;29(11):2089–2094.

53

  Hemispherectomy and Hemispherotomy Techniques in Pediatric Epilepsy Surgery Oğuz Çataltepe

Summary Hemispherectomy is one of the most effective surgical interventions in the management of children with unilateral multilobar or hemispheric epilepsy. Hemispherectomy, or resection of the entire hemisphere, was first performed by Dandy in 1928 and later McKenzie used the hemispherectomy technique for the first time to treat an epilepsy patient with infantile hemiplegia. Since the description of “anatomical hemispherectomy,” many other variations and modifications of this procedure have been developed. Rasmussen developed “functional hemispherectomy” first and then next generation of neurosurgeons developed various “hemispherotomy” techniques to further reduce the resection volume. All of these modifications aimed gradual reduction in the amount of the resected brain tissue while still achieving complete disconnection of the entire hemisphere. All these hemispherectomy techniques provide very satisfactory seizure control with a relatively low complication rate in a very challenging pediatric patient population. The pathology appears to be the most significant determinant of outcome, and incomplete disconnection appears to be the most frequent reason for surgical failure. Keywords:  hemispherectomy, functional hemispherectomy, hemispherotomy

„„ Introduction Hemispherectomy is one of the most effective surgical interventions in the management of children with unilateral multilobar or hemispheric epilepsy. Since Krynauw’s successful application of this technique for children with infantile hemiplegia in 1950,1 hemispherectomy has been used in the surgical management of hemispheric epilepsy with remarkably high success rates. Original surgical technique was “anatomical hemispherectomy,” mainly removal of the entire abnormal hemisphere.2 Since then, there have been many variations and modifications of the hemispherectomy procedure. The modifications all aimed gradual reduction in the amount of the resected brain tissue while still achieving complete disconnection of the entire hemisphere. The first effective application of this concept was defined by Rasmussen3 in the 1970s and it was

called “functional hemispherectomy.” Rasmussen’s ­ functional hemispherectomy technique was further modified by next generation of neurosurgeons to further reduce the resection volume. In the 1990s, “hemispherotomy” techniques were ­ developed to disconnect all neuronal fibers to functionally isolate the damaged hemisphere without much cortical resection (Video 53.1).4,​5,​6,​7 The evolution of hemispheric surgical interventions from anatomic hemispherectomy to hemispherotomy is a fascinating part of pediatric epilepsy surgery. Here, we will summarize the development of these techniques and review their indications and applications in pediatric epilepsy surgery. Chapters 54 through 61 of this book provide additional information about this topic with in-depth descriptions of the main variations of hemispherectomy and hemispherotomy techniques.

„„ Hemispheric Epilepsy Surgery: From Resection to Disconnection Hemispherectomy, or resection of the entire hemisphere, was first performed by Dandy2 in 1928. Dandy used this technique to treat a patient with hemispheric glioma. In 1938, Canadian neurosurgeon McKenzie8,​9 used the hemispherectomy technique for the first time to treat an epilepsy patient with infantile hemiplegia. Thereafter, first hemispherectomy series, including only epilepsy patients (12 children with infantile hemiplegia), was published by Krynauw1 in 1950 and the procedure became more popular in the following years. However, several reports regarding delayed and lifethreatening complications in hemispherectomy patients were published in the late 1960s. These reports were alarming and number of cases started to decline. Then anatomic hemispherectomy was almost completely abandoned after publication of

Video 53.1 Peri-insular hemispherotomy. (This video is provided courtesy of Oğuz Çataltepe.) ht t ps://www.thieme.de/de/q.ht m?p=opn/ tp/255910102/9781626238176_c053_v001&t=video

498

IVe  Hemispheric Surgery Techniques Oppenheimer and Griffith’s report10 in 1966 and subsequent reports on superficial cerebral hemosiderosis in postmortem studies.3,​5 After this era, several neurosurgeons started to develop new strategies to modify or replace anatomical hemispherectomy, including hemidecortication, modified anatomical hemispherectomy, and functional hemispherectomy. Some of them were effective and some others were unsuccessful. This search led to the development of the h ­ emispherotomy techniques in use today.11 Current hemispherotomy techniques r­epresent the latest stage in the conceptual and technical e ­ volution of ­functional hemispherectomy.

„„ Epilepsy Syndromes Associated with Hemispheric Lesions Hemispherectomy is an effective surgical procedure for the treatment of hemispheric epilepsy syndromes. The typical candidate for hemispherectomy is a patient with hemiplegia secondary to a unilaterally damaged hemisphere because of a congenital or acquired lesion. The most common conditions causing unilateral multilobar or hemispheric epilepsy are generally seen in infants with catastrophic epilepsy, such as infantile spasms, hemiconvulsion-hemiplegia-epilepsy (HHE) syndrome, Sturge-Weber syndrome, hemimegalencephaly, multilobar cortical dysplasia, and congenital hemiplegia from a perinatal infarction. In addition, some acquired conditions may also cause intractable hemispheric epilepsy, such as Rasmussen’s encephalitis, late ischemic events, and trauma-related hemispheric injuries.

Infantile Spasms Infantile spasms are almost entirely seen in the first year of life and are associated with developmental delay, regression, and medically refractory seizures. The patients present with a typical electroencephalographic (EEG) pattern: hypsarrhythmia. Seizures in infantile spasms are seen in clusters; they occur even during sleep, and cause exhaustion and lethargy. Various types of myoclonic seizures, such as flexor and extensor spasms with a cry, are seen and are followed by a brief episode of akinesia. Many conditions can cause infantile spasms, such as neurocutaneous syndromes, congenital brain malformations, metabolic and degenerative diseases, and hypoxic-ischemic ­ insults.12

Hemiconvulsion-Hemiplegia-Epilepsy Syndrome HHE syndrome is most frequently seen within the first 2 years of life. The initial phase of the syndrome presents with unilateral, prolonged hemiconvulsive seizures that involve the face, arms, and legs. The second phase is characterized by hemiplegia, and the third phase is characterized by partial epileptic seizures. The syndrome progresses to chronic epilepsy within 1 to 2 years. Although there are many possible causes for HHE syndrome, including meningitis, subdural effusions, trauma, and hemispheric lesions, no cause can be determined in many other cases. The etiology of this condition is still poorly u ­ nderstood.

In the course of the disease, hemiatrophia cerebri develops gradually after hemiconvulsive seizures and hemiplegia.12,​13

Sturge-Weber Syndrome Sturge-Weber syndrome is a progressive neurocutaneous disorder associated with pial angiomatosis involving the cerebral cortex, along with a cutaneous angioma in the trigeminal nerve territory on the face and scalp. Facial angioma (port wine stain) is seen in 90% of these cases. Facial and ­leptomeningeal ­angiomas occur mostly uni- and ipsilaterally but can also be seen bilaterally in up to 20% of patients. Pial angiomatosis mostly involves the parietooccipital region, but it can be extensive and may involve entire hemisphere in some cases. Sturge-Weber patients have a very peculiar leptomeningeal vascular bed, with hypertrophic pial vessels and frequently absent major venous sinuses and cortical bridging veins. This peculiar vascular anatomy creates a strong retrograde venous flow into the ventricle. This abnormal hemodynamic induces hypoxia in the surrounding brain tissue because of the diversion of cerebral blood flow away from the parenchyma and associated venous stasis. This abnormal blood circulation eventually causes cellular damage in the brain parenchyma and secondary seizures. The most common symptoms in Sturge-Weber patients are seizure (75–90%), developmental delay, hemiparesis, and various ophthalmologic problems, such as glaucoma and optic atrophy. Seizures are usually the earliest symptoms in Sturge-Weber patients, with 70% of the patients experiencing seizures in the first year of life. Seizures may even occur during the newborn period. If seizures start in infancy, then prognosis is more g ­ uarded. Most seizures are simple/complex partial seizures, with frequent secondary generalization, and are often unresponsive to medications (only 10% respond well to medications). Patients may develop hemiplegia after an episode of serial seizures in the first year of life. Therefore, vigorous treatment is essential to prevent postconvulsive damage during infancy. Although multilobar excision/disconnection or hemispherectomy is the main treatment modality in Sturge-Weber patients with severe epilepsy, waiting until the child is 1 year old before proceeding to surgery may also be a reasonable approach in some cases to prove the intractability of the seizures.12,​13,​14,​15

Hemimegalencephaly Hemimegalencephaly is an extensive neuronal migration disorder involving the entire hemisphere. This abnormal, unilaterally enlarged hemisphere generally has no cortical lamination but wide, thickened, and flattened cortex with shallow gyri. Other abnormal histological and radiological findings include reduced number of sulci, reduced white matter volume, subcortical heterotopia, calcifications, poor gray–white matter differentiation, hypoplastic corpus callosum, and an ipsilaterally enlarged or shrunken ventricle. The frontal and occipital lobes in the abnormal hemisphere are frequently hyperplastic, unlike the hypoplastic temporal lobe. Hemimegalencephaly can be seen as an isolated entity or may be associated with Klippel-Trenaunay syndrome, hypomelanosis of Ito, linear nervous sebaceous of Jadassohn, or Proteus syndrome. Medically intractable seizures are the most common

53  Hemispherectomy and Hemispherotomy Techniques in Pediatric Epilepsy Surgery finding and generally start during infancy.16 Severe epileptic encephalopathy and developmental delay are also common in these patients.16 If the seizures are not well controlled, patients may develop hemiparesis, hemianopia, and mental retardation. High mortality rates in the first months of life are seen in these patients because of the continuous seizures.12,​13,​17

Cortical Dysplasia Unilateral, multilobar, or extensive cortical dysplasia is another congenital condition associated with multilobar epilepsy in early childhood. Hemispherectomy or multilobar resections are frequently the best treatment options for these patients. Further information about this condition can be found in ­ ­Chapters 41 and 42.

Rasmussen’s Encephalitis Rasmussen’s encephalitis is a chronic, progressive condition that is characterized by intractable epilepsy and progressive atrophy in one hemisphere. The syndrome was described by ­Rasmussen in 1958. Although seizures in Rasmussen’s encephalitis frequently start with a generalized tonic–clonic seizure, they ­usually c­ ontinue as partial epilepsy. Rasmussen’s encephalitis is a progressive disease that causes hemiplegia in the majority of cases. The results of initial imaging studies might be normal, but follow-up imaging studies reveal unilateral ventricular enlargement, followed by hyperintense changes in MRI and, finally, focal atrophy in the primary sensorimotor cortex and insula. Mesial temporal involvement in these cases is frequently seen very late, and occipital involvement is seen even later.13,​18,​19

Porencephalic Cyst Perinatal vascular insults, such as internal cerebral artery and middle cerebral artery infarcts, intracerebral hemorrhage secondary to arteriovenous malformations and congenital ­ coagulopathies, and traumatic brain injuries can cause large, hemispheric porencephalic cysts.14,​18 These patients frequently have unilaterally enlarged ventricles and severe brain atrophy secondary to extensive tissue loss with large porencephalic cystic areas. Medically intractable seizures and hemiplegia are frequent findings and the patients are ideal candidates for ­hemispherectomy and, especially, hemispherotomy procedures.

„„ Preoperative Assessment Hemispherectomy is a very extensive surgical intervention with dramatic and gratifying results. However, it is associated with significant morbidity and mortality risks. Therefore, preoperative assessment of the patients to select ideal candidates is of utmost importance. Assessment of hemispherectomy candidates should include the following questions: Is the patient’s condition medically intractable? Does the patient’s clinical status justify such an extensive procedure? Are the patient’s electrophysiological findings strongly suggestive of a unilateral hemispheric origin of the seizures? Do structural and ­functional imaging studies show unilateral hemispheric

­damage? Is the contralateral hemisphere structurally, functionally, and electrographically healthy? Do the patient and family fully understand the extent of the intervention, the associated risks, and potential results? Does the patient have reliable family support? The epilepsy surgery team should determine the answers of these questions using available preoperative assessment tools, tests, and techniques. If the answers are affirmative, then the patient is deemed an acceptable surgical candidate for ­hemispherectomy.

Medical Intractability As in all epilepsy surgery cases, the first step in preoperative assessment is determining the intractability of the seizures. Although some patients may require extensive trials to prove that the seizures are intractable, children who have seizures secondary to hemispheric lesions rarely need exhaustive trials to prove medical intractability. Determining the intractability of seizures to major antiepileptic drugs (AEDs) may be relatively easy in patients with hemispheric lesions, such as Rasmussen’s encephalitis, Sturge-Weber syndrome, and cortical dysplasia because of the very nature of these conditions. Conversely, patients with other conditions may require more time and effort to document the intractability of their seizures.

Clinical Status The ideal candidate for hemispherectomy is a medically intractable epilepsy patient with hemiplegia and hemianopsia secondary to unilateral hemispheric damage. If the patient has no motor weakness or has only mild hemiparesis and partial hemianopsia, then surgery will provide good seizure control but with impairment of the patient’s neurological status. The decision for surgery in these patients is not straightforward, and opinions about the suitability of these patients for surgery are frequently controversial. Nevertheless, some of these patients may be selected as surgical candidates for hemispherectomy because of the severity of their seizures and debilitating effect of these seizures on the functional status of these patients. Certain progressive conditions, such as Rasmussen’s encephalitis, have a well-known natural course that eventually results in motor and cognitive worsening. Patients with catastrophic infantile epilepsy syndromes may present with hundreds of daily seizures, resulting in no functional life. These seizures pose a very high risk of damage to the developing brain. Even if these patients do not have hemiplegia, they may still be considered candidates for hemispherectomy procedure because early surgery may prevent an eventual decline in cognitive function and psychomotor development.5,​15,​17,​18,​19,​20

Physical Examination Patients with unilateral hemispheric damage generally have significant distal extremity weakness but relatively good proximal strength. Upper extremity weakness is also more pronounced than lower extremity weakness in these patients. Shoulder function is generally good, with a normal range of

499

500

IVe  Hemispheric Surgery Techniques elbow movements; patients can lift their arms up to shoulder level horizontally. However, wrist function is typically minimal, and fine finger movements are absent. In the lower extremities, these patients have good proximal strength, with good major joint movements but no toe movements. The vast majority of these patients have a variable degree of spasticity, but they walk either independently or with help.5,​13 Most likely reason for differences in proximal and distal function in the extremities is respective locations that control these functions in the brain. Although fine finger movements and repeated alternating movements, such as finger–thumb oppositions and foot tapping, are mainly cortical functions, gross motor movements, such as major joint movements, originate from subcortical structures, with ipsilateral motor participation as well.11 Therefore, distal impairment in these patients is much more pronounced than proximal impairment. Another major neurological deficit of these patients is hemianopsia. Detailed ophthalmologic examination is important, both to assess the baseline status of vision and to counsel the parents preoperatively. In addition, some of the syndromes causing unilateral hemispheric damage may also be associated with other ophthalmological findings, such as retinal damage, extraocular muscle weakness, and optic pathway damage. It is also very important to verify preoperatively that the visual field deficit is not bilateral.13 However, ophthalmological examination may not be feasible in many children because of their age or developmental status.

Electroencephalographic Assessment Preoperative ictal and interictal EEG studies help to verify the unilaterality of the epileptogenic zone and to determine its extent. Determining the extent of the epileptogenic zone is especially important for deciding whether hemispherectomy is needed or a limited cortical resection or disconnection will suffice. Most importantly, it should be preoperatively proven that the patient’s seizures are unilateral and the epileptogenic zone is contained within the damaged hemisphere. Predictors of good outcome in these patients are the presence of ipsilateral suppression of electrical activity associated with multifocal epileptogenic abnormalities confined to the damaged hemisphere, bilateral synchronous discharges spreading from the abnormal hemisphere without contralateral slowing, absence of generalized discharges, presence of bilateral independent spiking, and abnormal background activity in the “good” hemisphere. Although having sporadic epileptiform activities, some abnormal secondary or independent EEG findings, and nonepileptiform abnormalities in the “good” hemisphere, do not exclude those patients as candidates for hemispherectomy but may imply an unfavorable outcome, especially if independent interictal sharp wave activity in the “good” hemisphere is present.13,​14,​18,​21

Structural Imaging Anatomical imaging, mainly magnetic resonance imaging (MRI), provides detailed structural information about the damaged hemisphere. It is very helpful to determine the etiology and extent of the lesion, assess the integrity of the “good” hemisphere, and visualize the anatomical details

of the damaged hemisphere before deciding on the most appropriate surgical approach for the patient. A ­ natomical details such as ventricle size, the shape and depth of the sylvian fissure, displacement and distortion of anatomical landmarks, the thickness of the corpus callosum, surface ­ anatomy including cortical thickness and sulcal depth, and ­severity of p ­ arenchymal atrophy are of utmost significance for the neurosurgeon. Knowledge of these details helps the neurosurgeon to visualize the patient’s anatomy and preoperatively design the best surgical strategy. Some findings, such as the presence of atrophy in the ipsilateral cerebral peduncle and medulla, are also helpful to predict a lower risk for postoperative worsening.11 Conversely, other abnormal MRI findings, such as hyperintense areas in deep gray nuclei in the “good” hemisphere, may imply mitochondrial or metabolic disease; surgical determinations in these cases should be made very carefully. Positron emission tomography (PET) scans may also be helpful for further assessment of these cases. Magnetic resonance venography is a valuable tool, especially in Sturge-Weber patients, to enhance the understanding of venous drainage patterns. If needed, cerebral angiography also should be performed in these patients.5,​11,​13,​14,​17

Functional Imaging Functional imaging provides information about the extent and location of functionally damaged areas, shifted locations of some cortical functions, metabolic status of the “good” hemisphere, and expected postoperative outcome. Most commonly used techniques to assess the functional status of the brain are functional MRI, PET, and single photon emission computed tomography (SPECT) scans. If hypometabolic areas and epileptogenic zones are contained within damaged hemisphere, this is interpreted as a good predictor. If hypometabolism is also seen in the “good” hemisphere, this is interpreted as a warning for more extensive involvement. Further details regarding these studies and their application in preoperative assessment of patients can be found in ­Chapter 21.13,​14,​17

Wada Test The Wada test is important for verifying that “good” hemisphere can carry on speech and memory functions satisfactorily after surgical disconnection of damaged hemisphere is performed. However, Wada test is not always feasible in children because of their age and developmental status. It should be emphasized that Wada test should be performed only on damaged hemisphere, not on “good” hemisphere, to avoid any risk of ischemic injury to the latter.13,​14,​17 Further information regarding Wada test and its application can be found in Chapter 26.

Neuropsychological Evaluation Neuropsychological evaluation is a routine part of preoperative assessment of epilepsy surgery candidates. It is especially helpful in hemispherectomy candidates because one hemisphere in these patients is already severely damaged, and determining the function level of “good” hemisphere provides critical ­information.

53  Hemispherectomy and Hemispherotomy Techniques in Pediatric Epilepsy Surgery Neuropsychological assessment is performed to determine baseline level of cognitive function status, to locate certain cortical functions, to describe the extent and location of functional impairment, and to counsel parents regarding postoperative expectations. Early shift of language function is frequently seen in patients with early severe hemispheric damage. Complete lateralization of speech is acquired by the age of 5 to 6 years; thereafter, a complete shift of speech function is very difficult. Therefore, age of the patient at the time of hemispheric insult is critical for postoperative recovery of speech function. In addition to age, other factors, such as site of damage, extent and severity of epileptogenic activities, and progression of the disease, are critical in shifting of the speech function. A finding of severe cognitive impairment on neuropsychological assessment may imply diffuse, bilateral hemispheric involvement or structural or electrophysiological abnormality and may suggest a poor prognosis.5,​11,​14,​15,​18

„„ Surgical Planning The goal of surgery in pediatric epilepsy is not only controlling the seizures but also protecting the immature brain from deleterious effects of the seizures and AEDs during its most vulnerable period. The impact of ictal and interictal epileptogenic activities on an immature brain has become a growing concern in the last two decades, and earlier surgical intervention in pediatric epilepsy patients with hemispheric damage has increasingly become an acceptable option. Increasing support for earlier surgery in catastrophic pediatric epilepsy patients arises not only from concerns about the burden of epilepsy on immature brain but also from increased awareness of plasticity window of developing brain.15,​18 The primary concern in hemispherectomy cases is whether the patient will have any additional neurological deficit after surgery. If the child is younger than 3 years, then new neurological deficit is generally not expected. However, in late-onset cases, such as Rasmussen’s encephalitis, timing of surgery may not be straightforward. Several factors should be considered in these cases to determine the best timing for hemispherectomy: severity of the seizures and their current effect on the patient’s functional status, potential effects of the seizures on cognitive and neuropsychological development of the child, availability of adequate AED trials, and natural course of the disease. It is also important to remember that shorter the time interval between seizure onset and surgery, higher the success rate. This is especially true in patients with Rasmussen’s encephalitis.11 Again, it is critical to remember that earlier seizure control in children provides the best psychosocial environment with the least seizure burden, resulting in optimal psychosocial development.11,​14,​15,​16,​17,​18,​20,​21

„„ Surgical Approaches Once surgical intervention for hemispheric epilepsy has been decided, the last step is choosing the most appropriate surgical

technique. As mentioned previously, surgical intervention for hemispheric pathologies has evolved from anatomical hemispherectomy to hemispherotomy. Currently, a variety of hemispherectomy techniques have been practiced. We will have a brief overview of these techniques here, and in-depth information about each of these techniques can be found in Chapters 54 through 61.

Anatomical Hemispherectomy Anatomical hemispherectomy was first performed by Dandy in 1928.2 The history and details of the technique can be found in Chapter 55. Although it has fallen out of favor steadily, anatomical hemispherectomy has enjoyed a limited revival in the last decade. A recent anatomical hemispherectomy series with long-term follow-up reported much smaller complication rates than the rates in earlier series. It has been claimed that delayed complication rates reported in earlier studies were most likely overestimated.10,​18,​20,​22,​23

Hemidecortication Hemidecortication was described by Ignelzi and Bucy in 1968 as an alternative to the anatomical hemispherectomy.9 The procedure starts with sylvian cistern dissection and middle cerebral artery occlusion, followed by occlusion of the anterior and posterior cerebral arteries. The entire cerebral cortex is then removed en bloc, leaving the basal ganglia and thalamus intact with white matter coverage over the anterior horn and body of the lateral ventricle. Two other modifications of this technique were described in the 1990s. Winston et al24 described their hemidecortication technique in 1992 as “de-gloving” of the entire cerebral cortex by dissecting it around the lateral ventricle, leaving only a layer of white matter around the ventricular system after developing a plane of dissection from the edges of the insula by opening the sylvian fissure. De-gloving is first performed on the frontal, parietal, and occipital lobes and then on the temporal fossa by removing the entire cortical mantle. In 1996, Carson et al19 published a Johns Hopkins series, with some modifications. The details of this technique can be found in Chapter 56. The main problems with hemidecortication are excessive blood loss, heightened risk of infection, and, especially, technical difficulty of performing a complete cortical resection in the medial and basal sides of the lobes.

Functional Hemispherectomy Functional hemispherectomy was developed by Rasmussen and the details of his technique were published in 1983.3 It can be best defined as partial anatomical resection and full disconnection of the damaged hemisphere. After its introduction, this technique has become the most widely used hemispherectomy technique. It preserves the anterior frontal, posterior parietal, and occipital lobes while resecting the central lobe and the anterior twothirds of the temporal lobe, including mesial structures; disconnecting all commissural and projection fibers with a complete callosotomy; and severing the connection with the brainstem and thalamus through frontobasal and occipitoparietal cuts. The

501

502

IVe  Hemispheric Surgery Techniques rational for this technique was to perform a functionally complete but anatomically incomplete h ­ emispherectomy by leaving the large part of the brain intact to avoid well-known complications associated with anatomical hemispherectomy secondary to an enormous postsurgical cavity. Rasmussen reported seizure-free outcome with low complication rates in 75% of his patients. Other modifications of the functional hemispherectomy technique have also been described, including Comair’s transsylvian functional hemispherectomy technique.13 Further details of functional hemispherectomy technique can be found in Chapter 57.

Hemispherotomy The main goal of anatomical hemispherectomy in epilepsy surgery is the functional disconnection of the damaged hemisphere (Video 53.1). Hemispheric resection is usually performed to achieve this goal and tissue resection is not the main objective of this procedure, as it is in neurooncological surgery. Rasmussen came up with the hemispheric disconnection concept by describing his functional hemispherectomy technique and showing that full functional disconnection with comparable seizure-free outcome rates is still feasible without resecting the entire hemisphere. The next generation of neurosurgeons continues to develop new, creative surgical techniques based on this line of thought by further decreasing the amount of resected brain volume while still performing full disconnection of the hemisphere. In 1992, Delalande et al4,​5 introduced the term “hemispherotomy” to describe the evolving hemispheric surgery techniques designed to minimize resection volume while still fully disconnecting the hemisphere. Since 1992, several other neurosurgeons have developed different hemispherotomy techniques, with each technique characterized by further decrease in the resected brain volume with smaller cortical incisions.4,​5,​25,​26,​27,​28,​29 Hemispherotomy is the latest step in the natural evolution of hemispheric surgery techniques. The main differences between the various hemispherotomy techniques pertain to the volume of resected brain tissue, the access route to the lateral ventricle, resecting or not resecting the insular cortex, resecting or disconnecting the hippocampus, and preserving or sacrificing the vascular structures in the peri-insular area. Although the following chapters in this section include detailed descriptions of each technique, we will briefly summarize these approaches here to emphasize their common characteristics as well as their differences. Whether hemispherotomy or functional hemispherectomy is the correct term to describe these new techniques is debatable. Regardless, these techniques can be organized into two main groups: techniques using a vertical approach and techniques using lateral approaches.

Vertical Approach: Transventricular Vertical Hemispherotomy Transventricular vertical hemispherotomy, the most commonly used vertical approach, was described by Delalande et al in 1992.4,​5 This surgical technique, like others, has evolved, and several changes were made since the initial d ­ escription.5,​

30 In its current form, the surgery starts with a small linear, paramedian incision parallel to the sagittal suture. A small (3 × 5 cm) frontoparietal craniotomy is performed 1 to 2 cm lateral to midline by staying one-third anterior and twothirds posterior to the coronal suture. A small (3 × 2 cm) cortical resection is then performed, by staying away from the midline, to enter the ipsilateral lateral ventricle. The body and splenium of the corpus callosum are divided intraventricularly, followed by a cut in the posterior column of the fornix at the level of the trigone by reaching the choroidal fissure behind the pulvinar. A vertical incision lateral to the thalamus is then made, guided by the choroid plexus in the temporal horn. This incision extends from the trigone to the most anterior part of the temporal horn by completely unroofing the ventricle. Corpus callosotomy is then completed by dividing the genu and rostrum of the corpus callosum. The posterior part of the gyrus rectus is resected subpially to expose the anterior cerebral artery (ACA) and optic nerve. The final step to complete the hemispheric disconnection is a straight incision anterolaterally through the caudate nucleus from the rectus gyrus to the anterior temporal horn by following the ACA.4 More details about this technique can be found in Chapter 59. Delalande et al4 reported 83 patients who underwent this procedure, including 30 patients (36%) with multilobar cortical dysplasia (hemimegalencephaly), 25 patients (30%) with ­Rasmussen syndrome, 10 patients (12%) with Sturge-­Weber syndrome, and 18 patients (22%) with ischemic-vascular sequelae. The postoperative seizure-free outcome rate was 74%. The best results, a 92% seizure-free outcome, were achieved in patients with Rasmussen’s encephalitis and Sturge-Weber syndrome. The postoperative hydrocephalus rate was 16%, and the mortality rate was 3.6%.

Lateral Approaches Perisylvian Transcortical Transventricular Hemispherical Deafferentation The perisylvian transcortical transventricular hemispherical deafferentation technique was described by Schramm et al in 1995.7 The goal of the procedure was described as deafferenting nearly all of the cortical structures of the damaged hemisphere from its connections to the basal ganglia and the contralateral hemisphere. The first step is a classic anteromesial temporal lobectomy, followed by a circular transcortical incision starting from the temporal horn and ending at the tip of the frontal horn. Major superficial veins and some of the middle cerebral artery branches are preserved. The next step of the procedure is a posterior basal disconnection. The posterior end of the hippocampal resection at the choroidal fissure is carried through the white matter subpially, crossing the calcarine sulcus and reaching the splenium. Then, a complete corpus callosotomy is performed intraventricularly. The last step to complete the deafferentation of the hemisphere is a frontobasal disconnection between the sphenoid wing and the rostral end of the callosotomy by staying subpial and anterior to the A1 segment.

53  Hemispherectomy and Hemispherotomy Techniques in Pediatric Epilepsy Surgery

Transsylvian Keyhole Functional Hemispherectomy This technique was reported on by Schramm et al in 2001 as a modification of his original technique, described earlier.24,​31 In this new modification, anteromesial temporal lobectomy is replaced with selective amygdalohippocampectomy. More information about this technique can be found in Chapter 58.

Peri-insular Hemispherotomy Peri-insular hemispherotomy (Video 53.1) was described by Villemure and Mascott in 1995.6 It was defined as a conceptual and technical evolution of the functional hemispherectomy described by ­Rasmussen and represented the latest stage in its development.11,​19,​26,​32 More information about this technique can be found in Chapter 60. Shimizu and Maehara also described a modification of Villemure’s peri-insular hemispherotomy technique.29

„„ Special Considerations Regarding the Hemispheric Surgical Approaches Anatomical hemispherectomy has still been used in some centers thanks to significantly decreased complication rates in modern patient series.17,​18,​23 However, many complications and concerns remain. Compared with other hemispherectomy techniques, anatomical hemispherectomy is associated with higher blood loss, risk of coagulopathy, longer hospital stays, higher rates of hydrocephalus, lower seizure-free rate in infarction cases, and risk of sagittal sinus thrombosis with extensive midline exposure.7,​28,​33 Conversely, the lowest reoperation rate is seen with this technique. In the Cleveland Clinic Foundation series, all hemimegalencephaly patients undergoing functional hemispherectomy had persistent seizures at some degree, although this was not the case with hemimegalencephaly patients undergoing anatomical hemispherectomy, all of who became seizure free.17,​34 Anatomical hemispherectomy may also be more appropriate for some multilobar cortical dysplasia and hemimegalencephaly patients because of the difficulty of disconnection procedures in these cases secondary to a hypoplastic temporal horn and irregular ventricles, abnormally thick cortical mantle, distorted anatomical landmarks with an abnormal ACA trajectory, and malformed Sylvian fissure with abnormally large veins. Hemidecortication technique is also generally associated with a significant number of complications, including higher blood loss, and higher infection rates. It is also associated with a higher risk of incomplete disconnection because of difficult anatomical orientation.7,​19,​35,​36 Because functional hemispherectomy techniques are associated with considerable cortical resection, these techniques are more appropriate for multilobar cortical dysplasia and hemimegalencephaly cases. By creating more space, these techniques avoid complications secondary to postoperative cerebral swelling. Hemispherotomy techniques have many advantages, such as requiring a smaller craniotomy, shorter operative time,

and intensive care unit stay than other techniques. It is also ­associated with a decreased risk of infection, lower hydrocephalus rates, and less traumatic results for the patient.4,​25,​26,​37 The limitations of hemispherotomy techniques are smaller exposure, challenges with transventricular access route, postoperative brain swelling, difficult anatomical orientation, difficulty of confirming full disconnection even with postoperative MRI, and higher reoperation rates. Hemispherotomy techniques are most ideal for atrophic hemispheres with enlarged ventricles but can be performed by experienced hands in all cases.4,​11,​30,​37

„„ Complications Although complication rates are much lower in modern series than in earlier series, hemispherectomy and its variations remain extremely extensive procedures that are associated with a significant risk of complications and even death. Although the vast majority of these patients present with significant hemiparesis, they may still experience increase in their motor deficits after surgery. Postoperative increase in motor weakness is ­frequently temporary and generally not seen if the damage happened before 3 years of age. Most patients become ambulatory postoperatively, but distal weakness may persist and digital dexterity never returns. If the patient has only partial hemianopsia preoperatively, complete hemianopsia after surgery is inevitable. Another common problem is hydrocephalus, which has been reported in the range from 0 to 28% in large series.5,​38,​39,​40 One of the largest retrospective multi-institutional studies found postoperative hydrocephalus rate of 23% among 736 patients who underwent hemispherectomy surgeries.41 Meningitis, aseptic meningitis, blood loss, hypovolemia, coagulation problems, and disseminated intravascular coagulation—most likely related to thromboplastin release secondary to extensive ­disconnection— are other potential complications associated with hemispherectomy procedures. Mortality rates between 0 and 5.7% in recent patient series have been reported. Brain shift, brainstem lesions, hypovolemia secondary to excessive blood loss, hemodynamic instability, brain swelling, hypoxia, and infection are reported as mortality reasons.4,​11,​13,​18,​22,​25,​28,​29,​32,​35,​37,​42,​43,​44

„„ Outcomes Almost all hemispherectomy techniques provide high rate of seizure-free outcomes. The most common factor affecting the outcome appears to be the pathology. The best outcomes are seen in patients with Rasmussen’s encephalitis and Sturge-­ Weber syndrome; the lowest seizure-free outcome rates are seen in hemimegalencephaly and multilobar cortical dysplasia cases.5,​39,​40,​41 An extensive overview of outcome results of several epilepsy surgery centers based on etiology shows the lowest seizure-freedom rate occurring in the cortical dysplasia group (56.5%). Broken down, this rate was 77% in patients with R ­ asmussen’s encephalitis, 82.1% in Sturge-Weber syndrome, 76.1% in vascular insults, and 77.3% in hemiatrophy patients.36 A systematic review of previously published 29 hemispherectomy series comprising 1,161 pediatric epilepsy

503

504

IVe  Hemispheric Surgery Techniques patients found the overall rate of seizure freedom in this population as 73.4%.45 A comparison of outcomes based on the surgical technique is much more difficult to make because of differences in the experience levels of the surgeons and differences in pathologies. One of the largest available data source is a combined patient pool from multiple centers and includes 333 hemispherectomy cases.36 The overall seizure-free outcome in this group was 70.4%. The highest seizure-free outcome was seen in the hemispherotomy group (85.7%), followed by modified Adams’ hemispherectomy (78.3%), functional hemispherectomy (66.1%), anatomical hemispherectomy (64.3%), and hemidecortication (60.7%) groups. In another large series,28 the overall seizure-free outcome rate was found to be 79%, with the lowest seizure-free rate occurring in the anatomical hemispherectomy group (59%). The seizure-freedom rates were 73% in the functional hemispherectomy group and 83% in the hemispherotomy group in the same study. Surgical failure after hemispherectomy might be related to several factors, including incomplete disconnection, preserved insular cortex, or the presence of active epileptogenic foci in the opposite hemisphere. Incomplete disconnection might be related to the presence of still-connected cortical tissue to the contralateral hemisphere, to the ipsilateral basal ganglia and thalamus, or directly to descending fibers. Most common incomplete disconnections were found at the basal frontal and insular regions in one study.46 Postoperative neurological deficits are another critical topic as we discussed earlier. Moosa et al reviewed changes in motor

function of 115 children following hemispherectomy with an average follow-up of 6 years.47 Eighty-three percent were able to walk unaided, an increase of 21% from the preoperative cohort. Sixty-four percent had unchanged or improved hemiparesis. Seventy percent had satisfactory spoken language skills. Timing of surgery is now accepted as a critical factor; the shorter the time between seizure onset and surgery, the higher the s­ uccess rate. This is partially related to positive effect of earlier seizure control on psychosocial and cognitive development, and partially related to potential help of brain plasticity on speech and motor function before the sixth to eighth years of age.2,​3,​48

„„ Conclusion Hemispherectomy and its variations provide very satisfac­ tory seizure control with a relatively low complication rate in a very challenging pediatric patient population. In published series, comparing outcomes based on different surgical techniques is not straightforward because of differences in the level of surgical expertise, case numbers, patient selection criteria, duration of seizures, and even etiologies.5 However, pathology appears to be the most significant determinant of outcome, and incomplete disconnection appears to be the most frequent reason for surgical failure. The choice of ­hemispherectomy technique is largely related to the ­surgeon’s training and experience with a particular technique. The type of the pathology may also dictate the surgical approach.

References 1. Krynauw RA. Infantile hemiplegia treated by removing one cerebra l hemisphere. J Neurol Neurosurg Psychiatry 1950;13(4):243–267 2. Dandy WE. Removal of right cerebral hemisphere for certain tumors with hemiplegia. JAMA 1928;90(11):823–825 3. Rasmussen T. Hemispherectomy for seizures revisited. Can J ­Neurol Sci 1983;10(2):71–78 4. Delalande O, Bulteau C, Dellatolas G, et al. Vertical parasagittal hemispherotomy: surgical procedures and clinical longterm outcomes in a population of 83 children. Neurosurgery 2007;60(2, Suppl 1):ONS19–ONS32, discussion ONS32 5. De Ribaupierre S, Delalande O. Hemispherotomy and other disconnective techniques. Neurosurg Focus 2008;25(3):E14 6. Villemure JG, Mascott CR. Peri-insular hemispherotomy: surgical principles and anatomy. Neurosurgery 1995;37(5):975–981 7. Schramm J, Behrens E, Entzian W. Hemispherical deafferentation: an alternative to functional hemispherectomy. Neurosurgery 1995;36(3):509–515, discussion 515–516 8. Daniel RT, Villemure JG. Peri-insular hemispherotomy: potential pitfalls and avoidance of complications. Stereotact Funct Neurosurg 2003;80(1–4):22–27 9. Ignelzi RJ, Bucy PC. Cerebral hemidecortication in the treatment of infantile cerebral hemiatrophy. J Nerv Ment Dis 1968;147(1):14–30 10. Oppenheimer DR, Griffith HB. Persistent intracranial bleeding as a complication of hemispherectomy. J Neurol Neurosurg Psychiatry 1966;29(3):229–240 11. Villemure JG, Daniel RT. Peri-insular hemispherotomy in paediatric epilepsy. Childs Nerv Syst 2006;22(8):967–981 12. Arzimanoglou A. Major types of epileptic seizures in childhood and corresponding epileptic syndromes. In: Arzimanoglou A, Guerrini R, Aicardi J, eds. Aicardi’s Epilepsy in Children. 3rd ed.

Philadelphia, PA: Lippincott Williams & Wilkins; 2004:12–37, 167–168, 290–292, 308–310 13. Comair YG. Transsylvian functional hemispherectomy: patient selection and results. In: Luders HO, Comair YG, eds. Epilepsy Surgery. 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001:699–704 14. Montes JL, Farmer JP, Andermann F, Poulin C. Hemispherectomy. In: Wyllie E, ed. The Treatment of Epilepsy: Principles and Practice. 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001:1147–1159 15. Graveline C, Hwang PA, Fitzpatrick T, Jay V, Hoffman HJ. Sturge-­ Weber syndrome: implications of functional studies on neural plasticity, brain maturation, and timing of surgical treatment. In: Kotagal P, Luders HO, eds. The Epilepsies: Etiologies and Prevention. San Diego, CA: Academic Press; 1999:61–70 16. Bulteau C, Otsuki T, Delalande O. Epilepsy surgery for hemispheric syndromes in infants: hemimegalencephaly and hemispheric cortical dysplasia. Brain Dev 2013;35(8):742–747 17. Nagel SJ, Elbabaa SK, Hadar EJ, Bingaman WE. Hemispherectomy techniques. In: Luders HO, ed. Textbook of Epilepsy Surgery. ­London, UK: Informa; 2008:1121–1129 18. Peacock WJ. Hemispherectomy for the treatment of intractable seizures in childhood. Neurosurg Clin N Am 1995;6(3):549–563 19. Carson BS, Javedan SP, Freeman JM, et al. Hemispherectomy: a hemidecortication approach and review of 52 cases. J Neurosurg 1996;84(6):903–911 20. González-Martínez JA, Gupta A, Kotagal P, et al. Hemispherectomy for catastrophic epilepsy in infants. Epilepsia 2005;46(9):1518–1525 21. Peacock WJ, Wehby-Grant MC, Shields WD, et al. Hemispherectomy for intractable seizures in children: a report of 58 cases. Childs Nerv Syst 1996;12(7):376–384

53  Hemispherectomy and Hemispherotomy Techniques in Pediatric Epilepsy Surgery 22. Davies KG, Maxwell RE, French LA. Hemispherectomy for intractable seizures: long-term results in 17 patients followed for up to 38 years. J Neurosurg 1993;78(5):733–740 23. Di Rocco C, Iannelli A. Hemimegalencephaly and intractable epilepsy: complications of hemispherectomy and their correlations with the surgical technique. A report on 15 cases. Pediatr Neurosurg 2000;33(4):198–207 24. Winston KR, Welch K, Adler JR, Erba G. Cerebral hemicorticectomy for epilepsy. J Neurosurg 1992;77(6):889–895 25. Schramm J, Kral T, Clusmann H. Transsylvian keyhole functional hemispherectomy. Neurosurgery 2001;49(4):891–900, discussion 900–901 26. Villemure JG, Vernet O, Delalande O. Hemispheric disconnection: callosotomy and hemispherotomy. Adv Tech Stand Neurosurg 2000;26:25–78 27. Wen HT, Rhoton AL Jr, Marino R Jr. Anatomical landmarks for hemispherotomy and their clinical application. J Neurosurg 2004;101(5):747–755 28. Cook SW, Nguyen ST, Hu B, et al. Cerebral hemispherectomy in pediatric patients with epilepsy: comparison of three techniques by pathological substrate in 115 patients. J Neurosurg 2004;100(2, Suppl Pediatrics):125–141 29. Shimizu H, Maehara T. Modification of peri-insular hemispherotomy and surgical results. Neurosurgery 2000;47(2):367–372, discussion 372–373 30. Dorfer C, Czech T, Dressler A, et al. Vertical perithalamic hemispherotomy: a single-center experience in 40 pediatric patients with epilepsy. Epilepsia 2013;54(11):1905–1912 31. Schramm J, Kuczaty S, Sassen R, Elger CE, von Lehe M. Pediatric functional hemispherectomy: outcome in 92 patients. Acta Neurochir (Wien) 2012;154(11):2017–2028 32. Kestle J, Connolly M, Cochrane D. Pediatric peri-insular hemispherotomy. Pediatr Neurosurg 2000;32(1):44–47 33. Brian JE Jr, Deshpande JK, McPherson RW. Management of cerebral hemispherectomy in children. J Clin Anesth 1990;2(2):91–95 34. Carreño M, Wyllie E, Bingaman W, Kotagal P, Comair Y, Ruggieri P. Seizure outcome after functional hemispherectomy for malformations of cortical development. Neurology 2001;57(2):331–333 35. Kossoff EH, Vining EP, Pyzik PL, et al. The postoperative course and management of 106 hemidecortications. Pediatr Neurosurg 2002;37(6):298–303

36. Holthausen H, May TW, Adams TB, et al. Seizures post hemispherectomy. In: Tuxhorn I, Holthausen H, Boenigk H, eds. Pediatric Epilepsy Syndromes and Their Surgical Treatments. London, UK: John Libbey; 1997:749–773 37. Binder DK, Schramm J. Transsylvian functional hemispherectomy. Childs Nerv Syst 2006;22(8):960–966 38. Vadera S, Griffith SD, Rosenbaum BP, et al. National trends and in-hospital complication rates in more than 1600 hemispherectomies from 1988 to 2010: a Nationwide Inpatient Sample Study. Neurosurgery 2015;77(2):185–191, discussion 191 39. Lin Y, Harris DA, Curry DJ, Lam S. Trends in outcomes, complications, and hospitalization costs for hemispherectomy in the United States for the years 2000–2009. Epilepsia 2015;56(1):139–146 40. Lew SM, Koop JI, Mueller WM, Matthews AE, Mallonee JC. Fifty consecutive hemispherectomies: outcomes, evolution of technique, complications, and lessons learned. Neurosurgery 2014;74(2):182–194, discussion 195 41. Lew SM, Matthews AE, Hartman AL, Haranhalli N; Post-­ Hemispherectomy Hydrocephalus Workgroup. Posthemispherectomy hydrocephalus: results of a comprehensive, multiinstitutional review. Epilepsia 2013;54(2):383–389 42. De Almeida AN, Marino R Jr, Aguiar PH, Jacobsen Teixeira M. Hemispherectomy: a schematic review of the current techniques. Neurosurg Rev 2006;29(2):97–102, discussion 102 43. Devlin AM, Cross JH, Harkness W, et al. Clinical outcomes of hemispherectomy for epilepsy in childhood and adolescence. Brain 2003;126(Pt 3):556–566 44. Villemure JG, Adams CBT, Hoffman HJ, Peacock WJ. ­Hemispherectomy. In: Engel Jr. J, ed. Surgical Treatment of the Epilepsies. 2nd ed. New York, NY: Raven Press; 1993:511–518 45. Griessenauer CJ, Salam S, Hendrix P, et al. Hemispherectomy for treatment of refractory epilepsy in the pediatric age group: a systematic review. J Neurosurg Pediatr 2015;15(1):34–44 46. Vadera S, Moosa ANV, Jehi L, et al. Reoperative hemispherectomy for intractable epilepsy: a report of 36 patients. Neurosurgery 2012;71(2):388–392, discussion 392–393 47. Moosa AN, Jehi L, Marashly A, et al. Long-term functional outcomes and their predictors after hemispherectomy in 115 children. Epilepsia 2013;54(10):1771–1779 48. Beier AD, Rutka JT. Hemispherectomy: historical review and recent technical advances. Neurosurg Focus 2013;34(6):E11

505

54

  Multifocal Epilepsy and Multilobar Resections Rafael Uribe, George I. Jallo, and Caitlin Hoffman

Summary The concept of multifocality in epilepsy is inherently contrary to the principles of epilepsy surgery. Most patients with multifocal or generalized epilepsies are usually only candidates for palliative treatments like vagal nerve stimulation or corpus callosotomy. Experience in patients with tuberous sclerosis has paved the way to understanding that patients with multifocal epilepsy can be treated by focal resections. Nonetheless, the literature on multilobar resections is still limited by the heterogeneity of conditions treated, the variable surgical approaches that are reported, and the small size of patient series. Most reports of multilobar resections are embedded in larger series of pediatric epilepsy, extratemporal epilepsy, and focal cortical dysplasia. The concept of multifocality in epilepsy can apply to different theoretical clinical scenarios. Keywords:  multifocal, seizures, surgery

„„ Introduction The purpose of the presurgical workup undertaken in patients with refractory epilepsy is to assess if seizures originate from a discrete area of the brain that can be ultimately resected.1 Ideal cases are those in which: • The epileptogenic zone has a discrete anatomical location. yy A lesion is visible on MRI in a “non-eloquent” area. yy Seizure semiology is consistent with known clinical ­correlates. yy The electrophysiological, clinical, and imaging data are all concordant. However, that ideal scenario is more often the exception than the rule. This is particularly true in pediatric patients where extratemporal lesions and dual pathology are more common2 and the pattern of seizure spread is faster due to varying myelination patterns. The concept of multifocality in epilepsy is inherently contrary to the principles of epilepsy surgery. Most patients with multifocal or generalized epilepsies are usually only candidates for palliative treatments like vagal nerve stimulation or corpus callosotomy. Experience in patients with tuberous sclerosis3,​4,​5 has paved the way to understanding that patients with multifocal epilepsy can be treated by focal resections. Nonetheless, the literature on multilobar resections is still limited

by the heterogeneity of conditions treated, the variable surgical approaches that are reported, and the small size of patient series. Most reports of multilobar resections are embedded in larger series of pediatric epilepsy,6,​7,​8,​9,​10 extratemporal epilepsy,11 and focal cortical dysplasia12 as noted by Sarkis et al.13 The concept of multifocality in epilepsy can apply to different theoretical clinical scenarios: • Patients with more than one independent seizure focus within a single lobe. yy Patients with multiple independent foci that affect more than one lobe in the same hemisphere. yy Patients with a single focus that spans more than one lobe in one hemisphere. yy Patients with multiple independent foci in both ­hemispheres. Multiple anatomical lesions do not necessarily equate with multifocal epileptiform discharges on surface electroencephalography (EEG). A good example is tuberous sclerosis, where the background can be normal or show only focal slowing over cortical tubers. Major et al showed through direct electrographic recording of cortical tubers that some of those lesions are electrically silent and only the surrounding parenchyma is epileptogenic.14 On the other hand, lesions that span multiple lobes can also present with focal semiology restricted to a single lobe. Catenoix et al15 presented a series of seven patients with multilobar lesions who presented with mesiotemporal semiology; all patients were interrogated using stereoelectroencephalography (SEEG) and found to have mesial temporal discharges. All patients in that series underwent temporal lobectomy and were classified as Engel class I after surgery, with a mean follow-up time of 37.4 months.15 For these reasons, multifocal epilepsy presents a unique challenge for the epilepsy surgery team. In this chapter, we will review the indications and outcomes of multilobar resections in patients with refractory epilepsy.

„„ Indications It is important to define the goal of surgery in patients with suspected multifocal disease. If the patient has multiple types of seizures, the goal of surgery is not to eliminate all seizures but only the most disabling ones. In patients with only one predominant semiology, complete elimination of seizures might

54  Multifocal Epilepsy and Multilobar Resections be attainable if a focal epileptogenic zone can be proven to be responsible for the symptoms, even in patients with multilobar lesions. The expectations must be discussed with the family and the patient prior to undertaking any invasive testing. The conditions that require multilobar resections are wide ranging and include perinatal stroke, congenital malformations (e.g., Sturge-Weber syndrome), cortical dysplasia, tumors, neuronal migration disorders, head trauma, and vascular anomalies.13,​16 What separates patients who are candidates for a hemispherectomy from those who only require a sub-hemispheric resection is the degree of remaining neurological function in the affected hemisphere.17

„„ Epidemiology Defining the epidemiology of multifocal epilepsy treated by multilobar resections is extremely challenging. None of the largest published case series of multilobar resections reach more than 100 patients. The series published by Cho et al looked at the prognostic role of functional neuroimaging in a sample of 90 patients who underwent multilobar resection in 18-year period (1995–2013).18 Sarkis et al published the Cleveland Clinic’s experience with multilobar resections from 1994 to 2010. During this 16-year period, 63 patients were treated with multilobar resections.13 UCLA published their experience with multilobar resections from 1986 to 2000 which included 52 patients.19 Cossu et al published the experience in Milan from 1996 to 2004, and identified 21 patients who underwent multilobar resections from a total cohort of 113 pediatric patients.7 In addition to sample size, patient characteristics and underlying etiologies vary within published series. A precise incidence of multilobar resections is not possible to calculate reliably. Series like the one published by Maillard et al detail clearly the proportion of patients treated with multilobar resections,20 while others like Paolicchi et al only mention seizure outcomes without giving a detailed account of the types of resections that were performed.6

„„ Workup The workup of patients with multifocal epilepsy who are potential candidates for resective surgery follows the same algorithm described for other patients with medically intractable epilepsy. A detailed history and exam, followed by dedicated neuroimaging and long-term video EEG, will constitute the first phase of evaluation.21 Once the first phase of the study is completed, the epilepsy team will review the available noninvasive data to formulate a hypothesis about the possible ictal onset zone. The use of advanced imaging techniques including PET, MEG, and SPECT will be considered if a candidate seizure focus that requires additional testing has been identified.22 Different implantation strategies have been proposed including SEEG15 and subdural grids and strips combined with depth electrodes. Weiner and co-workers proposed a three-stage procedure in patients with multifocal epilepsy, where patients are reimplanted with grids and strips after focal resection, to assess the need for resection of additional foci.1,​23 Both methodologies have relative advantages and disadvantages. SEEG affords a less invasive method to sample bilaterally and spares patients from

undergoing large craniotomies, with the potential of tailoring a smaller craniotomy at the time of lesion resection. Functional mapping, however, is limited with SEEG.24

„„ Surgical Strategies The surgical approaches to multilobar resections vary depending on the extent of the lesion (if one is clearly visible) and/or the ­ ­ epileptogenic zone. The approaches listed by Sarkis et al are the most comprehensive and include ­frontotemporal, temporoparietal, frontoparietal, and occipital plus (­temporo-parieto-occipital [PTO], parieto-occipital, temporo-occipital).13 Occipital-plus resections were the most commonly reported in their series (57%), followed by frontotemporal resections (21%).13 Nilsson et al found that extended frontal resections (frontotemporal, frontoparietal, frontotemporo-parietal) were the most common types of procedures (20/57, 35%) in their series.25 Posterior quadrant resections (particularly TPO) were the most common in the series reported by Leiphart et al (46%).19 These large series suggest that posterior quadrant epilepsy accounts for the most common type of multilobar resections. The morbidity of multilobar resections is high compared to less invasive surgical procedures. Sarkis et al reported new or worsened visual field defects in 19% of patients and a new or worsened motor deficit in 11%.13 Nilsson et al reported a 5.3% rate of major complications including one MCA infarct, one epidural abscess, and one patient with hemiparesis. Minor complications occurred in 17.5% of patients including wound infections, transient neurological deficits, urinary tract infection, shunt malfunction, and a cerebrospinal fluid leak.25 Leiphart et al did not mention complication rates in their surgical series.19 Serletis et al reported no complications in their series of 11 patients who underwent multilobar resections for orbitofrontal pathology.26 Shaver et al presented a series of pediatric patients who were treated with reoperation after an initial failed epilepsy surgery. Out of twenty included patients, four underwent multilobar resections. Three of those patients had associated surgical complications and therefore the authors conclude that patients undergoing multilobar surgery after an initial failed resection have a higher morbidity rate.27 The large series published by Cho reported a rate of complication of 15.5% (14 patients) including four cases with persistent paresis, four with visual field defects (not specified if new or worsened), four cases with mild language disturbances, three cases of surgical site infection that required drainage, and one case of prosopagnosia.18 The use of advanced imaging techniques include PET, MEG, and SPECT are indicated prior to a surgical intervention to localize the seizure focus.

„„ Pathology The most common pathological substrate reported in the major series corresponds to malformations of cortical development (MCD). Forty-six percent of the samples in Sarkis et al’ series corresponded to MCDs.13 Fifty-four percent of the cases in the UCLA series corresponded to cortical dysplasia.19 The Swedish series reported cortical malformations as the most common etiology, being found in 28.1% of the analyzed samples.25

507

508

IVe  Hemispheric Surgery Techniques

„„ Outcomes There is a paucity of data regarding outcomes of patients undergoing multilobar resections due to the heterogeneity of presentation, underlying diagnosis, involved cortical areas, and approach to diagnosis and treatment. The greatest body of literature stems from the patients with posterior quadrant epilepsy. A recent systematic review by Harward et al looked at seizure outcomes in that patient population. A total of 27 studies were included for analysis, totaling 584 patients with a mean seizure freedom rate of 65% (range: 25–100%). Seizure freedom was most commonly achieved in pediatric patients (2 y

3.9 y (1 mo to 10 y)

7.1 y (1.7–13.9 y)

Complete Results FU period

N = 74

N = 119

N = 75

No ↓seizure frequency

13 (17.6%)

24 (20.2%)

17 (22.7%)

>50% ↓seizure frequency

49 (66.2%)

70 (58.8%)

48 (64.0%)

14 (18.9%)

21 (17.6%)

>75% ↓seizure frequency Seizure-free with AED

29 (38.7%) 7 (9.3%)a

Abbreviations: LGS, Lennox–Gastaut syndrome; IS-LGS, infantile spasms evolving to Lennox–Gastaut syndrome; IS, infantile spasms; SE-MISG, severe epilepsy multiple independent spike foci; HHE, hemiconvulsions hemiplegia epilepsy syndrome; CPS2G, complex partial seizures with secondary ­generalization; PS ± 2G, partial seizure with/without secondary generalization; 2G?, difficult to be classified but possible secondary generalized epilepsy. Source: Reproduced with permission from Wong et al.13 a Out of 75 patients, 7 reached stable seizure free at average of 5.3 years (1.3–12 years).

62  Corpus Callosotomy 1999 with average follow-up period of 3.9 years, and 75 cases operated between 1991 and 2003 with average follow-up of 7.1 years. Constantly, over half of the three groups of patients reviewed in different periods and follow-up durations had more than 50% reduction of seizure frequency. The rates of seizure free patients with anticonvulsants dropped from 18.9 to 9.3% in the patient group with longer follow-up periods. It means that after callosotomy, seizure control still fluctuates in many of the patients.

„„ Conclusion The concept and technique of corpus callosotomy has been well established since the 1980s. It is a palliative procedure that is applied predominantly in children with LGS and drop attacks. Compared with anterior callosotomy, single-stage complete callosotomy is more effective with a broader spectrum of seizure control including drop attacks. A recent new approach,

the selective posterior callosotomy was reported in 2016 with effective control of drop attacks.41 Single-stage complete callosotomy is applied in prepubertal children and in older children with severe developmental delay. In suitable selected patients, after callosotomy, ­ significant neuropsychological changes are evident only on formal testing. These neuropsychological changes are usually ignored by patient and family members and rarely affect daily life. In patients without diffuse cerebral abnormality and EEG epileptic activities limiting over the anterior cerebral hemispheres, it may be suitable to select anterior callosotomy or staged complete callosotomy. Improved seizure control after callosotomy is always associated with improved quality of life of the patients, satisfaction of parents, and improvement in the family’s quality of life. The surgical procedure is quite safe in experienced hands, with minimal surgical complications and acceptable adverse events. Early identification of medically resistant epilepsy and appropriate selection of the surgical candidates for corpus callosotomy are important for satisfactory outcomes.

References 1. Berg AT, Shinnar S, Levy SR, Testa FM, Smith-Rapaport S, ­Beckerman B. Early development of intractable epilepsy in children: a prospective study. Neurology 2001;56(11):1445–1452

15. Tanriverdi T, Olivier A, Poulin N, Andermann F, Dubeau F. Longterm seizure outcome after corpus callosotomy: a retrospective analysis of 95 patients. J Neurosurg 2009;110(2):332–342

2. Sillanpää M, Schmidt D. Natural history of treated childhood-­ onset epilepsy: prospective, long-term population-based study. Brain 2006;129(Pt 3):617–624

16. Pinard JM, Delalande O, Chiron C, et al. Callosotomy for epilepsy after West syndrome. Epilepsia 1999;40(12):1727–1734

3. Ramos-Lizana J, Rodriguez-Lucenilla MI, Aguilera-López P, Aguirre-Rodríguez J, Cassinello-García E. A study of drug-­resistant childhood epilepsy testing the new ILAE criteria. Seizure 2012;21(4):266–272 4. Rosati A, De Masi S, Guerrini R, De Masi S, Guerrini R. Antiepileptic drug treatment in children with epilepsy. CNS Drugs 2015;29(10):847–863 5. Lambrechts DAJE, de Kinderen RJA, Vles JSH, de Louw AJ, ­Aldenkamp AP, Majoie HJM. A randomized controlled trial of the ketogenic diet in refractory childhood epilepsy. Acta Neurol Scand 2017;135(6):678 6. Kellermann TS, Wagner JL, Smith G, Karia S, Eskandari R. Surgical management of pediatric epilepsy: decision-making and outcomes. Pediatr Neurol 2016;64:21–31 7. Sugano H, Arai H. Epilepsy surgery for pediatric epilepsy: optimal timing of surgical intervention. Neurol Med Chir (Tokyo) 2015;55(5):399–406 8. Agadi S, Quach MM, Haneef Z. Vitamin-responsive epileptic encephalopathies in children. Epilepsy Res Treat 2013;2013:510529 9. Zimmerman RS, Sirven JI. An overview of surgery for chronic seizures. Mayo Clin Proc 2003;78(1):109–117 10. Benifla M, Rutka JT, Logan W, Donner EJ. Vagal nerve stimulation for refractory epilepsy in children: indications and experience at The Hospital for Sick Children. Childs Nerv Syst 2006;22(8):1018–1026 11. Asadi-Pooya AA, Sharan A, Nei M, Sperling MR. Corpus callosotomy. Epilepsy Behav 2008;13(2):271–278 12. Graham D, Tisdall MM, Gill D. Corpus callosotomy outcomes in pediatric patients: A systematic review. Epilepsia 2016;57(7):1053–1068 13. Wong TT, Kwan SY, Chang KP, et al. Corpus callosotomy in children. Childs Nerv Syst 2006;22(8):999–1011 14. Wheless JW. Managing severe epilepsy syndromes of early childhood. J Child Neurol 2009;24(8, Suppl):24S–32S, quiz 33S–36S

17. Kasasbeh AS, Smyth MD, Steger-May K, Jalilian L, Bertrand M, ­ Limbrick DD. Outcomes after anterior or complete corpus ­ callosotomy in children. Neurosurgery 2014;74(1):17–28, ­discussion 28 18. Asadi-Pooya AA, Malekmohamadi Z, Kamgarpour A, et al. Corpus callosotomy is a valuable therapeutic option for patients with Lennox-Gastaut syndrome and medically refractory seizures. Epilepsy Behav 2013;29(2):285–288 19. Luat AF, Asano E, Kumar A, Chugani HT, Sood S. Corpus callosotomy for intractable epilepsy revisited: the Children’s Hospital of Michigan Series. J Child Neurol 2017;32(7):624–629 20. Douglass LM, Salpekar J. Surgical options for patients with ­Lennox-Gastaut syndrome. Epilepsia 2014;55(Suppl 4):21–28 21. Liang S, Zhang S, Hu X, et al. Anterior corpus callosotomy in school-aged children with Lennox-Gastaut syndrome: a prospective study. Eur J Paediatr Neurol 2014;18(6):670–676 22. Rathore C, Abraham M, Rao RM, George A, Sankara Sarma P, Radhakrishnan K. Outcome after corpus callosotomy in children with injurious drop attacks and severe mental retardation. Brain Dev 2007;29(9):577–585 23. Turanli G, Yalnizoğlu D, Genç-Açikgöz D, Akalan N, Topçu M. Outcome and long term follow-up after corpus callosotomy in childhood onset intractable epilepsy. Childs Nerv Syst 2006;22(10):1322–1327 24. Oguni H, Olivier A, Andermann F, Comair J. Anterior callosotomy in the treatment of medically intractable epilepsies: a study of 43 patients with a mean follow-up of 39 months. Ann Neurol 1991;30(3):357–364 25. Fuiks KS, Wyler AR, Hermann BP, Somes G. Seizure outcome from anterior and complete corpus callosotomy. J Neurosurg 1991;74(4):573–578 26. Spencer SS, Spencer DD, Sass K, Westerveld M, Katz A, Mattson R. Anterior, total, and two-stage corpus callosum section: differential and incremental seizure responses. Epilepsia 1993;34(3):561–567 27. Jalilian L, Limbrick DD, Steger-May K, Johnston J, Powers AK, Smyth MD. Complete versus anterior two-thirds corpus cal-

587

588

IVf  Other Disconnective Procedures losotomy in children: analysis of outcome. J Neurosurg Pediatr 2010;6(3):257–266 28. Sauerwein HC, Lassonde M. Neuropsychological alterations after split-brain surgery. J Neurosurg Sci 1997;41(1):59–66 29. Schaller K, Cabrilo I. Corpus callosotomy. Acta Neurochir (Wien) 2016;158(1):155–160 30. Oguni H, Andermann F, Gotman J, Olivier A. Effect of anterior callosotomy on bilaterally synchronous spike and wave and other EEG discharges. Epilepsia 1994;35(3):505–513 31. Clarke DF, Wheless JW, Chacon MM, et al. Corpus callosotomy: a palliative therapeutic technique may help identify resectable epileptogenic foci. Seizure 2007;16(6):545–553 32. Hur YJ, Kang HC, Kim DS, Choi SR, Kim HD, Lee JS. Uncovered primary seizure foci in Lennox-Gastaut syndrome after corpus callosotomy. Brain Dev 2011;33(8):672–677 33. Van Wagenen WP, Herren RY. Surgical division of commissural pathways in the corpus callosum: relation to spread of an epileptic attack. Arch Neurol Psychiatry 1940;44(4):740–759 34. Mathews MS, Linskey ME, Binder DK. William P. van Wagenen and the first corpus callosotomies for epilepsy. J Neurosurg 2008;108(3):608–613 35. Bogen JE, Fisher ED, Vogel PJ. Cerebral commissurotomy. A second case report. JAMA 1965;194(12):1328–1329 36. Luessenhop AJ, Dela Cruz TC, Fenichel GM. Surgical disconnection of the cerebral hemispheres for intractable seizures. Results in infancy and childhood. JAMA 1970;213(10):1630–1636 37. Wilson DH, Reeves A, Gazzaniga M. Division of the corpus callosum for uncontrollable epilepsy. Neurology 1978;28(7):649–653 38. Wilson DH, Reeves AG, Gazzaniga MS. “Central” commissurotomy for intractable generalized epilepsy: series two. Neurology 1982;32(7):687–697 39. Maxwell RE, Gate JR, Gumnit RJ. Corpus callosotomy at the ­University of Minnesota. In: Engel Jr. J, ed. Surgical Treatment of the Epilepsy. New York, NY: Raven Press; 1986:659–666 40. Wyler AR. Corpus callosotomy. In: Wyllie E, ed. The Treatment of Epilepsy: Principles and Practice. Philadelphia, PA; London, UK: Lea & Febiger; 1993:1120–1125 41. Paglioli E, Martins WA, Azambuja N, et al. Selective posterior callosotomy for drop attacks: a new approach sparing prefrontal connectivity. Neurology 2016;87(19):1968–1974 42. Raybaud C. The corpus callosum, the other great forebrain commissures, and the septum pellucidum: anatomy, development, and malformation. Neuroradiology 2010;52(6):447–477 43. Hofer S, Frahm J. Topography of the human corpus c­allosum revisited—comprehensive fiber tractography using d ­ ­iffusion tensor magnetic resonance imaging. Neuroimage 2006; 32(3):989–994 44. Feichtinger M, Schröttner O, Eder H, et al. Efficacy and safety of radiosurgical callosotomy: a retrospective analysis. Epilepsia 2006;47(7):1184–1191 45. Bodaghabadi M, Bitaraf MA, Aran S, et al. Corpus callosotomy with gamma knife radiosurgery for a case of intractable generalised epilepsy. Epileptic Disord 2011;13(2):202–208 46. Pendl G, Eder HG, Schroettner O, Leber KA. Corpus callosotomy with radiosurgery. Neurosurgery 1999;45(2):303–307, discussion 307–308 47. Sood S, Marupudi NI, Asano E, Haridas A, Ham SD. Endoscopic corpus callosotomy and hemispherotomy. J Neurosurg Pediatr 2015;16(6):681–686 48. Chandra PS, Kurwale NS, Chibber SS, et al. Endoscopic-assisted (through a mini craniotomy) corpus callosotomy combined

with anterior, hippocampal, and posterior commissurotomy in L­ ennox-Gastaut syndrome: a pilot study to establish its safety and efficacy. Neurosurgery 2016;78(5):743–751 49. Smyth MD, Vellimana AK, Asano E, Sood S. Corpus callosotomy—open and endoscopic surgical techniques. Epilepsia 2017;58(Suppl 1):73–79 50. Park HJ, Kim JJ, Lee SK, et al. Corpus callosal connection mapping using cortical gray matter parcellation and DT-MRI. Hum Brain Mapp 2008;29(5):503–516 51. Chao YP, Cho KH, Yeh CH, Chou KH, Chen JH, Lin CP. Probabilistic topography of human corpus callosum using cytoarchitectural parcellation and high angular resolution diffusion imaging tractography. Hum Brain Mapp 2009;30(10): 3172–3187 52. Pannek K, Mathias JL, Bigler ED, Brown G, Taylor JD, Rose S. An automated strategy for the delineation and parcellation of commissural pathways suitable for clinical populations utilising high angular resolution diffusion imaging tractography. Neuroimage 2010;50(3):1044–1053 53. Fabri M, Pierpaoli C, Barbaresi P, Polonara G. Functional topography of the corpus callosum investigated by DTI and fMRI. World J Radiol 2014;6(12):895–906 54. Lassonde M, Sauerwein C. Neuropsychological outcome of corpus callosotomy in children and adolescents. J Neurosurg Sci 1997;41(1):67–73 55. Leen WG, Klepper J, Verbeek MM, et al. Glucose transporter-1 deficiency syndrome: the expanding clinical and genetic spectrum of a treatable disorder. Brain 2010;133(Pt 3):655–670 56. Ferrie CD, Patel A. Treatment of Lennox-Gastaut syndrome (LGS). Eur J Paediatr Neurol 2009;13(6):493–504 57. Lemmon ME, Terao NN, Ng YT, Reisig W, Rubenstein JE, Kossoff EH. Efficacy of the ketogenic diet in Lennox-Gastaut syndrome: a retrospective review of one institution’s experience and summary of the literature. Dev Med Child Neurol 2012;54(5): 464–468 58. Kossoff EH, Shields WD. Nonpharmacologic care for patients with Lennox-Gastaut syndrome: ketogenic diets and vagus nerve stimulation. Epilepsia 2014;55(Suppl 4):29–33 59. Caraballo RH, Fortini S, Fresler S, et al. Ketogenic diet in patients with Lennox-Gastaut syndrome. Seizure 2014;23(9):751–755 60. Arya R, Greiner HM, Horn PS, Turner M, Holland KD, Mangano FT. Corpus callosotomy for childhood-onset drug-resistant epilepsy unresponsive to vagus nerve stimulation. Pediatr Neurol 2014;51(6):800–805 61. Sass KJ, Novelly RA, Spencer DD, Spencer SS. Postcallosotomy language impairments in patients with crossed cerebral dominance. J Neurosurg 1990;72(1):85–90 62. Jea A, Vachhrajani S, Widjaja E, et al. Corpus callosotomy in children and the disconnection syndromes: a review. Childs Nerv Syst 2008;24(6):685–692 63. Lassonde M, Sauerwein H, Geoffroy G, Décarie M. Effects of early and late transection of the corpus callosum in children. A study of tactile and tactuomotor transfer and integration. Brain 1986;109(Pt 5):953–967 64. Cendes F, Ragazzo PC, da Costa V, Martins LF. Corpus callosotomy in treatment of medically resistant epilepsy: preliminary results in a pediatric population. Epilepsia 1993; 34(5):910–917 65. Pilcher WH, Roberts DW, Flanigin HF, et al. Complications of epilepsy surgery. In: Engel Jr. J, ed. Surgical Treatment of the ­ Epilepsies. 2nd ed. New York, NY: Raven Press; 1993:565–581

63

  Endoscope-Assisted Corpus Callosotomy with Anterior, Hippocampal, and Posterior Commissurotomy P. Sarat Chandra, Jitin Bajaj, Heri Subianto, and Manjari Tripathi

Summary Corpus callosotomy is an effective palliative surgery in drug refractory epilepsy. The surgery is performed after ruling out the patients for curative resections through extensive presurgical evaluation. The most common indication is Lennox-Gastaut syndrome, and best outcome is for drop attacks. Complete corpus callosotomy with pan commissural sectioning is safe and efficacious. Transient akinetic mutism, buccal apraxia, and disconnection syndrome are the major side effects. Endoscopic procedure is minimally invasive with less morbidity. Keywords:  corpus callosotomy, endoscopic, hemispherotomy, Lennox-Gastaut syndrome

„„ Introduction Corpus callosotomy (CC) is a “palliative” surgery aimed to reduce the burden of seizures in patients with nonlocalizing bihemispheric epilepsy.1,​2,​3,​4 It is perhaps the only procedure of choice for Lennox-Gastaut syndrome (LGS) with multiple drop attacks apart from vagal nerve stimulation.5,​6,​7,​8 The extent of CC remains controversial. Many authors advocate only anterior CC to prevent disconnection syndromes, while compromising the seizure outcome.3,​9 Some other authors feel that a posterior disconnection is better in view of larger volume of fibers crossing, faster recovery, and also the ability to divide the hippocampal commissure during this procedure.10 Approximately one-third of patients undergoing anterior callosotomy have seizure recurrence necessitating a second surgery to complete the callosal sectioning.1,​2,​11,​12,​13,​14,​15,​16,​17,​18 A complete CC while being more effective is still considered as a palliative procedure.11,​14 In our center, we have always performed a complete CC since the beginning of our epilepsy surgery program for the following three reasons: (1) we believe that a complete section of callosum would offer the best possible chance of seizure freedom; (2) it would be difficult for us to motivate most of our patients for a second surgery for staged callosal sectioning; (3) as recent studies have suggested, callosal sectioning in failed cases may

localize the epileptogenic foci. Thus, the number of surgeries would reduce if a complete callosal sectioning is performed in the first instance itself.11,​14,​19,​20,​21 Commissural sectioning (anterior commissurotomy [ACT]; hippocampal commissurotomy [HCT]; and posterior commissurotomy [PCT]) has been tried in late 1970s with success but abandoned due to morbidities (mostly related to disconnection) in few cases.22 In animal models, division of corpus callosum has been useful in the management of secondarily generalized epilepsies, but there is lack of homogeneity in experimental studies to see the effect of anterior hippocampal and posterior commissurotomy on the generalization of discharges.23 We also prefer to combine CC with anterior, middle, and posterior commissurotomy.24,​25 We do this especially for patients with LGS with multiple drop attacks and severe encephalopathy. The main reason is that we are providing the maximum “benefit” of “interhemispheric” disconnection for these patients which is the maximum chance of seizure freedom. We developed an endoscopic technique for performing CC (since 2012) also combined (if required) with anterior, middle, and posterior commissurotomy (pan-commissurotomy). In this chapter, we provide a brief description of this technique along with our results.

„„ Indications The indications of endoscopic CC are same as any indications for open microsurgical CC. However, it is important to ascertain the exact site of cranial opening especially to avoid large draining veins. In our setting, we have been performing a complete CC right from beginning for the reasons mentioned earlier. In addition, we prefer to perform a complete CC with anterior, hippocampal, and posterior commissurotomy in patients with LGS as per the following indications: • No single lateralization/localization of epileptiform zone/ network. yy Drop attacks being the predominant seizure type. yy Intelligent/Social quotients less than 50.

590

IVf  Other Disconnective Procedures yy High seizure frequency defined as at least once or twice per day. yy Parental consent for the procedure.

„„ Principles of Endoscope-Assisted Corpus Callosotomy and Anterior, Middle, and Posterior Commissurotomy 1. Cranial opening: The cranial size and opening is similar to that of endoscopic hemispherotomy. The cranial opening is also positioned more anterior to that in traditional open vertical parasagittal hemispherotomy and is performed with the help of neuronavigation to avoid encountering major bridging veins. 2. Interhemispheric approach: Unlike CC being performed in hemispherotomy, in corpus callostomy per se, the callosal sectioning is performed strictly in the midline through the cavum. The septae must be carefully split on either side and one must enter the cavum. The cavum is present even in cases where it is not seen on imaging. 3. Preserving vessels and normal structures: Unlike hemispherotomy, where a whole hemisphere may be sacrificed, one should remember that in CC both of the hemispheres are functioning in a highly compromised manner. Hence, any degree of vascular or brain damage should be avoided. More so due to epileptic encephalopathy, damage to midline vessels like septal vessels could lead to significant cognitive damage.16,​26 4. Avoid blood getting into ventricles: This is another important principle to remember. The surgeon should confine the surgery to the extraventricular space in the intercaval region. If the ventricle is entered inadvertently, a tailed cotton patty must be placed in the ventricle to prevent blood from trickling into the ventricle. If a posterior commissurotomy has been planned,24 then similarly a cotton patty must be placed in the aqueduct. Presence of blood may cause fever, which can again prolong hospital stay. 5. Sectioning of commissures along with CC: Complete callosal sectioning (first introduced by van Wagenen and Herren in 1940)27 is a very effective “palliative” procedure.1,​2,​11,​12,​16,​17 Various combinations of corpus callosotomies with commissurotomies have been tried in literature: complete CC + HCT,28 complete CC,29 complete CC + ACT,30 complete CC + HCT.31 Most of these authors reported 10 to 20% of primary nonresponders and close to 30% of patients further relapses in next few years with outcome mostly remaining stable thereafter.17,​18 6. Transmission of seizures through the commissures has been cited as the most common cause for failure of CC.18,​32,​33,​34 Using DTI, Jang and Kwon in 2013 and 2014 demonstrated a much wider connections of the fornix to include the cerebral cortex (precentral gyrus, postcentral gyrus, and posterior parietal cortex) and also the brainstem through the ­thalamus.35,​36,​37 Anterior commissure (AC) connects the highly epileptogenic temporal lobes32 while hippocampal commissure, mostly considered as rudimentary in humans, connects both the hippocampi (again highly e ­ pileptogenic)

and joins both the bodies of the fornices just under the posterior part of the body of the corpus callosum.6,​33 Thus, division of anterior and hippocampal commissure may be expected to have significant effect to reduce the seizure burden. The posterior commissure (PC) section has not been described till date for seizure control. The PC is known to carry pupilloconstrictor fibers to mediate bilateral light reflex. A detailed anatomical study using anterograde fiber degeneration and retrograde axon transport (using horse peroxidase) revealed a much larger distribution of PC to pars reticularis, periaqueductal grey matter connecting both cranial and caudal reticular ­system, ventral lateral geniculate nucleus, H field of Forel, and zona incerta.38 Hence, it is possible that it may be responsible for subcortical spread and contralateral transmission of seizures. Embryologically, the commissures develop shortly before the development of corpus callosum. They also act as a seed to the development of CC. Thus, it is possible to have corpus callosal agenesis with intact commissures, but not commissural agenesis w ­ ithout intact corpus callosum.26,​30,​38,​39 In an interesting personal communication, Sharan et al40 reported a case to the author (P.S.C.) about a person with epilepsy with corpus callosal agenesis. Here, an SEEG placement confirmed the spread of seizures through the AC. In our practice, we currently perform complete CC with anterior and HCT in all cases. We also add posterior commissurotomy only in children with LGS with severe progressive encephalopathy. 7. Complications associated with corpus callosal sectioning along with commissurotomies: The main complication of complete CC and commissurotomies is the morbidity associated with acute disconnection.26,​29,​30,​31,​41,​42 We have often observed axial (e.g., pooling of saliva in oral cavity) and appendicular (paucity of movement of nondominant hand to command) apraxias. In addition, we often found the children taking a longer time to reverse from anesthesia and because of this we prefer to electively ventilate these patients. Similar complications and seizure outcome profiles were also reported in contemporary series.22,​29,​ 43 Most of our patients had moderate to severe mental retardation with severe epilepsy. Following CC with pan commissurotomy, there was no deterioration (rather mild improvement) after surgery. We believe that this could have been because of the fact that most of the children were already severely compromised in their cognitive status. Thus, we perform CC + commissurotomy even though the children were initially associated with acute disconnection. 8. Postoperative drain: We usually do not prefer to place an intraventricular drain as the bleeding is minimal. However, if the ventricle is opened and there has been a significant passage of blood into the ventricle, then a drain may be placed for a period of 24 to 36 hours or earlier if the blood clears off.

„„ Surgical Technique Presurgical Workup This is similar to what has been described in Chapter 61. It is important to rule out a lesion while considering CC. Often, when an epilepsy protocol imaging is performed (thin slices, T1, T2,

63  Endoscope-Assisted Corpus Callosotomy with Anterior, Hippocampal, and Posterior Commissurotomy and FLAIR), one can often pick up subtle lesions ­especially when the case is discussed in a comprehensive epilepsy surgery conference. Not uncommonly, a more advanced investigation like SPECT, PET, and MEG can often pick up a suspicious area, which can then be confirmed as a lesion on detailed analysis on MRI. It is also important to analyze the MRI on a high-­resolution ­monitor and not on films, as only limited images would be printed on physical films. The presence of drop attacks in the background of LGS should provide impetus for the team to consider for CC. In our center, in children with LGS with significant epileptic encephalopathy, we prefer to combine a complete CC with anterior, ­hippocampal, and posterior commissurotomy. The idea is to provide the maximal degree of interhemispheric disconnection so as to offer the best chance to prevent interhemispheric spread of seizures.

Procedure The patient is placed supine in a neutral position with head fixed in head clamp. A 5 to 6 cm transverse skin incision is placed a centimeter in front of coronal suture, and the

c­ raniotomy was 4 × 3 cm longitudinal. The final site of the craniotomy was determined using neuronavigation (to avoid veins). Stepwise procedure is shown in Fig. 63.1; Video 63.1. We prefer to use the same type of endoscopic system and instruments as described in endoscopic-assisted hemispherotomy. Following dural opening, interhemispheric fissure is accessed and cisternal cerebrospinal fluid is released to make the brain lax. We prefer to start the corpus callosal sectioning at the posterior part and then proceed to the splenial part, proceeding next to the anterior part and genu. Unlike hemispherotomy, the division in CC should be performed strictly in the midline going into the cavum rather than the ventricle. Once the corpus callosum was sectioned Video 63.1 The technique of endoscopic complete corpus callosotomy with anterior, hippocampal and posterior commissurotomy. (This video is provided courtesy of P. Sarat Chandra.) https://www.thieme.de/de/q.htm?p=opn/tp/255910102/ 9781626238176_c063_v001&t=video

Fig. 63.1  The stepwise execution of the procedure from skin incision (a), craniotomy (b), and bone flap size (c), followed by sectioning of corpus callosum (d). After completing the complete section of corpus callosum, anterior (e) and posterior (g) commissures were identified (arrows in e and g) and sectioned respectively (f, h). Postoperative MRI shows sectioned corpus callosum (i) and posterior commissure.

591

592

IVf  Other Disconnective Procedures ­completely, the septae on either side of the cavum were separated. First, the AC was divided. The hippocampal commissure was divided by separating the septae and following it posteriorly. The hippocampal commissure is located at the level of the posterior part of the body of CC or just in front of the splenium. The PC is last to be divided. It is located over the dorsal and superior part of the third ventricle. This may be divided using a microscissors. Once divided, the aqueduct becomes clearly visible. All patients were put on ventilator for at least 24 hours after surgery. This was to avoid any immediate postextubation morbidity as all the patients were quite sick, mentally retarded with several seizures a day. All underwent immediate postoperative MRI and a postoperative CT scan within 6 hours after surgery. Fig. 63.2 shows a patient of LGS with bilateral pachygyria; Fig. 63.3 shows division of anterior commissure, anterior part and splenial part of corpus callosum, and PC; and Fig. 63.4 shows postoperative axial section of MRI with division of genu of the corpus callosum.

Results All patients were assessed for social quotient (SQ) using Vineland Social Maturity Scale (VSMS). For seizure outcome characterization and parental satisfaction, the scale adopted by Iwasaki et al was used.44 Child behavior check list (CBCL) score (ASEBA, Inc. USA) was adopted for behavioral assessments (113 items for parents). CBCL scores greater than 60 indicated borderline or at-risk kids while scores greater than 64 indicate significant clinical behavioral problems. All patients were assessed at 3, 6, 9, and 12 months initially and every 6 months. Routine EEGs were performed at 3 months and later on as per clinical decision of the neurologist 1 year later. Detailed­ neuropsychological assessments were conducted at 1 year. Seizure outcomes were recorded at last outpatient follow-up. Postoperative MRI was performed once immediately at the time of surgery and then scheduled on follow-up between 3 and 7 months.

Fig. 63.2  (a, b) MRI coronal images at the level of anterior and posterior part of the corpus callosum in a 2-year-old child with LennoxGastaut syndrome Lennox-Gastaut syndrome (LGS) with bilateral pachygyria.

Fig. 63.3  The same patient underwent endoscopic complete corpus callosotomy with anterior, hippocampal, and posterior commissurotomy. Division of anterior commissure and anterior part of corpus callosum (a) and also splenial portion of the corpus callosum and posterior commissure (arrow) (b) are shown.

63  Endoscope-Assisted Corpus Callosotomy with Anterior, Hippocampal, and Posterior Commissurotomy

Neuropsychological Outcomes Aggression in behavior of patients was noted in eight (first 3 months), and this reduced in 6 to 9 months in three patients. The mean IQ preoperatively was 26.3 ± 11.1, which did not deteriorate following surgery (mean score after surgery: 27 ± 8.4 in 6 months and 28.8 ± 9.5 in 1 year). Behavioral parameters, in particular social contacts, attention span, and learning, also did not show deterioration. Detailed SQ analysis (at 6 and 12 months) did not show any deterioration. CBCL scores also revealed no significant deterioration (rather improved mildly after surgery). The mean preoperative CBCL was 69.25 ± 2.5 as compared to the postoperative score of 61.81 ± 3.8 at 1-year follow-up. On parental questionnaire, 28 parents were satisfied with the surgical results and agreed to recommend this surgery to others.

Control Cohort

Fig. 63.4  Axial section MRI of the same patient shown in Fig. 63.2 following surgery showing division of the genu of the corpus callosum (arrow).

A total of 18 patients (mean age: 11 ± 6.9 [range: 2–25], 12 males) underwent CC, ACT, and HCT in our center. Sixteen patients underwent posterior commissurotomy as well along with CC, AC, and HC. The latter group included children with LGS with severe to profound epileptic encephalopathy with multiple seizure types (tonic seizures [n = 12], tonic–clonic [n = 11], absence [n = 4], myoclonic seizures [n = 4], and focal dyscognitive seizures [n = 3]) and high seizure frequency (mean seizure frequency was 24.5 ± 19.8/day [range: 1–60]). The mean duration of epilepsy was 11 ± 5.4 years.

Seizure Outcomes Mean follow-up was 19 ± 5.7 months (range: 15–32 months). There was a complete improvement in drop attacks in all patients. Significant decrease (>90%) in seizure frequency was noted in 27 patients, moderate reduction (>50%) in 6 patients, and increased seizure frequency seen in 1 patient. Decrease in frequency was observed in all types of seizures in these patients (tonic, tonic–clonic, absence, and myoclonic seizures). One patient initially underwent a complete callosotomy only. ­Following this, the patient went into status on the third postoperative day. Bedside EEG showed continuous nonconvulsive status epilepticus with ictal discharges from bilateral centroparietal areas. She was taken up for surgery in emergency; an additional ACT, HCT, and PCT were performed. Following this, the seizures stopped. One patient encountered increased ­duration of seizure with change of seizure type from spasms to unilateral tonic seizures of long duration.

The study group was compared with a similar cohort (n = 16, mean age of 12 ± 4.7 years, range of 4–21 years), where a complete CC only was performed. The mean age of onset of seizures was 26.7 ± 29.1 months, and mean seizure frequency was 10.3 ± 4.8/ day (much less than study group). All had drop attacks with multiple other seizure types. Following a CC, drop attacks were relieved (>90%) in 10 of 16 (62%), and other seizure types were relieved (>90%) in 7 of 16 (43%) at a mean follow-up of 16.4 months (13.2– 21 months). On applying the “Fisher’s exact test,” there was a significant difference in study and control group for drop attacks (p = 0.003), being better in the case group. However, there was no significant difference for other seizure types (p = 0.240).

Complications Six patients had evidence of acute disconnection, characterized by confusion and limb apraxia of the nondominant side, buccal apraxia (with pooling of saliva). This improved during hospital stay to their preoperative functional levels. Two patients developed hyperammonemic encephalopathy, which required discontinuation of valproate and administration of lactulose. One patient developed bacterial meningitis, which was treated appropriately. Mean hospital stay was 9.5 ± 5.1 days (5–20 days). No mortality or long-term procedural morbidity was recorded. One patient of MRI-negative epilepsy who earlier underwent respective surgery following SEEG placement came back with multiple seizures. He underwent CC + AC + HC + PC. However, there was no improvement, and he continued to have seizures. Due to poor neurological status, he had to be tracheotomized. While it was not possible to examine the detailed extraocular movements due to severe deranged cognitive status of the patients, none of the patients had contralateral light reflex. This did not affect them or in their care giving in any manner.

„„ Discussion Complete callosal sectioning (first introduced by van Wagenen and Herren in 194027) was described as an effective procedure for breaking secondary bilateral synchrony and alleviating drop

593

594

IVf  Other Disconnective Procedures attacks with more than 90% improvement in drops with reasonable long-term remission.1,​2,​11,​12,​16,​17 Various authors tried different combinations: complete CC + HCT,28 complete CC,29 complete CC + ACT,30 and complete CC + HCT.31 Most of these authors reported 10 to 20% of primary nonresponders and close to 30% of patients further relapses in next few years with outcome mostly remaining stable thereafter.17,​18 The common reasons cited for callosal sections failing to alleviate drop attacks or their recurrence is the possibility of transmission of epileptiform activity through other interhemispheric pathways like anterior, posterior, and hippocampal commissures.18,​32,​33,​34 As mentioned earlier, using DTI, Jang and Kwon in 2013 and 2014 demonstrated a much wider connections of the fornix to include the cerebral cortex (precentral gyrus, postcentral gyrus, and posterior parietal cortex) and also the brainstem through the thalamus.35,​36,​37 AC connects temporal lobes32 while hippocampal commissure (PC), mostly considered as rudimentary in humans, connects both the hippocampi and joins both the bodies of the fornices just under the posterior part of the body of the corpus callosum.6,​33 Thus, division of anterior and hippocampal commissure may be expected to have significant effect to reduce the seizure burden. The PC is known to carry pupilloconstrictor fibers to mediate bilateral light reflex. However, a detailed anatomical study using anterograde fiber degeneration and retrograde axon transport (using horse peroxidase) revealed a much larger distribution of PC to pars reticularis, periaqueductal grey matter connecting both cranial and caudal reticular system, ventral lateral geniculate nucleus, H field of Forel, and zona incerta.38 Hence, it is possible that it may be responsible for subcortical spread and contralateral transmission of seizures. The main complication of the earlier series combining a complete CC with either ACT or HCT commissurotomies was the morbidity associated with acute disconnection.26,​29,​30,​31,​41,​42 Similar complications and seizure outcome profiles were also reported in contemporary series.22,​29,​43 Although we observed acute disconnection syndromes, it did not alter functional status of our patients. Even though all our patients had moderate to severe mental retardation with severe epilepsy, there was no deterioration (rather mild improvement) after surgery. One of the reasons could be because patients were already severely compromised in their cognitive status that an additional disconnection may not alter their quality of life or leave a permanent disability. On the contrary, relief of disabling seizures was perceived as the biggest factor of improvement by parents in postoperative period. Recently, surgeries utilizing smaller craniotomies are being performed more frequently for epilepsy surgeries like p ­ eri-­insular hemispherotomies and even for anterior corpus callosotomies.6,​11 More recently, an endoscopic-assisted hemispherotomy through a transcallosal route has been described.45,46 We were the first to introduce the concept of endoscopic CC along with anterior, hippocampal, and posterior commissurotomy.24,​25 Our study is the first of its kind to demonstrate utility and

safety of this approach for CC, AC, HC, and PC division. We also believe that endoscope-assisted approach through a mini craniotomy helps in minimizing the unnecessary brain exposure, and reduces the blood loss. It does become important to localize the venous anatomy preoperatively (MRI with gadolinium along with susceptibility-weighted imaging) to plan the optimal location for performing the craniotomy. Overall, we feel that with some experience, endoscope-assisted approach through a “mini craniotomy” is feasible and is of advantage in performing this procedure. Neurological complications are reported in all major series in the magnitude of 2 to 5% with permanent sequel in 5% of patients.2,​11,​13 We did not encounter any motor deficits or permanent deficits. However, postoperative akinetic state, buccal apraxia manifested as drooling of saliva, and limb apraxia were common. We feel that all these changes are due to acute disconnection syndromes, and resolve over the subsequent weeks. In addition, improvement in seizure outcome and cognitive status compensated adequately this short-term morbidity. However, despite the limitations mentioned later, the study did succeed in the objective of establishing the safety, efficacy, and acceptable morbidity of CC combined with anterior, hippocampal, and posterior commissurotomies. It also demonstrated a similar efficacy compared to complete CC alone, if not better. Long-term follow-up of larger cohorts and comparison studies are needed for better understanding. The authors also for the first time demonstrated a key hole endoscopic-assisted CC with anterior, hippocampal, and posterior commissurotomy.

„„ Conclusion Complete CC combined with anterior, hippocampal, and posterior commissurotomy performed in patients with severe drop attacks and nonlocalizing epilepsy has been demonstrated to be safe and efficacious in LGS. Drop attacks ceased completely in all patients and there was a significant improvement in all other seizure types (>90% reduction in 66% of cases). This was also accompanied with a significant improvement in cognition. We would also be hesitant at this stage to subject this procedure in patients with well-preserved cognitive status. Larger future studies especially those involving blinding this procedure with CC only may be helpful in further establishing its role.

„„ Acknowledgment This study is part of center of excellence for epilepsy and is funded by the Department of Science and ­ Technology (­ Ministry of Science and Technology). The authors also wish to acknowledge Dr. Shri Vidya Malviya for providing ­statistical inputs.

63  Endoscope-Assisted Corpus Callosotomy with Anterior, Hippocampal, and Posterior Commissurotomy

References 1. Tanriverdi T, Olivier A, Poulin N, Andermann F, Dubeau F. Longterm seizure outcome after corpus callosotomy: a retrospective analysis of 95 patients. J Neurosurg 2009;110(2):332–342 2. Jalilian L, Limbrick DD, Steger-May K, Johnston J, Powers AK, Smyth MD. Complete versus anterior two-thirds corpus ­callosotomy in children: analysis of outcome. J Neurosurg Pediatr 2010;6(3):257–266 3. Oguni H, Olivier A, Andermann F, Comair J. Anterior callosotomy in the treatment of medically intractable epilepsies: a study of 43 patients with a mean follow-up of 39 months. Ann Neurol 1991;30(3):357–364 4. Reutens DC, Bye AM, Hopkins IJ, et al. Corpus callosotomy for intractable epilepsy: seizure outcome and prognostic factors. Epilepsia 1993;34(5):904–909

20. Hausmann M, Corballis MC, Fabri M, Paggi A, Lewald J. Sound lateralization in subjects with callosotomy, callosal agenesis, or hemispherectomy. Brain Res Cogn Brain Res 2005; 25(2):537–546 21. Lin JS, Lew SM, Marcuccilli CJ, et al. Corpus callosotomy in multistage epilepsy surgery in the pediatric population. J Neurosurg Pediatr 2011;7(2):189–200 22. Wilson DH, Reeves A, Gazzaniga M, Culver C. Cerebral commissurotomy for control of intractable seizures. Neurology 1977;27(8):708–715 23. Papo I, Quattrini A. The role of corpus callosum in experimental epileptogenesis. A brief survey of literature. J Neurosurg Sci 1997;41(1):27–30

5. Asadi-Pooya AA, Malekmohamadi Z, Kamgarpour A, et al. Corpus callosotomy is a valuable therapeutic option for patients with Lennox-Gastaut syndrome and medically refractory seizures. Epilepsy Behav 2013;29(2):285–288

24. Chandra PS, Kurwale NS, Chibber SS, et al. Endoscopic-assisted (through a mini craniotomy) corpus callosotomy combined with anterior, hippocampal, and posterior commissurotomy in ­Lennox-Gastaut syndrome: a pilot study to establish its safety and efficacy. Neurosurgery 2016;78(5):743–751

6. Asadi-Pooya AA, Sharan A, Nei M, Sperling MR. Corpus callosotomy. Epilepsy Behav 2008;13(2):271–278

25. Chandra PS, Tripathi M. Endoscopic epilepsy surgery: emergence of a new procedure. Neurol India 2015;63(4):571–582

7. Benedetti-Isaac JC, Torres-Zambrano M, Fandiño-Franky J, et al. Vagus nerve stimulation therapy in patients with drug-resistant epilepsy and previous corpus callosotomy [in Spanish] Neurocirugia (Astur) 2012;23(6):244–249

26. Harbaugh RE, Wilson DH, Reeves AG, Gazzaniga MS. Forebrain commissurotomy for epilepsy. Review of 20 consecutive cases. Acta Neurochir (Wien) 1983;68(3–4):263–275

8. Benifla M, Rutka JT, Logan W, Donner EJ. Vagal nerve stimulation for refractory epilepsy in children: indications and experience at The Hospital for Sick Children. Childs Nerv Syst 2006;22(8):1018–1026 9. Turanli G, Yalnizoğlu D, Genç-Açikgöz D, Akalan N, Topçu M. Outcome and long term follow-up after corpus c­allosotomy in childhood onset intractable epilepsy. Childs Nerv Syst 2006;22(10):1322–1327 10. Paglioli E, Martins WA, Azambuja N, et al. Selective posterior callosotomy for drop attacks: a new approach sparing prefrontal connectivity. Neurology 2016;87(19):1968–1974 11. Kasasbeh AS, Smyth MD, Steger-May K, Jalilian L, Bertrand M, Limbrick DD. Outcomes after anterior or complete corpus ­callosotomy in children. Neurosurgery 2014;74(1):17–28, ­discussion 28 12. Shim KW, Lee YM, Kim HD, Lee JS, Choi JU, Kim DS. Changing the paradigm of 1-stage total callosotomy for the treatment of pediatric generalized epilepsy. J Neurosurg Pediatr 2008;2(1):29–36 13. Spencer SS, Spencer DD, Sass K, Westerveld M, Katz A, Mattson R. Anterior, total, and two-stage corpus callosum section: differential and incremental seizure responses. Epilepsia 1993;34(3):561–567 14. Bower RS, Wirrell E, Nwojo M, Wetjen NM, Marsh WR, Meyer FB. Seizure outcomes after corpus callosotomy for drop attacks. Neurosurgery 2013;73(6):993–1000 15. Cukiert A, Cukiert CM, Burattini JA, et al. Long-term outcome after callosotomy or vagus nerve stimulation in consecutive prospective cohorts of children with Lennox-Gastaut or Lennox-like syndrome and non-specific MRI findings. Seizure 2013;22(5):396–400 16. Maehara T, Shimizu H. Surgical outcome of corpus callosotomy in patients with drop attacks. Epilepsia 2001;42(1):67–71 17. Sunaga S, Shimizu H, Sugano H. Long-term follow-up of seizure outcomes after corpus callosotomy. Seizure 2009;18(2):124–128 18. Wong TT, Kwan SY, Chang KP, et al. Corpus callosotomy in children. Childs Nerv Syst 2006;22(8):999–1011 19. Chen PC, Baumgartner J, Seo JH, Korostenskaja M, Lee KH. Bilateral intracranial EEG with corpus callosotomy may uncover seizure focus in nonlocalizing focal epilepsy. Seizure 2015;24:63–69

27. Van Wagenen WP, Herren RY. Surgical division of commissural pathways in the corpus callosum: relation to spread of an epileptic attack. Arch Neurol Psychiatry 1940;44(4):740–759 28. Marcus EM, Watson CW. Symmetrical epileptogenic foci in monkey cerebral cortex. Mechanisms of interaction and regional variations in capacity for synchronous discharges. Arch Neurol 1968;19(1):99–116 29. Wilson DH, Reeves A, Gazzaniga M. Division of the corpus callosum for uncontrollable epilepsy. Neurology 1978;28(7):649–653 30. Musgrave J, Gloor P. The role of the corpus callosum in bilateral interhemispheric synchrony of spike and wave discharge in feline generalized penicillin epilepsy. Epilepsia 1980;21(4):369–378 31. Gates JR, Leppik IE, Yap J, Gumnit RJ. Corpus callosotomy: clinical and electroencephalographic effects. Epilepsia 1984;25(3):308–316 32. Adam C. [How do the temporal lobes communicate in medial temporal lobe seizures?] Rev Neurol (Paris) 2006;162(8–9):813–818 33. Gloor P, Salanova V, Olivier A, Quesney LF. The human dorsal hippocampal commissure. An anatomically identifiable and functional pathway. Brain 1993;116(Pt 5):1249–1273 34. Wada JA. Transhemispheric horizontal channels for transmission of epileptic information. Jpn J Psychiatry Neurol 1991;45(2):235–242 35. Jang SH, Kwon HG. Perspectives on the neural connectivity of the fornix in the human brain. Neural Regen Res 2014;9(15):1434–1436 36. Jang SH, Kwon HG. Neural connectivity of the posterior body of the fornix in the human brain: diffusion tensor imaging study. Neurosci Lett 2013;549:116–119 37. Yeo SS, Seo JP, Kwon YH, Jang SH. Precommissural fornix in the human brain: a diffusion tensor tractography study. Yonsei Med J 2013;54(2):315–320 38. Shoumura K, Imai H, Kimura S, Suzuki T, Ara M. Posterior commissural connections of area pretectalis and neighboring structures in cat, with special reference to pupilloconstrictory pathway via posterior commissure. Jpn J Ophthalmol 1987;31(2):289–304 39. Gonçalves Ferreira AJ, Farias JP, Carvalho MH, Melancia J, Miguéns J. Corpus callosotomy: some aspects of its microsurgical anatomy. Stereotact Funct Neurosurg 1995;65(1–4):90–96 40. Sharan A. SEEG Confirmation of transmission of seizures through anterior commissure in a patient with agenesis of corpus­

595

596

IVf  Other Disconnective Procedures callosum (personal communication). Thomas Jefferson University, PA; 2016 41. Harbaugh RE, Wilson DH. Telencephalic theory of generalized epilepsy: observations in split-brain patients. Neurosurgery 1982;10(6, Pt 1):725–732 42. Amacher AL. Midline commissurotomy for the treatment of some cases of intractable epilepsy. Preliminary report. Childs Brain 1976;2(1):54–58 43. Wilson DH, Reeves AG, Gazzaniga MS. “Central” commissurotomy for intractable generalized epilepsy: series two. Neurology 1982;32(7):687–697 44. Iwasaki M, Uematsu M, Nakayama T, et al. Parental satisfaction and seizure outcome after corpus callosotomy in patients

with infantile or early childhood onset epilepsy. Seizure 2013;22(4):303–305 45. Chandra PS, Kurwale N, Garg A, Dwivedi R, Malviya SV, Tripathi M. Endoscopy-assisted interhemispheric transcallosal hemispherotomy: preliminary description of a novel technique. Neurosurgery 2015;76(4):485–494, discussion 494–495 46. Chandra PS, Subianto H, Bajaj J, Girishan S, Doddamani R, Ramanujam B, Chouhan MS, Garg A, Tripathi M, Bal CS, Sarkar C, Dwivedi R, Sapra S, Tripathi M. Endoscope-assisted (with robotic guidance and using a hybrid technique) interhemispheric transcallosal hemispherotomy: a comparative study with open hemispherotomy to evaluate efficacy, complications, and outcome. J Neurosurg Pediatr. 2018 Nov 1:1–11

64

  Endoscopic Disconnection of Hypothalamic Hamartomas Georg Dorfmüller and Sarah Ferrand-Sorbets

Summary Surgical excision of hypothalamic hamartoma (HH) bears a higher risk of neurological, vascular, or endocrine complications. Since HHs are nongrowing masses of dystopic and disorganized but mature neuroglial tissue with a high epi­ leptogenicity, a surgical disconnection should be sufficient to interrupt the seizure propagation from its onset within the hamartoma. The advantage of the transventricular endoscopic disconnection of HH over other minimally invasive techniques is the continuous visual control during the entire procedure. In general, the endoscopic disconnection is easily tolerated even in young children, necessitates a short stay in the hospital, and will give immediate results following surgery if a sufficient extent of the HH attachment has been interrupted. Image guidance is indispensable for an optimal approach; as the size of the ventricles is frequently not enlarged, navigation guidance is necessary to enter into the optimal ventricular trajectory. The side of the endoscopic entry to the lateral ventricle is chosen opposite to the side of the main hypothalamic attachment of the hamartoma, in order to have the best access to the plane of HH attachment. Patients with HH with exclusive or predominant ­attachment to the hypothalamus within the third ventricle (type 2 of Delalande’s surgical HH classification) are ­excellent candidates for the transventricular endoscopic approach for disconnection. Overall seizure-free outcome in our series of 126 children and adults with HH is 70%. Keywords:  hypothalamic hamartoma, endoscopic disconnection, gelastic seizures, Delalande’s classification

„„ Introduction Hypothalamic hamartoma (HH) is a congenital tumor-like mass of dystopic and disorganized cerebral tissue composed of mature neuronal and glial cells. Beside a certain volume increase which parallels normal brain growth, there is no

further progression. These lesions are attached to the hypothalamus on one or both sides, in most cases at the floor of the third ventricle, to the tuber cinereum or to the mammillary bodies and may be responsible for endocrine dysfunctions, notably precocious puberty, behavioral disturbances and epilepsy, the latter being very frequently resistant to drug treatment. The seizures can begin from the first day of life. Seizures are typically characterized by gelastic (laughing) fits in most cases, but other seizure types can be present as well. Smaller HH could remain undiagnosed in the pre-MRI era. First clear etiological link between HH and epilepsy has been proven over 20 years ago almost simultaneously by depth electrode recording of seizures originating from the hamartoma itself1 and the confirming results of ictal single-photon emission computed tomography studies.2 Surgery for these nongrowing lesions was initially performed in children with precocious puberty.3 Larger HH extending below the floor of the third ventricle would be approached by one of the classical skull base approaches. However, accessibility of HH located mainly within the third ventricle was problematic and associated with a higher risk of surgery-related morbidity, like the tumors of the third ventricle. Two new approaches were proposed in the beginning of this century: microscopic transcallosal and endoscopic transventricular approaches. Jeff Rosenfeld presented the transcallosal anterior interforniceal approach in order to access to the third ventricle for microsurgical resection of the HH.4 On the other hand, Olivier Delalande proposed a transventricular endoscopic approach in order to disconnect the HH from the wall of the third ventricle initially using an electrocoagulation probe (Fig. 64.1),5,​6,​7 later using thulium laser.8 For this purpose, Delalande classified all sessile HH into four subtypes, according to their plane of attachment and size (see later).5 In this chapter, we will describe the technique of endoscopic transventricular disconnection of HH, its indications, as well as the surgical results in our single-center cohort of 130 mostly pediatric patients.

598

IVf  Other Disconnective Procedures

„„ Indication of Endoscopic Transventricular Disconnection In general, the sessile or intrahypothalamic type HH (with respect to the form of hypothalamic attachment) is associated with epilepsy, whereas those of the pedunculated or parahypothalamic type have been related primarily to precocious puberty.9 In 2003, Delalande presented a classification of the sessile hamartoma type, the one pediatric epilepsy surgery centers are confronted with, into four subtypes according to the orientation of its attachment to the hypothalamus and its size (Fig. 64.2).5 Assigning a HH to one of the four subtypes is easier when the MRI is studied in the coronal plane (Fig. 64.3). This classification was proposed in order to facilitate the evaluation of the surgical approach to be chosen, and in particular to estimate the efficacy of a transventricular endoscopic approach.

Fig. 64.1  Endoscopic transventricular approach to the hypothalamic hamartoma which is attached to the ventricular wall and the hypothalamus on the opposite side. From the drawing, it is clear that the access to an ipsilaterally attached hamartoma is not recommended through this trajectory. In cases with bilateral attachment, the entry point should be chosen closer to the midline. (Reproduced with permission from Dorfmüller et al.7)

„„ Description of the Endoscopic Transventricular Disconnection Technique This procedure can be performed with a neuronavigation system or with stereotactic robot guidance. We use the StealthStation (Medtronic) or the Rosa robot (Zimmer B ­ iomet). In our opinion, image guidance is indispensable for an optimal approach: since the size of the ventricles is frequently not enlarged, navigation guidance is necessary to enter into the

Fig. 64.2  Classification schemes of the four subtypes of hypothalamic hamartoma according to Delalande’s surgical classification. (Reproduced with permission from Dorfmüller et al.7)

64  Endoscopic Disconnection of Hypothalamic Hamartomas

Fig. 64.3  MRI of the four subtypes of hypothalamic hamartoma (HH) according to the Delalande’s classification. (a) Coronal T2-weighted MRI of a type 1 hamartoma, with bilateral horizontal attachment to the inferior surface of the hypothalamus. The hamartoma mass is entirely located below the floor of the third ventricle. In these lesions, a skull base approach is indicated for resection or disconnection. Type 2 hamartoma: T1-weighted MRIs with horizontal slice (b) and coronal slice (c). The lesion is entirely located within the third ventricle and is attached to the left hypothalamus. Due to its vertical plane of attachment, an endoscopic disconnection through a contralateral transventricular approach will have the best chance, as compared to all other HH types, to achieve a complete or near-complete disconnection, with expectably good seizure outcome. Type 3 hamartoma: T1-weighted (d) and fluid-attenuated inversion recovery (FLAIR) coronal (e) MRIs. These lesions are partly within the third ventricle and partly extending below the floor. The lines of attachment to the hypothalamus, one-sided or bilateral, are in several directions, with more vertical and more horizontal components. An endoscopic approach for disconnection is possible, but the disconnection can be achieved only partly. The surgical results on seizure outcome are therefore less predictable and additional resective procedures through a transcallosal interforniceal or a skull base approach might be necessary to achieve seizure freedom. Type 4 hamartoma on a coronal T2-weighted (f) and FLAIR image (g). These mass lesions are always attached to both sides of the hypothalamus, with dilatation of the third ventricle which is almost filled out, frequent dilatation of one or both foramen of Monro. Like type 3 hamartomas, they extend below the floor of the third ventricle into the interpeduncular cistern. Endoscopic disconnection will be only incomplete and orientation is difficult because of loss of the anatomical landmarks within the third ventricle and lack of space in the third ventricle to move the endoscope when the HH mass is close to the foramen of Monro. However, we have been able to obtain in a few patients with this giant hamartoma type a significant seizure reduction, sometimes more than 90%, through repeated endoscopic disconnection surgery alone.

optimal ventricular trajectory. Furthermore, as we emphasized earlier,6 the exact placement of the burr hole should be in line with the foramen of Monro and the center of the hamartoma as the target of this endoscopic trajectory (Fig. 64.4). This will enable to the best possible degree endoscope movements during the disconnection in order to expose the most anterior and posterior parts of the HH without forcing the structures that form the foramen of Monro, notably the column of the fornix anteriorly and the prominent veins posteriorly. Damage to the former may elicit transient or permanent memory disorders; damage to the latter would provoke bleeding during the procedure with reduced visibility and even the need to end this surgery precociously. The patient is in a supine position and the head is straight, fixed in the Mayfield head holder. Image guidance either with a neuronavigation system or a stereotactic robot-assisted system (we use alternately the StealthStation or the Rosa robot in the frameless mode; Fig. 64.5) will now be installed. The side of the endoscopic entry to the lateral ventricle is chosen opposite to the side of the main hypothalamic attach-

ment of the hamartoma, in order to have the best access to the plane of HH attachment (Fig. 64.1). After having planned the trajectory, a paramedian frontal burr hole is placed 2.5 to 3.0 cm lateral to the midline and usually in front of the coronal suture, according to defined trajectory. After opening of the dura and punctiform cortical coagulation, the rigid neuroendoscope (Decq set, Karl Storz) is introduced and advanced under image guidance until the frontal horn of the lateral ventricle is entered. After having introduced the working insert with a 30-degree telescope attached to the camera (Karl Storz), the anatomy of the anterior lateral ventricle is appreciated, the foramen of Monro being in front of us. Sometimes, a small slit-like shape of the foramen will prompt us to carefully infuse the ventricle with a saline solution that has been connected to the endoscope before its introduction. This allows a moderate ventricular dilatation with widening of the foramen of Monro and easier passage into the third ventricle. By passing the foramen of Monro and entering into the third ­ventricle, the shape and extension of the hamartoma will be immediately appreciated (Fig. 64.6). In most cases, the principal

599

600

IVf  Other Disconnective Procedures

Fig. 64.4  The endoscopic trajectory (arrow) should be in line with the foramen of Monro and the center of the hypothalamic hamartoma (dotted line) in order to limit the stressing forces at the level of the foramen of Monro during the endoscope inclination while disconnecting the most anterior and posterior parts of the hamartoma. In very large lesions, it will even be necessary to plan the disconnection in two steps through two distinct burr holes, allowing to disconnect the most anterior and posterior parts, respectively.

Fig. 64.6  View from the right foramen of Monro into the third ventricle. Its anterior wall is formed by the anterior column of the fornix (FX) which is coming from the midline to bend lateroinferiorly. Anterior to the fornix, the massa intermedia (MI, or interthalamic adhesion) is usually prominent. The hypothalamic hamartoma (HH) is usually immediately identified at the floor. In this case, there is a one-sided attachment (AT) to the left ventricular wall (VW), the right side of the hamartoma being free in the ventricle. Anteriorly, one can notice in the depth the infundibular recess (IR).

Fig. 64.5  The neuroendoscope has been introduced and is being image guided during the procedure with robotic assistance (ROSA robot in the frameless stereotactic mode). The shaft of the endoscope is attached to the tool adapter of the ROSA. The slow endoscope movements during the procedure are induced by the surgeon by pushing the endoscope gently, whereas the robot will stabilize all movements isocentrically at the level of the burr hole. The level of the isocenter and the margins of security of endoscope movements within the ventricle have been defined during the trajectory planning before surgery.

attachment to the hypothalamus is one-sided, as should be studied before surgery on MRI. But even then, there is frequently a bilateral attachment of the most posterior part of the HH. For the disconnection at the line of attachment, we use a fine electrocoagulation probe at very low energy level (coagulation electrode 1 mm diameter, Karl Storz; Lamidey Medical Surgilec MC4 monopolar electric coagulator). Before beginning with the vertical coagulation once the hamartoma’s surface is entered, the depth of the lesion along this line should be known, in order to avoid entering into the space below the hamartoma which might provoke heating injury to the third cranial nerve, the optic tract, or vascular structures. Also, a more posterior and deep coagulation trajectory could reach the cerebral peduncle and brain stem. A series of coagulations is performed along a line which represents the interface between hamartoma and third ventricular wall (Fig. 64.7). This procedure of disconnection can also be realized with an endoscopically applied thulium laser probe in the pulsating mode (Fig. 64.8) (Revolix Junior 15-W thulium:YAG surgical laser unit, LISA Laser products OHG), which we have reported earlier in more detail.8 The two methods can also be combined during one procedure. Usually, we begin with a series of electrocoagulations in order to trace the disconnection line, and thereafter use the thulium laser to perform longitudinal cuts that connect the coagulation holes and create a more continuous and profound cleavage. With a small Fogarty catheter (smallest size, green), we use to enter into the cleavage and inflate the balloon with about 0.8 mL saline for a few seconds. This will spread away the hamartoma mass from the ventricular wall, thus interrupting more fibers and thereby enhancing the degree of disconnection. Usually, the ballooning is repeated two or three times in the anterior, middle, and posterior part of the hamartoma.

64  Endoscopic Disconnection of Hypothalamic Hamartomas

Fig. 64.7  (a, b) The thermocoagulation probe is introduced and deep lesions at the level of the hamartoma attachment to the ventricular wall, corresponding to less than the entire depth of the attachment plane, are realized.

Fig. 64.8  (a, b) Disconnection with the thulium laser (Revolix) in the pulsatile mode. Due to the rigidity of the fiber probe, horizontal movements along the hamartoma attachment can be performed during the disconnecting vaporization thus producing multiple adjacent cuts.

During the procedure, we have observed in about a third of our patients an increase in the heart rate, which will persist until the end of the procedure, but was in no case prolonged over more than a few hours thereafter. This phenomenon is most likely due to manipulation of the ventricular wall and/or volume changes during irrigation of physiological saline during the procedure, although we are very careful to avoid transient intracranial hypertension through the irrigating system while the working channels of the endoscope would be closed. This surgical procedure lasts about 30 to 60 minutes. After having verified the absence of any intraventricular bleeding source, the endoscope is carefully withdrawn and the wound closed, after having filled the burr hole with autogenous bone powder. The effect of HH disconnection on seizure activity can be seen in the first postoperative days, sometimes with spectacular accompanying amelioration of behavioral disturbances. On the other side, seizures during the first postoperative week may still be present and it does not predict an unfavorable outcome. Antiepileptic medications should be continued and we see the patients and their family at 3 months postoperatively, in order to appreciate the first interval input for expected seizure outcome. We do perform a postoperative MRI at this point, although even high-resolution T1- or T2-weighted sequences with a 3T machine cannot trace the area of disconnection reliably. A second endoscopic disconnective procedure, either from the same side or from the opposite side in bilaterally attached HH, can be an option. However, we usually wait for

a period of about 9 to 12 months before deciding for second surgery as seizure activity can vary considerably in the first postoperative months. Our policy of drug tapering is more conservative, that is, we start later and over a longer period. In this kind of epilepsy surgery, as compared to resective epilepsy surgery, HH disconnection will rarely be 100% in anatomical terms. Therefore, the threshold for recurring seizures postoperatively is lower. When dealing with larger and more complex HH of the type 3 or 4, where only part of the mass will have attachment within the third ventricle and a larger attachment is horizontal at the inferior surface, we propose endoscopic disconnection of the accessible upper part as an initial procedure. A resection or disconnection of the inferior part from a combined pterional/subfrontal approach can then be proposed as a second step.

„„ Surgical Results From 1998 to 2015, a total of 126 mostly pediatric patients with a HH have been surgically treated for a drug-resistant epilepsy at our institution. According to Delalande’s classification, 3% of the patients had a hamartoma of type 1, 60% of type 2, 27% of type 3, and 10% of type 4. Hamartoma disconnection was performed using either monopolar coagulation, the thulium laser, or, in a few cases, ultrasonic dissection, with the aid of a neuronavigated or robot-guided transventricular endoscopic route. Patients of the type 1 HH and a few of the type

601

602

IVf  Other Disconnective Procedures 3 or 4 were operated through an open pterional/­subfrontal approach, either in a single surgery or following endoscopic disconnection. Mean age at surgery in all pediatric and adult patients was 14 years. Mean age at seizure onset was 2.6 years. Gelastic seizures (laughing fits) were either the single manifestation or among other seizure types in 84% of the patients. Mental retardation was present in 41%. Endoscopic transventricular disconnection was performed in 96% of the patients: in a single procedure in 60% and in more than one procedure (2–5) in 40%. With a mean follow-up of 3.6 years, 70% of the patients with type 2 HH became seizure free, as compared to 58% for all HH types. Postoperative complications in our entire patient cohort included the great majority of these being transient, shortterm memory deficits (6); hemiparesis (4); hormonal dysfunction (12); third nerve palsy (5); meningitis (2); and hydrocephalus (1).

„„ Conclusion As for tumors within the third ventricle, open surgery for HH in this location bears a higher risk of neurological, vascular, or endocrine complications. Since HH are nongrowing masses of dystopic and disorganized but mature neuroglial tissue with a high epileptogenicity, a surgical disconnection should be sufficient to interrupt the seizure propagation from its onset within the hamartoma, not unlike other cerebral disconnections (lobar, multilobar, or hemispheric) that are practiced in epilepsy surgery. Patients with HH with exclusive or predominant attachment to the hypothalamus within the third ventricle (type 2 of Delalande’s surgical HH classification) are excellent candidates for the transventricular endoscopic approach for disconnection. In our population of 126 children and adults who were treated until 2015, 70% of the patients with type 2 HH became seizure free after one or more endoscopic surgeries. Surgery-related morbidity can be kept relatively low if one is aware of the instrument inclination and respective depth of the HH at each endoscope position during the procedure, in order to avoid interference with cranial nerves, major vessels, or the motor pathway. Of course, image-guided navigation does not replace a thorough preoperative study of the MRI

Fig. 64.9  A case of repeated endoscopic disconnection. One can see the scar formation of the first procedure at the hamartoma’s leftsided attachment.

images before surgery, in order to have best possible appreciation of the neuroanatomy in the close vicinity to the margins of the HH. In the case of repeated surgery, the scar formation at the ventricular wall along the disconnection line can be easily recognized and does not render an endoscopic re-disconnection more complicated than during the first procedure, particularly with respect to tissue consistency or bleeding tendency (Fig. 64.9). The advantage of the transventricular endoscopic disconnection of HH over other minimally invasive techniques is the continuous visual control during the entire procedure. In general, the endoscopic disconnection is easily tolerated even in young children, necessitates a short stay in the hospital, and will give immediate results following surgery if a sufficient extent of the HH attachment has been interrupted.

References 1. Munari C, Kahane P, Francione S, et al. Role of the hypothalamic hamartoma in the genesis of gelastic fits (a video-stereo-EEG study). Electroencephalogr Clin Neurophysiol 1995;95(3):154–160 2. Kuzniecky R, Guthrie B, Mountz J, et al. Intrinsic epileptogenesis of hypothalamic hamartomas in gelastic epilepsy. Ann Neurol 1997;42(1):60–67 3. Northfield DW, Russell DS. Pubertas praecox due to hypothalamic hamartoma: report of two cases surviving surgical removal of the tumour. J Neurol Neurosurg Psychiatry 1967;30(2):166–173 4. Rosenfeld JV, Harvey AS, Wrennall J, Zacharin M, Berkovic SF. Transcallosal resection of hypothalamic hamartomas, with control of seizures, in children with gelastic epilepsy. Neurosurgery 2001;48(1):108–118 5. Delalande O, Fohlen M. Disconnecting surgical treatment of hypothalamic hamartoma in children and adults with refractory

epilepsy and proposal of a new classification. Neurol Med Chir (Tokyo) 2003;43(2):61–68 6. Procaccini E, Dorfmüller G, Fohlen M, Bulteau C, Delalande O. Surgical management of hypothalamic hamartomas with epilepsy: the stereoendoscopic approach. Neurosurgery 2006; 59(4, Suppl 2):ONS336–ONS344, discussion ONS344–ONS346 7. Dorfmüller G, Fohlen M, Bulteau C, Delalande O. Surgical disconnection of hypothalamic hamartomas Neurochirurgie 2008;54(3):315–319 8. Calisto A, Dorfmüller G, Fohlen M, Bulteau C, Conti A, Delalande O. Endoscopic disconnection of hypothalamic hamartomas: safety and feasibility of robot-assisted, thulium laser-based procedures. J Neurosurg Pediatr 2014;14(6):563–572 9. Valdueza JM, Cristante L, Dammann O, et al. Hypothalamic hamartomas: with special reference to gelastic epilepsy and surgery. Neurosurgery 1994;34(6):949–958, discussion 958

65

  Multiple Subpial Transections in Children with Refractory Epilepsy Zulma S. Tovar-Spinoza and James T. Rutka

Summary The multiple subpial transection (MST) is a surgical technique used to treat patients with epileptogenic zones in functionally critical cortical areas where resective surgery can cause unacceptable neurological deficits. It is based on selective destruction of the short horizontal fiber connections and aims to ­prevent synchronization and spread of epileptogenic discharges while allowing ­preservation of the normal cortical functions. Subpially directed parallel cuts are performed until the entire proposed epileptogenic zone has been transected. MST can be performed as a standalone procedure or can be combined with cortical resection or lesionectomy. Focal seizures located in the eloquent cortex are the main indication for MST. Other indications include Landau-­ Kleffner syndrome, cortical dysplasia, epilepsia partialis continua, and ­Rasmussen’s encephalitis and malignant rolandic-sylvian epilepsy. Keywords:  multiple subpial transection, eloquent cortex, ­Landau-Kleffner syndrome

„„ Introduction The multiple subpial transection (MST) technique was described by Frank Morrell as a surgical procedure to treat patients with epileptogenic zones in functionally critical cortical areas where resective surgery can cause unacceptable neurological deficits.1,​2,​3 In theory, the MST technique is based on selective destruction of the short horizontal fiber connections and aims to prevent synchronization and spread of epileptogenic discharges while allowing preservation of the normal cortical functions.1,​2,​3,​4

„„ Fundamentals of the Multiple Subpial Transection Technique The rationale for MST is based on the following facts and assumptions: • Columnar organization of the cortex: MST was based on experimental evidences describing functional cortical units composed of vertically oriented neuronal elements and vertical afferent and efferent fibers.5,​6,​7,​8,​9 yy Synchronized neuronal discharges: An epileptogenic focus requires paroxysmal synchronous discharges from a critical

volume of neurons and tangential horizontal interneuronal projections to spread seizure activity.10,​11,​12,​13 yy The critical mass of cortical cells: A critical volume of contiguous neurons (1 cm3) is necessary to sustain synchronous spikes. It is believed that cortical areas greater than 5 mm width or tangential connections greater than 5 mm are indispensable for the generation of paroxysmal neuronal discharges.14,​15,​16 This observation explains the transection interval of 5 mm classically suggested for MST. yy Spread of epileptogenic discharges: The lateral radiating cortical–cortical interneuronal connections in all cortical layers—but mainly in layers IV through V of the cerebral cortex—are the main projections of spreading the seizure activity.17,​18 Morrell reasoned that interrupting this pattern of propagation of seizure activity might eliminate the spread of the epileptic discharges.11,​12,​13 yy Preservation of cortical blood supply: Anatomically, the gyral blood supply enters perpendicularly to the cortical surface, and the arterial flow and venous drainage have a parallel trajectory with the axonal fibers. Thus, MST technique would not disrupt the vascular supply to the cortex and preserve the integrity of the subpial bank.2

„„ Indications for the Multiple Subpial Transection Technique Although MST is a well-known technique, its application is still not widespread. MST was the least frequently performed epilepsy surgery procedure (on the order of 0.6%) among all cases reported by Harvey et al in their recently published survey from the Pediatric Epilepsy Surgery Subcommission of the International League Against Epilepsy.19

Focal Seizures Located in the Eloquent Cortex Currently, this is the main indication for MST. In children, as in adults, the MST can be performed as a standalone procedure or can be combined with cortical resection or lesionectomy.20,​21,​22 In one of the largest series, Shimizu and Maehara23 presented their experience with 31 children who underwent MST. The proce-

604

IVf  Other Disconnective Procedures dure was performed along with resective surgeries (lobectomy, corticectomy, or lesionectomy) in 25 patients and as a standalone procedure in the remaining 6 patients. Of 25 patients who were followed up for more than 1 year, 10 of them had Engel class I or II outcome, but no details were provided as to whether these patients had MST alone or combined with cortical resection. No mortality or morbidity was encountered during surgery or postoperatively. Later, Blount et al20 described their series of 30 consecutive children who underwent MST. The procedure was performed as a standalone therapy (4 patients) or in conjunction with cortical excisions (26 patients). Twenty-three children underwent invasive monitoring with subdural grid electrodes, and intraoperative electrocorticography was performed in the remaining seven children. The mean follow-up period for the group was 3.5 years (minimum of 30 months in all cases). All 20 patients in whom MST was performed in the primary motor cortex experienced transient hemiparesis lasting up to 6 weeks. No patient suffered a permanent motor deficit in the long-term follow-up period. In the 26 patients who underwent cortical resections followed by MST, 12 children (46%) were seizure free (Engel class I) after surgery. Eleven patients (42%; Engel classes II and III) continued to suffer seizures, but the improvement in seizures was still acceptable. In the 23 patients in whom subdural grids were placed to capture the ictal onset zone by invasive video-electro-encephalography (VEEG), MST area consisted of a mean of 37% of the surgically treated area under the grid. It is difficult to assess the effectiveness of MST as a distinct surgical procedure because the majority of published series include patients in whom MST is combined with lesionectomy or cortical resection. The use of MST as a standalone therapy has been described by Schramm et al,24 Smith,25 Lui et al,26 and Whisler.27 These series addressed both adult and pediatric populations. Schramm et al obtained 45 and 50% good results according to Engel’s and Spencer’s classifications, respectively. Seizure-free outcome was 5% in Schramm’s series, 37.5% in Smith’s series, and 63% in Whisler’s series. Téllez-Zenteno et al28 published a meta-analysis of several series of patients with MST, both pure and combined with resective surgery. The mean follow-up period was 5 years in this study. They concluded that MST has the lowest rate of long-term, seizure-free outcome (16%) among all epilepsy surgical procedures. Previously, Spencer et al29 reported their meta-analysis from six series and reviewed the data covering 211 patients collected from six different centers. Fifty-five patients in this study underwent pure MST, and 156 patients had MST and resection combined procedures. The overall seizure outcome was similar in the patients with partial seizures whether the patient had pure MST or MST with cortical resection. However, the patients with generalized seizures had better outcome if the MST was combined with cortical resections. Unfortunately, this study did not specify the postsurgical follow-up period for outcome analysis. This report and others concluded that MST could be considered an effective surgical treatment alternative for uncontrolled seizures arising in eloquent brain areas.22,​25,​26,​30,​31,​32,​33

Landau-Kleffner Syndrome Landau-Kleffner syndrome (LKS) has been considered as another main indication for MST in children.3 This rare syndrome consists of an acquired epileptic aphasia or verbal auditory a ­ gnosia that

may start abruptly or gradually and occurs in children with previously normal developmental history. A well-defined electrophysiological profile underlying this condition consists of frequent epileptiform discharges when awake, and generalized slow spike-wave discharges over the perisylvian region during slow wave sleep.34 Antiepileptic drugs can improve clinical seizures, but EEG findings can be frequently resistant to treatment.35 The MST technique was initially applied to the LKS patients by Morrell et al,3 and they reported their experience with 14 patients. Seven recovered age-appropriate speech ability and no longer needed speech therapy or special education classes. Another four had marked improvement in their speech but still required speech therapy. Three children had no changes in their preoperative conditions. Sawhney et al33 reported improvement in all three of their patients with LKS who underwent MST. Neville et al36 reported one case that had marked postoperative improvement in reading, vocabulary, sign language, and nonverbal subtests. Nass et al37 described their experience in seven patients with atypical forms of LKS and reported mild improvement in receptive language functions. Irwin et al38 reported five children with this condition who underwent MST and reported improvement in their language skills. Although none of them improved to an age-appropriate level, seizures and behavior disturbances were immediately improved in all cases. Castillo et al34 reported one case with significant linguistic improvement after a 2-year follow-up. In summary, MST application in LKS can result in improvements in communication skills and behavior after surgery. However, improvement in speech takes a considerably longer period of time.39,​40 Use of MST for severe autistic regression in childhood epilepsy is not always associated with improvement in cognitive and behavioral function. At best, gains observed are temporary, and further studies are needed to prove its worth in this condition.37

Syndrome of Malignant Rolandic-Sylvian Epilepsy This nonlesional syndrome described by Otsubo et al41 relates to children who have intractable rolandic partial seizures that progress to secondary generalization, fronto-centro-temporal spikes on EEG, localized spike sources in the rolandic-sylvian regions on magnetoencephalography (MEG), and neurocognitive problems. Resective surgery is required in nonfunctional cortex and as an additional management strategy. MSTs are also recommended throughout the eloquent cortex. The series included seven patients and reported that the combination of techniques could reduce or eliminate seizure activity, avoiding postoperative permanent motor deficits or further language deficits.

Other Indications MST has been reported in the treatment of patients with cortical dysplasia, epilepsia partialis continua (EPC), and Rasmussen’s encephalitis. Molyneux et al described the successful treatment of one patient with EPC using MST on epileptogenic focus over the left central cortex.42 This patient had a normal magnetic resonance imaging (MRI) scan, but a diagnosis of cortical dysplasia was made by biopsy. MSTs have also been attempted with variable results for EPC caused by Rasmussen’s encephalitis.33,​43

65  Multiple Subpial Transections in Children with Refractory Epilepsy

„„ Presurgical Evaluation Defining the seizure focus is the main goal of the presurgical evaluation for all patients who are candidates for epilepsy surgery procedures, including MST. Although different protocols exist depending on the institution, most epilepsy centers use many common techniques to assess the surgical candidates. Harvey et al19 recently published a survey collecting data from 20 programs in the United States, Europe, and Australia on 543 patients and all 20 centers reported using scalp EEG, VEEG, and MRI in the presurgical evaluation of all patients. Eight percent of the centers used ictal single proton emission computed tomography (SPECT), 85% used 2-deoxy-2[18F] ­f luoro-D-glucose (FDG)-positron emission tomography (PET), 70% used functional MRI (fMRI), usually for language localization, 35% used MEG/magnetic source imaging (MSI), and 50% performed an intracarotid amobarbital procedure (IAP; Wada test). Only three centers used all presurgical tests (ictal SPECT, FDG-PET, fMRI, MEG, and IAP). The epilepsy surgery protocol at the Hospital for Sick ­Children in Toronto includes ictal and interictal scalp EEG, ictal and interictal VEEG, and MRI with special sequences to evaluate myelination, cortical pathways, and to rule out neoplasms.21 In

Fig. 65.1  The multiple subpial transection (MST) knife is introduced through the incised pial hole and is subpially directed toward the sulcal margin, making a right-angle cut to the long axis of the gyrus to a depth of approximately 5 to 7 mm. The blade of the MST knife is maintained in a strictly vertical orientation and the tip of the MST knife is visualized through the pia mater as it is drawn back along the subpial space completing the transection. (Reproduced with permission from Tovar-Spinoza ZS, Rutka JT. Multiple subpial transections. In: Lozano A, Gildenberg P, Tasker R, eds. Textbook of Stereotactic and Functional Neurosurgery. 2nd ed. Berlin: Springer: 2009.)

addition to these, formal neurological and neuropsychological evaluations are performed to determine preoperative levels of verbal and memory functions and their lateralization. Language dominance is determined by Wada testing, fMRI, or f­ unctional MEG. MEG spikes source localization can be overlaid onto MRIs to generate an MSI.44 MSI can also be incorporated into ­neuronavigation systems to localize the focus of spike-wave disturbances at the time of surgery. The merging of the data from all of these evaluations will help define the epileptogenic focus amenable to surgical treatment.

„„ Surgical Technique The original technique of MST was described and developed by Morrell and Whisler.1,​2,​27 The bipolar electrocautery is used to cauterize a small pial point on the defined epileptogenic focus either at the side of the gyrus or at the crest.45 Then, the cortex is penetrated at this site with the tip of an 11 blade scalpel. A right angle blunt dissector is introduced through the incised pial hole and it is subpially directed toward the sulcal margin, making a right-angle cut to the long axis of the gyrus to a depth of approximately 5 to 7 mm (Fig. 65.1). The blade of the MST knife (Fig. 65.2) should be maintained in a strictly vertical orientation to avoid undercutting the cortex. The tip of the MST knife is visualized through the pia mater as it is drawn back along the subpial space completing the transection. The surgeon should avoid disrupting the pia matter or injuring sulcal vessels during the procedure. If the insertion point is made in the center of the gyrus, after the first half cut, the instrument is removed and reinserted, aiming in the opposite direction, and the remainder of the transection is completed. Another

Fig. 65.2  Multiple subpial transection knives.

605

606

IVf  Other Disconnective Procedures ­ ariation on the technique is to point the MST knife downward v using sharpened rather than blunt knives to diminish the damage of using blunt instruments.45 Parallel cuts (spaced 5 mm apart) are then made from this cut until the entire proposed epileptogenic zone has been transected. Pial bleeding at the blade insertion point can be controlled with a bipolar or a small piece of thrombin-soaked Gelfoam. Intraoperative ultrasound has been useful when MST is performed in extensive areas of the brain to rule out the presence of an acute intracerebral hematoma.21 The efficacy and extent of the procedure are judged intraoperatively by using electrocorticography before and after MST. However, the role of electrocorticography in predicting seizure outcome has been the subject of some controversy.20,​21,​46

„„ Complications In general, MST is associated with low morbidity. Although transient postoperative neurological deficits are common, permanent deficits lasting longer than 3 months are rare even when extensive regions of cortex are transected.20,​24,​27,​30,​32 Some of these complications include hemiparesis,25,​27,​29,​33,​47,​48 hemisensory deficits, memory deficits, visual field deficits,29 alteration in propioception,31 aphasia/dysphasia,24,​27,​29 and dysdiadochokinesis.47 Children usually show clinical improvement within 1 week of the procedure. In addition to small subarachnoid and intracerebral hemorrhages,27,​49 some unusually large and symptomatic hematomas have been reported.24,​27 Mild brain swelling is expected, but excessive brain swelling not related to hematoma formation has been reported only once.24

„„ Discussion There are many reasons for the wide range of variations in the application and results of MST. Differences in the presurgical evaluation, definition of the epileptogenic focus, indications, and surgical technique can considerably affect the seizure ­outcome.50 The surgical technique in MST is not standard. Different surgeons perform the procedure with some variations that can also potentially influence the outcome. Transection intervals and depth, transecting just the crown, or transecting the entire gyrus may easily affect the result of the procedure.

Again, failure to keep the transection perfectly vertical with respect to the gyral surface may result in sparing of the cortical tissue at the sulcus and the U fibers.31 Nontransected tissue bridges in layers IV and V or V and VI could allow for seizure propagation.18 Differences in the length of the curved MST dissectors might also play a factor in the completion of the transection ­appropriately.51 Another reason for variable results is patient selection. Although there is not a full consensus on the benefit of MST, the most common two indications for MST in children are epileptogenic focus in an eloquent area of the brain and LKS. The other indications are significantly more controversial, and the surgical criteria variable depending on the individual centers. One other reason for variable results is using MST technique as a standalone procedure or combined with resective surgeries. As discussed earlier, the role of the MST as standalone therapy is still controversial, and further studies are needed to clarify this controversy. In published studies, several authors have presented their outcome data based on different follow-up periods. It is a wellknown fact that seizures may change over time.2,​24,​52 Whether or not the initial seizure improvement seen in the first 1 to 2 years after surgery will persist over subsequent years is unclear. In addition, the lack of an accepted and standardized seizure outcome scale, and the use of different scales such as the Engel,53 modified Engel,52 or Spencer classification24 in series, makes comparisons difficult.

„„ Conclusion Current advances in the preoperative evaluation techniques in children with refractory epilepsy clearly show that pediatric patients benefit from definitive surgeries if the procedure is performed as early as possible. Certain groups advocate more aggressive approaches in children even if the epileptogenic focus is located in functionally critical cortical areas. However, frequently, the risk of a resective surgery in eloquent cortex is high and functionally unacceptable. MST constitutes a relatively safe surgical alternative in this patient group. Current data also suggest that MST is probably more effective when combined with cortical resection compared with its application as a standalone procedure. However, prospective studies are needed to assess the full potential of MST as standalone therapy in children.

References 1. Morrell F, Hanbery JW. A new surgical technique for the treatment of focal cortical epilepsy. Electroencephalogr Clin Neurophysiol 1969;26(1):120 2. Morrell F, Whisler WW, Bleck TP. Multiple subpial transection: a new approach to the surgical treatment of focal epilepsy. J Neurosurg 1989;70(2):231–239 3. Morrell F, Whisler WW, Smith MC, et al. Landau-Kleffner syndrome. Treatment with subpial intracortical transection. Brain 1995;118(Pt 6):1529–1546 4. Kaufmann WE, Krauss GL, Uematsu S, Lesser RP. Treatment of epilepsy with multiple subpial transections: an acute histologic analysis in human subjects. Epilepsia 1996;37(4):342–352

5. Asanuma H, Sakata H. Functional organization of a cortical efferent system examined with focal depth stimulation in cats. J Neurophysiol 1967;30(1):35–54 6. Asanuma H, Stoney SD Jr, Abzug C. Relationship between afferent input and motor outflow in cat motorsensory cortex. J Neurophysiol 1968;31(5):670–681 7. Hubel DH, Wiesel TN. Receptive fields, binocular interaction and functional architecture in the cat’s visual cortex. J Physiol 1962;160:106–154 8. Mountcastle VB. Modality and topographic properties of single neurons of cat’s somatic sensory cortex. J Neurophysiol 1957;20(4):408–434

65  Multiple Subpial Transections in Children with Refractory Epilepsy 9. Powell TP, Mountcastle VB. Some aspects of the functional organization of the cortex of the postcentral gyrus of the monkey: a correlation of findings obtained in a single unit analysis with cytoarchitecture. Bull Johns Hopkins Hosp 1959;105:133–162 10. Tharp BR. The penicillin focus: a study of field characteristics using cross-correlation analysis. Electroencephalogr Clin Neurophysiol 1971;31(1):45–55 11. Reichenthal E, Hocherman S. The critical cortical area for development of penicillin-induced epilepsy. Electroencephalogr Clin Neurophysiol 1977;42(2):248–251 12. Lueders H, Bustamante L, Zablow L, Krinsky A, Goldensohn ES. Quantitative studies of spike foci induced by minimal concentrations of penicillin. Electroencephalogr Clin Neurophysiol 1980;48(1):80–89 13. Lueders H, Bustamante LA, Zablow L, Goldensohn ES. The independence of closely spaced discrete experimental spike foci. Neurology 1981;31(7):846–851 14. Dichter M, Spencer WA. Penicillin-induced interictal discharges from the cat hippocampus. II. Mechanisms underlying origin and restriction. J Neurophysiol 1969;32(5):663–687 15. Dichter M, Spencer WA. Penicillin-induced interictal discharges from the cat hippocampus. I. Characteristics and topographical features. J Neurophysiol 1969;32(5):649–662 16. Goldensohn ES, Zablow L, Salazar A. The penicillin focus. I. Distribution of potential at the cortical surface. Electroencephalogr Clin Neurophysiol 1977;42(4):480–492 17. Ebersole JS, Chatt AB. The laminar susceptibility of cat visual cortex to penicillin induced epileptogenesis. Neurology 1980;30:355 18. Telfeian AE, Connors BW. Layer-specific pathways for the horizontal propagation of epileptiform discharges in neocortex. Epilepsia 1998;39(7):700–708 19. Harvey AS, Cross JH, Shinnar S, Mathern GW; ILAE Pediatric Epilepsy Surgery Survey Taskforce. Defining the spectrum of international practice in pediatric epilepsy surgery patients. Epilepsia 2008;49(1):146–155 20. Blount JP, Langburt W, Otsubo H, et al. Multiple subpial transections in the treatment of pediatric epilepsy. J Neurosurg 2004;100(2, Suppl Pediatrics):118–124 21. Benifla M, Otsubo H, Ochi A, Snead OC III, Rutka JT. Multiple subpial transections in pediatric epilepsy: indications and outcomes. Childs Nerv Syst 2006;22(8):992–998 22. Guénot M. [Surgical treatment of epilepsy: outcome of various surgical procedures in adults and children] Rev Neurol (Paris) 2004;160(Spec No 1):S241–S250 23. Shimizu H, Maehara T. Neuronal disconnection for the surgical treatment of pediatric epilepsy. Epilepsia 2000;41(Suppl 9):28–30 24. Schramm J, Aliashkevich AF, Grunwald T. Multiple subpial transections: outcome and complications in 20 patients who did not undergo resection. J Neurosurg 2002;97(1):39–47 25. Smith MC. Multiple subpial transection in patients with extratemporal epilepsy. Epilepsia 1998;39(Suppl 4):S81–S89 26. Liu Z, Zhao Q, Li S, Tian Z, Cui Y, Feng H. Multiple subpial transection for treatment of intractable epilepsy. Chin Med J (Engl) 1995;108(7):539–541 27. Whisler WW. Multiple subpial transection. Tech Neurosurg 1995;1:40–44 28. Téllez-Zenteno JF, Dhar R, Wiebe S. Long-term seizure outcomes following epilepsy surgery: a systematic review and meta-analysis. Brain 2005;128(Pt 5):1188–1198 29. Spencer SS, Schramm J, Wyler A, et al. Multiple subpial transection for intractable partial epilepsy: an international meta-analysis. Epilepsia 2002;43(2):141–145 30. Mulligan LP, Spencer DD, Spencer SS. Multiple subpial transections: the Yale experience. Epilepsia 2001;42(2):226–229 31. Pacia SV, Devinsky O, Perrine K, et al. Multiple subpial transections for intractable partial seizure: seizures outcome. J Epilepsy 1997;10:86–91

32. Rougier A, Sundstrom L, Claverie B, et al. Multiple subpial transection: report of 7 cases. Epilepsy Res 1996;24(1):57–63 33. Sawhney IMS, Robertson IJ, Polkey CE, Binnie CD, Elwes RD. Multiple subpial transection: a review of 21 cases. J Neurol Neurosurg Psychiatry 1995;58(3):344–349 34. Castillo EM, Butler IJ, Baumgartner JE, Passaro A, Papanicolaou AC. When epilepsy interferes with word comprehension: findings in Landau-Kleffner syndrome. J Child Neurol 2008; 23(1):97–101 35. Buelow JM, Aydelott P, Pierz DM, Heck B. Multiple subpial transection for Landau-Kleffner syndrome. AORN J 1996;63(4):727– 729, 732–735, 737–739, quiz 741–744 36. Neville BG, Harkness WF, Cross JH, et al. Surgical treatment of severe autistic regression in childhood epilepsy. Pediatr Neurol 1997;16(2):137–140 37. Nass R, Gross A, Wisoff J, Devinsky O. Outcome of multiple subpial transections for autistic epileptiform regression. Pediatr Neurol 1999;21(1):464–470 38. Irwin K, Birch V, Lees J, et al. Multiple subpial ­ transection in Landau-Kleffner syndrome. Dev Med Child Neurol 2001; 43(4):248–252 39. Harkness W. How to select the best surgical procedure for children with epilepsy. In: Epilepsy Surgery. Luders H, ed. Philadelphia, PA: Lippincott Williams & Wilkins; 2001:767–780 40. Grote CL, Van Slyke P, Hoeppner JA. Language outcome following multiple subpial transection for Landau-Kleffner syndrome. Brain 1999;122(Pt 3):561–566 41. Otsubo H, Chitoku S, Ochi A, et al. Malignant rolandic-sylvian epilepsy in children: diagnosis, treatment, and outcomes. Neurology 2001;57(4):590–596 42. Molyneux PD, Barker RA, Thom M, van Paesschen W, Harkness WF, Duncan JS. Successful treatment of intractable epilepsia partialis continua with multiple subpial transections. J Neurol Neurosurg Psychiatry 1998;65(1):137–138 43. Nakken KO, Eriksson AS, Kostov H, et al. [Epilepsia partialis continua (Kojevnikov’s syndrome)] Tidsskr Nor Laegeforen 2005;125(6):746–749 44. Otsubo H, Oishi M, Snead OCI. Magnetoencephalography. In: Miller J, Silbergeld D, eds. Epilepsy Surgery: Principles and Controversies Neurological Disease and Therapy. New York, NY: Marcel Dekker Inc; 2007:752–767 45. Wyler AR. Multiple subpial transections in neocortical epilepsy: Part II. Adv Neurol 2000;84:635–642 46. Wennberg R, Quesney LF, Lozano A, Olivier A, Rasmussen T. Role of electrocorticography at surgery for lesion-related frontal lobe epilepsy. Can J Neurol Sci 1999;26(1):33–39 47. Hufnagel A, Zentner J, Fernandez G, Wolf HK, Schramm J, Elger CE. Multiple subpial transection for control of epileptic seizures: effectiveness and safety. Epilepsia 1997;38(6): 678–688 48. Patil AA, Andrews R, Torkelson R. Isolation of dominant seizure foci by multiple subpial transections. Stereotact Funct Neurosurg 1997;69(1–4, Pt 2):210–215 49. Shimizu H, Suzuki I, Ishijima B, Karasawa S, Sakuma T. Multiple subpial transection (MST) for the control of seizures that originated in unresectable cortical foci. Jpn J Psychiatry Neurol 1991;45(2):354–356 50. Wyler A. Multiple subpial transections: a review and arguments for use. In: Miller J, Silbergeld D, eds. Epilepsy Surgery. New York, NY: Taylor & Francis Group, LLC; 2006:524–529 51. Patil AA, Andrews RV, Torkelson R. Surgical treatment of intractable seizures with multilobar or bihemispheric seizure foci (MLBHSF). Surg Neurol 1997;47(1):72–77, discussion 77–78 52. Orbach D, Romanelli P, Devinsky O, Doyle W. Late seizure recurrence after multiple subpial transections. Epilepsia 2001;42(10):1316–1319 53. Engel J, Van Ness PC, Rasmussen TB, Ojemann LM, et al. Outcome with respect to epileptic seizures. In: Engel J Jr, ed. Surgical Treatment of the Epilepsies. New York, NY: Raven Press; 1993:609–621

607

66

  Hippocampal Subpial Transection Hiroyuki Shimizu

Summary Hippocampal transection was first reported in 2006 as a new surgical treatment for temporal lobe epilepsy patients. The rationale of hippocampal transection is based on the theory of multiple subpial transection (MST). Vertical cuts with hippocampal transection disrupt longitudinal fibers within the hippocampus without damaging the vertical fibers and allow the disruption of seizure propagation without effecting hippocampal functions. Therefore, using this method, postoperative verbal memory can be preserved even in patients with left mesial temporal lobe epilepsy without hippocampal atrophy. We used hippocampal transection technique in 45 patients with ages ranging from 2 to 42 years and 78% of the patients had Engel class I outcome. Hippocampal transection is a good alternative technique in patients with mesial temporal epilepsy on the dominant side. Keywords:  mesial temporal epilepsy, hippocampal transection, longitudinal fiber, hippocampus

„„ Introduction Risk of postoperative verbal memory dysfunction in epilepsy patients with minimal or no hippocampal sclerosis is higher and can be observed in pediatric patients as well as in adults after left temporal lobectomy.1,​2,​3 Hippocampal transection was first reported in 2006 as a new surgical treatment for this group of temporal lobe epilepsy patients.4 Using this method, postoperative verbal memory can be preserved even in patients with left mesial temporal lobe epilepsy without hippocampal atrophy. We also applied this new surgical technique to pediatric patients and obtained good seizure outcome and preservation of verbal function. The overall surgical technique and surgical outcomes, including adult and pediatric patients, are presented here and a representative pediatric case is illustrated.

„„ Rationale of Hippocampal Subpial Transection The hippocampal pathway is closely related with verbal memory. It originates in the inferior temporal association cortex and projects directly onto CA1 pyramidal neurons after running through the perirhinal cortex.5 Hippocampus con-

tains two circuits: the transverse lamellae and longitudinal ­pathways. Transverse lamellae are organized perpendicularly to the length of hippocampus and include perforant pathways which are important for memory processing. On the other hand, longitudinal pathways have been found to be important for synchronization and propagation of seizure discharges.6,​7,​8,​9 Therefore, vertical cuts with hippocampal transection disrupt longitudinal fibers within the hippocampus without damaging the vertical fibers and allow the ­disruption of seizure propagation without effecting hippocampal functions.6,​7,​10

„„ Surgical Exposure of the Hippocampus Based on the previously mentioned neurofunctional data, we place a small corticotomy within 4.5 cm from the temporal tip. After aspirating the gray matter of the superior temporal gyrus along the sylvian fissure, the temporal stem is exposed. If the temporal stem is aspirated just beneath the gray matter of the insula, then the temporal horn can be easily opened (Fig. 66.1a). The anterolateral part of the ventricle is suctioned as widely as possible and the hippocampal head and amygdala are fully exposed.

„„ Electrocorticography over the Hippocampus and Amygdala Before starting surgical procedures, specially designed electrodes are placed over the hippocampus and amygdala. Electrocorticography (ECoG) is recorded in many points over the hippocampus. For this purpose, two small strip electrodes with four contacts for exploring the hippocampal body and tail and two square electrodes with four contacts for recording in the hippocampal head and amygdala are applied. Therefore, ECoG is recorded at 12 contacts all over the hippocampus and the amygdala (Fig. 66.1b, c).

„„ Hippocampal Transection The rationale of hippocampal transection is based on the theory of multiple subpial transection (MST) developed by Morrell

66  Hippocampal Subpial Transection et al.11 Because the pyramidal cell layer of the hippocampus is within 2 mm from the surface,5 we devised a ring transector 2 mm in diameter (Fig. 66.2a). Because the alveus covering the pyramidal layer is a very firm, fibrous tissue, the alveus is sharply cut with microscissors. A 2-mm ring transector is inserted through this slit and the pyramidal layer is transected 4 mm apart using the same distance used for MST. At the bilateral corners of the CA4 portion and near the subiculum, the pyramidal layer becomes deeper and a 4-mm ring transector is used for transection. Toward the posterior portion of the hippocampus, the width of the hippocampus becomes narrower and an oval-formed transector with a 4-mm-long diameter is applied for transection of the bilateral corners (Fig. 66.2b). Fimbria is left intact.

The transection areas of the pyramidal cell layer are determined based on the results of intraoperative ECoG. After areas with epileptic discharges are transected, ECoG is repeated to detect residual epileptic activity. If residual spikes are found, transection is also performed until prominent spike areas completely disappear (Fig. 66.2c).

„„ Seizure Outcome We performed hippocampal transection in 45 patients, left side in 23 and right side in 22. The patients consisted of 22 men and 23 women. Patient ages ranged from 2 to 42 years, with a mean of 25 years old. In all patients, preoperative MRI did not demon-

Fig. 66.1  (a) Access from the superior temporal gyrus to the temporal horn is shown. First a small corticotomy is made on the surface of the superior temporal gyrus within 4.5 cm from the tip. Along the sylvian fissure (dotted line), the gray matter of the superior temporal gyrus is aspirated to reach the temporal stem. By sectioning the temporal stem, the temporal horn is opened, and the hippocampus and amygdala are confirmed. (b) Two small strip electrodes with four contacts and two square plate electrodes with four contacts are arrayed over the hippocampus and the amygdala to record electrocorticography (ECoG). (c) Intraoperative ECoG demonstrated distribution of epileptic discharges.

609

610

IVf  Other Disconnective Procedures

Fig. 66.2  (a) Three types of ring transectors were designed: the upper transector shown here has a diameter of 2 mm, the middle transector has a 4-mm diameter, and the lower transector has an oval ring 4 mm long. (b) Transection of the pyramidal layer (small dotted area) is shown. The tough membrane of the alveus is cut with microscissors, and ring transectors are inserted through the slit. A transector with a 2-mm ring is used to expose the surface of the hippocampus. However, at the bilateral corners, 4-mm ring transectors are used as the pyramidal layer becomes deeper. (c) Surgical view of the transected hippocampal surface is shown. At the head of the hippocampus, transection lines are made along the hippocampal digitations.

strate any sign of atrophy or asymmetry of the hippocampus (Fig. 66.3). In six patients, organic lesions were confirmed in front of the hippocampus. In 12 cases, intracranial subdural electrodes were inserted, and laterality of the mesial temporal focus was determined. In the remaining patients, laterality was diagnosed based on scalp electroencephalography (EEG) with sphenoidal lead, single photon emission computed tomography (SPECT), and neuropsychometry data. Within 2 weeks after surgery, postoperative magnetic resonance imaging (MRI) was examined. Compared with preoperative MRIs, a tract along the sylvian fissure aspirating the gray matter of the superior temporal gyrus was confirmed. However, there was no deformity of the hippocampus after hippocampal transection (Fig. 66.4). In 36 patients who were followed up for more than 1 year, 28 (78%) patients were categorized in Engel class I, 4 (11%) patients were categorized in class II, 3 patients (8%) were categorized in class III, and 1 patient (3%) was categorized in class IV.

„„ Postoperative Verbal Memory The auditory verbal learning test (AVLT) is known as a very sensitive test for evaluating short-term verbal memory.12 AVLT was used as an objective index for evaluation of pre- and postoperative hippocampal function. In cases of left hippocampal transection, AVLT scores showed a transient decrease immediately after surgery. However, within 6 months, decreased scores generally recovered to preoperative levels. Approximately 6 months after surgery, transected hippocampus usually showed slight atrophy. Memory function recovered despite this atrophy and there was no decline of memory thereafter. In some cases, there was no worsening of AVLT even immediately after surgery. This may be because of the extent of surgical procedures added to hippocampal transection. If surgical procedures were confined to hippocampal transection only, postoperative AVLT scores were

66  Hippocampal Subpial Transection generally kept at preoperative levels. However, when other procedures, such as MST of the lateral temporal cortex or resection of the temporal basal area, were performed in addition to hippocampal transection, there was a greater tendency toward immediate decline of AVLT scores and later recovery.

„„ Pediatric Data and Case Report We performed hippocampal transection in eight pediatric patients (18 years or less). Patients consisted of six boys and two girls, with ages ranging from 2 to 17 years (mean age = 10 years). Postoperative follow-up ranged from 1.7 to 5.8 years. All patients became seizure free except for one patient who demonstrated rare residual seizures. Because AVLT examination was difficult to perform in most of the children, their verbal memory function was evaluated by developmental quotient. Most patients showed various degrees of improvement in the speech–social category.

Fig. 66.3  In our series of hippocampal transection, preoperative MRI did not show any hippocampal atrophy or asymmetry in any of the patients.

An Illustrative Case This patient is a 5-year-old boy with no history contributory to the present illness. His development had been normal until his first seizure at the age of 2 years and 11 months. Thereafter, habitual seizures started with blank staring, followed by meaningless manual movements. Seizures usually continued for approximately 1 minute. When the patient was standing, he often collapsed gradually after the start of seizures. Seizures mainly occurred in the morning 5 to 10 times every day. He had speech delay, and he often showed aggressive behaviors. He was referred to our clinic to evaluate the possibility of surgical treatment. Neurological examination did not detect any abnormality. Scalp EEG showed high-amplitude multiple spikes in the left temporal area. MRI showed abnormal intensity in the left temporal pole, suggestive of cortical tuber or other type of dysgenesis. There was no asymmetry of the hippocampus (Fig. 66.5). SPECT demonstrated low perfusion in the left temporal area. According to neuropsychological examination, he showed developmental delay with a developmental quotient of 49 for both speech–social and cognition–adaptation ability. Based on the previously mentioned data, left temporal lobe epilepsy was diagnosed, and he underwent left temporal lobe surgery. According to intraoperative ECoG, epileptic areas were treated with preservation of brain functions. The left hippocampus was transected, and the temporal tip with temporal horn area on MRI was resected in front of the inferior ventricle. MST was applied extensively over the lateral temporal cortex. The postoperative course was uneventful. The histopathology of the resected temporal pole showed cortical dysplasia. Since surgery, he has been completely seizure free for more than 2 years. He is now able to concentrate more in class, and his mental state has become calmer and more stable. His neuropsychological evaluation 2 years after surgery showed development in speech ability with a developmental quotient of 58 compared with a preoperative score of 49.

„„ Other Series Since the first edition of this book, several other groups have published their results using hippocampal transection

Fig. 66.4  (a, b) On MRI after hippocampal transection, the tract from the surface to the inferior horn (arrows) could be demonstrated. However, there was no deformity of the transected hippocampus (dotted arrow).

611

612

IVf  Other Disconnective Procedures

Fig. 66.5  (a, b) Preoperative MRI showing a dysplastic lesion in the left temporal pole (a). However, there was no atrophy of the hippocampus on the same side (b).

­technique.6,​7,​10,​13,​14 These are mainly adult series, but it is still worthwhile to summarize their results here. Patil and Andrews performed hippocampal transection in 15 patients with a minimum follow-up duration of 2 years.6,​7 They reported seizure-free outcome (Engel class 1) in 94.7% of the patients and improvement on verbal memory in 77.8% of them. ­Another series was published by Koubeissi et al which included 13 patients.10 Their seizure-free outcome rate was 77% and they reported no loss of memory or cognitive decline. Uda et al published their experience with hippocampal transection in 37 patients.14 This group also adopted some modifications in original hippocampal tran-

section technique. They used transsylvian approach to access the temporal horn and also transected the parahippocampal gyrus and entorhinal area using a blunt ring-shaped dissector inserting it from the innominate sulcus along the incision lines of the alveus. They emphasized the significance of complete transection while preserving fimbria carefully. Uda et al reported Engel class I seizure-free outcome in 25 patients (67.6%) with an additional Engel class II outcome in 10 patients (27%). They reported significant postoperative improvement in verbal memory in right-side cases and no change in left-side cases.14

References 1. Szabó CA, Wyllie E, Stanford LD, et al. Neuropsychological effect of temporal lobe resection in preadolescent children with epilepsy. Epilepsia 1998;39(8):814–819 2. Dlugos DJ, Moss EM, Duhaime AC, Brooks-Kayal AR. Language-­ related cognitive declines after left temporal lobectomy in children. Pediatr Neurol 1999;21(1):444–449 3. Gleissner U, Sassen R, Lendt M, Clusmann H, Elger CE, Helmstaedter C. Pre- and postoperative verbal memory in ­ pediatric patients with temporal lobe epilepsy. Epilepsy Res 2002;51(3):287–296

9. Umeoka SC, Lüders HO, Turnbull JP, Koubeissi MZ, Maciunas RJ. Requirement of longitudinal synchrony of epileptiform discharges in the hippocampus for seizure generation: a pilot study. J Neurosurg 2012;116(3):513–524 10. Koubeissi MZ, Kahriman E, Fastenau P, et al. Multiple hippocampal transections for intractable hippocampal epilepsy: seizure outcome. Epilepsy Behav 2016;58:86–90 11. Morrell F, Whisler WW, Bleck TP. Multiple subpial transection: a new approach to the surgical treatment of focal epilepsy. J Neurosurg 1989;70(2):231–239

4. Shimizu H, Kawai K, Sunaga S, Sugano H, Yamada T. Hippocampal transection for treatment of left temporal lobe epilepsy with preservation of verbal memory. J Clin Neurosci 2006;13(3):322–328

12. Rosenberg SJ, Ryan JJ, Prifitera A. Rey Auditory-Verbal Learning Test performance of patients with and without memory impairment. J Clin Psychol 1984;40(3):785–787

5. Duvernoy H. The Human Hippocampus. Berlin, Germany: Springer; 1998;26–37

13. Sunaga S, Morino M, Kusakabe T, Sugano H, Shimizu H. Efficacy of hippocampal transection for left temporal lobe ­ epilepsy without hippocampal atrophy. Epilepsy Behav 2011;21(1):94–99

6. Patil AA, Chamczuk AJ, Andrews RV. Hippocampal transections for epilepsy. Neurosurg Clin N Am 2016;27(1):19–25 7. Patil AA, Andrews R. Long term follow-up after multiple hippocampal transection (MHT). Seizure 2013;22(9):731–734 8. Amaral D, Lavenex P. Hippocampal neuroanatomy. In: Andersen P, Morris R, Amaral D, bliss T, O’Keefe J, eds. The Hippocampus Book. Oxford, UK: Oxford University Press; 2006

14. Uda T, Morino M, Ito H, et al. Transsylvian h ­ippocampal transection for mesial temporal lobe epilepsy: surgical ­ ­indications, procedure, and postoperative seizure and memory outcomes. J Neurosurg 2013;119(5):1098–1104

67

  Vagal Nerve Stimulation Jarod L. Roland, David D. Limbrick Jr., and Matthew D. Smyth

Summary Vagal nerve stimulation (VNS) is a safe and effective procedure for palliating medically refractory epilepsy in patients who are not candidates for more invasive cranial procedures. The initial studies of VNS in humans were limited to individuals over 12 years of age with partial epilepsy. The pediatric age range has since been further studied and FDA approval as of 2017 includes patients 4 years of age and older. Nevertheless, numerous studies since that time have expanded the indications to younger children with various epilepsy syndromes. In this chapter, we summarize the relevant literature leading to the development and modern application of VNS. To begin, we elucidate the early theoretical conceptions that motivated animal studies, which set the stage for the initial human trials. We then give an overview of subsequent clinical trials that broadened the scope of application. We focus our discussion on studies most relevant to pediatric patients, much of which has now been composed in systematic reviews and meta-analysis. Then we describe the surgical procedure for VNS implantation, including potential complications. We conclude with details of the hardware and its typical use. The information reviewed in this chapter should give the reader a firm foundation to fully understand VNS treatment and its use in pediatric patients. Keywords:  vagus nerve, vagal nerve stimulation, epilepsy, neuromodulation, pediatric

„„ Introduction Implantable devices for chronic vagal nerve stimulation (VNS) have been utilized since 1988 as an adjunct in the treatment of medically refractory epilepsy.1 Since that time, a substantial body of literature, including several randomized controlled trials (RCTs),2,​3,​4,​5,​6,​7 has confirmed the usefulness of VNS in refractory epilepsy. Here, we review the state of the art in vagal nerve stimulation and its application in pediatric epilepsy.

„„ Anatomy and Physiology of Vagal Nerve Stimulation Approximately 80% of the vagus nerve is composed of afferent fibers,8 including myelinated A and B fibers and unmyelinated C fibers. General sensory afferents from the pharynx and larynx, external auditory apparatus, and posterior fossa menin-

ges ­travel through the vagus to the spinal trigeminal tract and nucleus. Perhaps more relevant to VNS physiology, visceral sensory fibers ascending through the vagus project diffusely throughout the central nervous system.9 Several targets of these projections are involved in pathways for rhythmic excitability or have the potential for epileptogenicity themselves. Converging data suggest that the effects on seizures result from the direct and indirect modulation of central seizure networks.10,​11 Though the precise mechanism of seizure suppression by VNS remains to be elucidated, progress has been made in identifying central effects of peripheral VNS. VNS has been shown to increase neuronal discharges from the locus ceruleus and dorsal raphe nucleus, and chemically induced lesions of the locus ceruleus attenuate VNS-induced seizure suppression. Cross-modulatory pathways exist between the nucleus of the tractus solitarius, the locus ceruleus, and the dorsal raphe nucleus, implicating ascending transmitter pathways in the action of VNS. Microdialysis experiments have demonstrated that VNS increases norepinephrine levels in the amygdala, hippocampus, and cerebral cortex. Indeed, locus ceruleus neurons are the primary source of norepinephrine to the hippocampus and the cortex, strongly implicating this nucleus in the mechanism of action of VNS. For more detailed review, see the study of Rufolli et al.12

„„ Preclinical History of Vagal Nerve Stimulation The first description of vagal nerve stimulation for seizure control was by Dr. James Corning in 1884.13 Based on the hypothesis that cerebral hyperemia produced seizures, Dr. Corning devised an “electrocompressor,” which combined a transcutaneous vagal nerve stimulator with mechanical carotid compression. Despite simultaneously decreasing cardiac output and occluding carotid flow, Corning’s effects on seizure reduction were inconsistent, and his techniques were abandoned. In the 1930s, Bailey and Bremer demonstrated that stimulation of the vagal nerve affected electroencephalography (EEG) patterns in cats.14 Subsequent reports confirmed that manipulation of the vagus by ligation15 or stimulation16,​17,​18 decreased EEG spikes in various animal models of seizures. Based largely on these reports, safety and efficacy studies were conducted in primates19 prior to proceeding with implanting the first VNS devices in humans.1,​20

614

IVg  Neuromodulation Procedures

„„ Clinical Trials for Vagal Nerve Stimulation in Epilepsy The initial case series describing the use of VNS in humans was published in 1990 and involved four adult patients with intractable partial epilepsy.1 Early case series reporting favorable outcomes20,​21,​22,​23 led to the formation of the Vagus Nerve ­ Stimulation Study Group to coordinate RCTs evaluating the safety and efficacy of VNS.24 In conjunction with Cyberonics Inc., three open-label and two double-blind RCTs were conducted. The first trial (E03) included 114 patients older than 12 years (mean: 33.3 years) with more than six seizures per month while on anti-epileptic drugs (AEDs).25 Patients were randomized to receive low-stimulation (30 seconds every 90 minutes; 1 Hz; 130 μs; ≤ 3.5 mA) or high-stimulation (30 seconds every 5 minutes; 30 Hz; 500 μs; up to 3.5 mA) VNS. After 14 weeks of VNS, the high-stimulation group exhibited a 30.9% reduction in seizure frequency compared to an 11.3% reduction in the low-stimulation group, a statistically significant difference. The second major trial (E05) enrolled 199 patients older than 12 years (mean: 33.2, range: 13–60) with intractable ­partial-onset seizures.2 Patients were again randomized to low- or high-stimulation for 3 months and evaluated for seizure reduction and device safety. The high-stimulation group exhibited a 28% reduction in seizure frequency versus a 15% decrease in the low-stimulation group. In addition, the high-stimulation group demonstrated improved global evaluation scores, but dysphonia and dyspnea also were more common in this group. These initial studies led the United States Food and Drug Administration (FDA) to approve VNS (Cyberonics Inc., Houston, TX) in 1997 as an adjunctive therapy for use in medically refractory epilepsy patients older than 12 years with partial-onset seizures. In 1999, long-term efficacy and safety results from the five Vagus Nerve Stimulation Study Group clinical trials were reported in a meta-analysis.24 A total of 454 patients were followed up for up to 3 years. Although only 1 to 2% of patients became seizure free, many derived long-term benefit from VNS. Median reduction of seizure frequency increased from 35% at 1 year to 44.1% at 3 years following VNS initiation. The number of patients achieving greater than 50% seizure reduction also increased with time from 23% at 3 months to 43% at 3 years. A subsequent meta-analysis published in 2011 reviewed 74 clinical studies including 3,321 patients.26 Again, a greater effect over time was observed with a 36% reduction in seizures at 3 to 12 months and 51% reduction at more than 1 year. Many of these studies broadened the indications for VNS and a significant benefit was observed in generalized epilepsy and children.

„„ Vagal Nerve Stimulation in Pediatric Epilepsy As noted earlier, the studies conducted by the Vagus Nerve Stimulation Study Group enrolled only patients older than ­ 12 years. Based on preliminary results from a small series of children in which VNS was placed for medically and surgically refractory seizures,27 the Pediatric VNS Study Group reported

a compassionate use, prospective, open safety trial in 1999.28 A total of 60 pediatric patients (