Head and Neck Cancer Imaging [3 ed.] 9783030647346, 9783030647353

Table of contents :
Preface
Contents
Epidemiology, Risk Factors, Pathology, and Natural History of Head and Neck Neoplasms
1 Epidemiology: Frequency Measures and Risk Factors
1.1 Frequency Measure: Incidence
1.2 Risk Factors for the Development of Head and Neck Malignancies
1.2.1 Risk Factors for Development of HNSCC
1.2.2 Risk Factors for Development of Glandular Neoplasms
2 Pathology and Natural History of Frequent Benign and Malignant Head and Neck Neoplasms
2.1 Epithelial Neoplasms of the Mucous Membranes
2.1.1 Tumour Typing and Clinical Behaviour
2.1.1.1 Benign Lesions
2.1.1.2 Premalignant Lesions
2.1.1.3 Malignant Lesions
2.1.1.4 Natural History Before and at Diagnosis
2.1.1.5 Natural History Following Diagnosis and Successful Treatment of Malignant HNSCC
2.1.1.6 Microscopical Negative Prognostic Findings
2.2 Glandular Neoplasms
2.2.1 Thyroid Neoplasia
2.2.1.1 Benign Disease: Multinodular Enlargement
2.2.1.2 Benign Disease: Uninodular Enlargement—The Solitary Thyroid Nodule
2.2.1.3 Malignant Disease
2.2.1.4 Papillary Thyroid Cancer (PTC)
2.2.1.5 Follicular Thyroid Cancer (FTC)
2.2.1.6 Hürthle Cell Carcinoma
2.2.1.7 Medullary Carcinoma
2.2.1.8 Anaplastic Carcinoma
2.2.2 Salivary Gland Neoplasia
2.2.2.1 Tumour Typing and Clinical Behaviour
2.2.2.2 Benign Tumours
Pleomorphic Adenoma
Warthin’s Tumour
2.2.3 Malignant Tumours
2.2.3.1 Mucoepidermoid Carcinoma
2.2.3.2 Adenoid Cystic Carcinoma
2.2.3.3 Acinic Cell Carcinoma
2.2.3.4 Adenocarcinoma Not Otherwise Specified (NOS)
References
Clinical and Endoscopic Examination of the Head and Neck
1 Introduction
2 Neck
3 Nose and Paranasal Sinuses
4 Nasopharynx
5 Oral Cavity
6 Oropharynx
7 Larynx
8 Hypopharynx and Cervical Oesophagus
9 Salivary Glands
10 Thyroid Gland
11 Role of Imaging Studies
References
Imaging Techniques
1 Introduction
2 Plain Radiography
3 Ultrasonography
4 Computed Tomography and Magnetic Resonance Imaging
4.1 Computed Tomography
4.1.1 Patient Positioning
4.1.2 Contrast Agent Injection
4.1.3 Data Acquisition and Image Reconstruction
4.1.3.1 General Comments
4.1.3.2 Dose Reduction
4.1.3.3 Multidetector Spiral CT
4.1.3.4 Dual Energy CT
4.1.4 Dynamic Maneuvers
4.1.5 Three-Dimensional Image Reformatting
4.2 Magnetic Resonance Imaging
4.2.1 Patient Positioning
4.2.2 Coils
4.2.3 Standard Sequences
4.2.4 Contrast Agents
4.2.5 Additional MRI Techniques
4.2.5.1 Dynamic Contrast-Enhanced Magnetic Resonance Imaging
4.2.5.2 Diffusion-Weighted Magnetic Resonance Imaging
5 Positron Emission Tomography
5.1 Physical Aspects
5.2 Radiopharmaceuticals
5.2.1 Imaging of Glucose Metabolism: 18Fluorodeoxyglucose
5.2.2 Imaging of Tumor Proliferation: 18Fluorothymidine
5.2.3 Imaging of Amino Acid Metabolism: 18FET and 11C-MET
5.2.4 Imaging of Hypoxia
5.2.5 Imaging of Molecular Targets
5.3 Technical Aspects of FDG-PET and Integrated FDG-PET/CT in Head and Neck Cancer
5.4 PET/MRI
References
Laryngeal Neoplasms
1 Introduction
2 Normal Laryngeal Anatomy
2.1 Laryngeal Skeleton
2.2 Mucosal Layer and Deeper Laryngeal Spaces
2.3 Normal Radiological Anatomy
3 Squamous Cell Carcinoma
3.1 General Imaging Findings
3.2 Neoplastic Extension Patterns of Laryngeal Cancer
3.2.1 Glottic Cancer
3.2.1.1 Local Tumor Spread
3.2.1.2 Lymphatic Spread
3.2.2 Supraglottic Cancer
3.2.2.1 Suprahyoid Epiglottis
3.2.2.2 Infrahyoid Epiglottis
3.2.2.3 Aryepiglottic Fold and Arytenoid
3.2.2.4 False Vocal Cords
3.2.2.5 Lymphatic Spread
3.2.3 Subglottic Cancer
4 Prognostic Factors for Local Outcome of Laryngeal Cancer
4.1 Treatment Options
4.1.1 Glottic Cancer
4.1.2 Supraglottic Cancer
4.2 Impact of Imaging on Treatment Choice and Prognostic Accuracy
4.3 Use of Imaging Parameters as Prognostic Factors for Local Outcome Independently from the TN-Classification
4.3.1 Predicting Local Outcome After Radiotherapy
4.3.1.1 Tumor Volume and Deep Tissue Infiltration
4.3.1.2 Cartilage Involvement
4.3.1.3 Imaging of the Tumoral Micro-Environment
4.3.2 Predicting Local Outcome After Surgery
4.3.3 Towards Risk Profiles Incorporating Imaging Findings
5 Posttreatment Imaging in Laryngeal Cancer
5.1 Expected Findings After Treatment
5.1.1 Expected Tissue Changes After Radiotherapy
5.1.2 Expected Findings After Laryngeal Surgery
5.1.2.1 Laser Resection
5.1.2.2 Partial Laryngectomy
5.1.2.3 Total Laryngectomy
5.2 Persistent or Recurrent Cancer
5.2.1 Imaging Strategies and Findings
5.2.2 Potential Value of Imaging Surveillance
5.3 Treatment Complications
5.3.1 Complications After Surgery
5.3.2 Complications After Radiotherapy
5.3.2.1 Laryngeal Necrosis
5.3.2.2 Other Complications After Radiotherapy
6 Non-squamous Cell Laryngeal Neoplasms
6.1 Minor Salivary Gland Neoplasms
6.2 Mesenchymal Malignancies
6.2.1 Chondrosarcoma
6.2.2 Other Mesenchymal Malignancies
6.3 Hematopoietic Malignancies
6.3.1 Lymphoma
6.3.2 Plasma Cell Neoplasms
6.3.3 Metastasis
References
Neoplasms of the Hypopharynx and Proximal Esophagus
1 Introduction
2 Anatomy
2.1 Descriptive Anatomy
2.2 Imaging Anatomy
3 Pathology
3.1 Non-squamous Cell Malignancies
3.2 Squamous Cell Malignancies
3.2.1 Risk Factors
3.2.2 Clinical Presentation
3.2.3 Growth Pattern
3.2.4 Nodal Chain Involvement
3.2.5 Detection of Distant Metastasis
3.2.6 TNM Classification
3.3 Secondary Involvement by Other Tumors
4 Cross-Sectional Imaging
5 Radiologist’s Role
5.1 Pretreatment
5.1.1 Submucosal Spread
5.1.2 Cartilage Involvement
5.1.3 Tumor Volume
5.2 During Treatment
5.3 Posttreatment
5.3.1 Post Surgery
5.3.2 Postradiation Therapy
5.4 Detection of Second Primary Tumors
References
Neoplasms of the Oral Cavity
1 How to Assess a Tumor in the Oral Cavity
1.1 Modalities
1.2 Tumor Localization
1.3 Tumor Measurements
1.4 Infiltration of Adjacent Structures
1.4.1 Lip Cancer
1.4.2 Gingival and Buccal Cancer
1.4.3 Retromolar Trigone Cancer
1.4.4 Hard Palate Cancer
1.4.5 Tongue Cancer
1.4.6 Floor of Mouth Cancer
1.4.7 Advanced Tumor Spread (T4a and T4b Stages)
1.5 Perineural Spread
1.6 Lymph Nodes
1.7 TNM, Eighth Edition
2 How to Report an MRI or CT of an Oral Cavity Malignancy
2.1 New Malignancies
2.2 Posttreatment Follow-Up
2.2.1 Radiotherapy Changes Versus Residual/Recurrent Tumor
2.2.2 Radionecrosis Versus Residual/Recurrent Tumor
3 The Black Swans: Diagnoses Other Than SCC
3.1 Infection or Inflammation
3.2 Osteoradionecrosis
3.3 Lymphoma
3.4 Salivary Gland Tumors
3.4.1 Sublingual Gland Tumors
3.4.2 Minor Salivary Gland Tumors
3.5 Other Differential Diagnoses
4 Useful Anatomical Landmarks
4.1 Buccal Mucosa Anatomy
4.2 Floor of Mouth
4.3 Extrinsic Tongue Muscles
References
Neoplasms of the Oropharynx
1 Introduction
2 Normal Anatomy
3 Squamous Cell Carcinoma
3.1 Tonsillar Cancer
3.2 Tongue Base Cancer
3.3 Soft Palate Cancer
3.4 Posterior Oropharyngeal Wall Cancer
3.5 Lymphatic Spread
4 Treatment
5 Post-treatment Imaging
6 Other Neoplastic Disease
6.1 Non-Hodgkin Lymphoma
6.2 Salivary Gland Tumours
6.3 Other
References
Nasopharyngeal Neoplasms
1 Introduction
1.1 Histologic Subtypes
1.2 Risk Factors
2 Nasopharyngeal Imaging Anatomy
3 Clinical Features and Pathologic Anatomy of the Nasopharynx
3.1 Clinical Presentation and Evaluation
3.2 Imaging Evaluation
3.2.1 Normal Appearance
3.3 Local Extension and Patterns of Spread
3.3.1 Anterior Spread
3.3.2 Lateral Spread
3.3.3 Posterior Spread
3.3.4 Superior Spread
3.3.5 Orbital and Paranasal Sinus Involvement
3.3.6 Perineural Tumour Spread and Intracranial Extension
3.3.7 Carotid Artery Encasement
3.4 T-Staging
4 Metastatic Disease
4.1 Nodal Metastases
4.1.1 Imaging Evaluation
4.2 Distant Metastases
4.2.1 Metastatic Workup
5 Staging and Treatment
6 Post-treatment Changes and Follow-Up
6.1 Post-treatment Changes
6.1.1 Resolution
6.1.2 Residual and Recurrent Tumour
6.2 Post-treatment Changes
6.2.1 Skull Base Osteoradionecrosis
6.2.2 Radiation-Induced Brain Necrosis
6.2.3 Radiation-Induced Tumours
6.2.4 Brain Stem and Spinal Cord Encephalomyelopathy
6.2.5 Radiation-Induced Cranial Neuropathy
6.2.6 Vascular Complications
6.2.7 Xerostomia and Trismus
6.2.8 Radiation-Induced Lung Disease
7 Future Directions?
8 Other Nasopharyngeal Neoplasms and Infections
8.1 Pleomorphic Adenoma
8.2 Inflammatory Pseudotumour
8.3 Lymphoma
8.4 Adenoid Cystic Carcinoma (ACC)
References
Parapharyngeal Space Neoplasms
1 Introduction
2 Anatomy
2.1 Fascial Layers and Compartments
2.2 Radiological Anatomy
3 Imaging Findings in Parapharyngeal Space Lesions
3.1 Primary Lesions of the Parapharyngeal Space
3.1.1 Prestyloid Lesions
3.1.2 Retrostyloid Lesions
3.2 Secondary Lesions of the Parapharyngeal Space
4 Conclusion
References
Malignant Lesions of the Masticator Space
1 Introduction
2 Imaging Techniques
3 General Imaging Features of MS Masses
4 Specific Imaging Features of Primary MS Malignancies
5 Specific Imaging Features of Secondary MS Malignancies
6 Post-treatment Imaging
7 Benign Lesions Mimicking MS Malignancies
8 Conclusion
References
Neoplasms of the Sinonasal Cavities
1 Introduction
2 Normal Radiological Anatomy
3 Indications for Imaging Studies
4 Imaging Appearance and Extension Patterns of Sinonasal Neoplasms
4.1 Appearance of the Tumor Mass on CT and MRI
4.2 Extension Toward Surrounding Structures
4.2.1 Nasoethmoidal Pattern
4.2.2 Maxillary Sinus Pattern
5 Tumor Types
5.1 Epithelial Tumors
5.1.1 Benign Epithelial Tumors
5.1.1.1 Sinonasal Papillomas
5.1.1.2 Sinonasal Ameloblastoma
5.1.1.3 Salivary Gland Adenomas
5.1.1.4 Respiratory Epithelial Adenomatoid Hamartoma (REAH) and Seromucinous Hamartoma (SH)
5.1.2 Malignant Epithelial Tumors
5.1.2.1 Squamous Cell Carcinoma
5.1.2.2 Sinonasal Undifferentiated Carcinoma and Neuroendocrine Carcinoma
5.1.2.3 Adenocarcinoma
5.1.2.4 Salivary Gland-Type Carcinomas
5.1.2.5 Nuclear Protein in Testis Midline Carcinoma (NUT) Carcinoma and Teratocarcinosarcoma
5.1.2.6 Staging of Sinonasal Carcinomas
5.2 Non-epithelial Tumors
5.2.1 Neuroectodermal Tumors
5.2.1.1 Olfactory Neuroblastoma
5.2.1.2 Ewing Sarcoma (ES)/Peripheral Neuroectodermal Tumor (PNET)
5.2.1.3 Mucosal Melanoma
5.2.2 Soft Tissue Tumors
5.2.2.1 Benign Soft Tissue Tumors
5.2.2.2 Malignant Soft Tissue Tumors
5.2.2.3 Borderline/Low-Grade Soft Tissue Tumors
5.2.3 Osseous and Cartilaginous Tumors
5.2.3.1 Benign Fibro-osseous Tumors
5.2.3.2 Osteosarcoma
5.2.3.3 Chondrosarcoma
5.2.4 Hematolymphoid Tumors
5.2.4.1 Lymphoma
5.2.4.2 Extramedullary Plasmacytoma
5.2.5 Metastasis
6 Treatment Monitoring
References
Parotid Gland and Other Salivary Glands Tumors
1 Introduction
2 Anatomy
3 Imaging Issues
4 Parotid Benign Tumors
4.1 Pleomorphic Adenoma or Benign Mixed Tumor
4.1.1 General Description
4.1.2 Histologically
4.1.3 Imaging Findings
4.1.4 Differential Diagnosis (See Sect. 6)
4.2 Warthin Tumor (Adenolymphoma)
4.2.1 General Description
4.2.2 Histologically
4.2.3 Imaging Findings
4.2.4 Differential Diagnosis
4.3 Other Benign Tumors
4.3.1 Lipoma
4.3.2 Facial Nerve Schwannoma
4.3.2.1 Differential Diagnosis
4.3.3 Oncocytoma
4.4 Congenital Tumors
4.4.1 Lymphangioma
4.4.2 Infantile Hemangioma
4.5 Cystic Tumors
4.5.1 Solitary Cystic Lesion
4.5.2 Dermoid Cysts
4.5.3 Epidermoid Cysts
4.5.4 Multiple Intraparotid Cystic Lesions
5 Parotid Malignant Tumors
5.1 Histologic Classification
5.2 Imaging Findings
5.2.1 Parotid Cancer
5.2.2 Non-Hodgkin Lymphoma
6 Strategy in Difficult Cases
7 Pseudo-Tumors of the Parotid Gland
7.1 Sjögren’s Syndrome
7.2 Sarcoidosis
7.3 Intraparotid Lymph Nodes
8 Tumors of the Other Salivary Glands
8.1 Minor Salivary Glands Tumor
8.2 Submandibular Gland Tumors
8.3 Sublingual Gland Tumor
9 Conclusion
References
Malignant Lesion of the Central and Posterior Skull Base
1 Introduction
2 Anatomy
2.1 Central Skull Base
2.2 Posterior Skull Base
3 Clinical Presentation
4 Normal Anatomical Variations
5 Pathology
5.1 Malignant Lesions Causing Diffuse or Multi-focal Skull Base Involvement
5.2 Mimics of Malignant Lesions Causing Diffuse or Multi-focal Skull Base Involvement
5.3 Non-region Specific, Localized Malignant Skull Base Lesions
5.4 Mimics of Non-region Specific, Localized Malignant Skull Base Lesions
5.5 Malignant Central Skull Base Lesions
5.6 Mimics of Malignant Central Skull Base Lesions
5.7 Malignant Lesions at the Junction of Central to Posterior Skull Base
5.8 Malignant Posterior Skull Base Lesions
5.9 Mimics of Malignant Posterior Skull Base Lesions
6 Imaging Protocol
7 Radiologist’s Role
References
Thyroid and Parathyroid Neoplasms
1 Introduction
2 Thyroid Anatomy
3 Thyroid Gland Imaging Modalities
4 Thyroid Nodules
5 Thyroid Cancer
5.1 Papillary Thyroid Cancer
5.2 Follicular Adenoma and Follicular Thyroid Cancer
5.3 Anaplastic Thyroid Carcinoma
5.4 Medullary Thyroid Cancer
5.5 Thyroid Lymphoma
5.6 Thyroid Metastases
6 Imaging of Papillary Microcarcinoma
7 Post-operative Thyroid Cancer Imaging
8 Parathyroid Imaging Modalities
9 Parathyroid Adenoma
10 Parathyroid Carcinoma
References
Neck Nodal Disease
1 Introduction
2 Nodal Group Classification and Pathways of Lymphatic Drainage
3 Imaging Modalities
3.1 CT and MRI
3.2 US and US-Guided Fine-Needle Aspiration Cytology (US-FNAC)
3.3 FDG-PET Imaging
3.4 Lymphoscintigraphy for Sentinel Node Localisation
4 Imaging Criteria for Malignant Nodes
4.1 Size and Nodal Clustering
4.2 Shape
4.3 Hilum
4.4 Vascular Pattern
4.5 Internal Heterogeneity
4.6 Border Irregularity
4.7 FDG-PET Uptake
5 Advanced Techniques
6 Micrometastases
7 Nodal Staging
8 Impact of Nodal Imaging on Patient Management
8.1 Detection of Metastatic Nodes
8.2 Extranodal Extension and Infiltration of Adjacent Structures
8.3 Identification of Patients at High Risk for Distant Metastases
9 Treatment Assessment
9.1 Prediction of Treatment Response to (Chemo)Radiotherapy
9.2 Post-treatment Assessment
9.3 Post-treatment Surveillance
10 Brief Overview of Non-HNSCC Lymphadenopathies
10.1 Lymphoma
10.2 Thyroid Cancer
10.3 Salivary Gland Carcinoma
10.4 Nasopharyngeal Carcinoma
10.5 Skin Cancer
11 Squamous Cell Carcinoma of Unknown Primary
12 Non-malignant Lymphadenopathy
13 Conclusion
References
Neck Lymphoma
1 Introduction
1.1 Epidemiology
1.2 Etiology
1.3 Pathology and Classifications
2 Hodgkin’s Lymphoma
3 Non-Hodgkin’s Lymphomas (NHL) and Specific Entities
3.1 B Cell Neoplasms
3.2 T Cell and Natural Killer (NK)-Cell Neoplasms
3.3 Hodgkin’s Lymphoma (Hodgkin’s Disease)
4 Workup
4.1 Diagnosis
4.2 Initial Imaging
4.3 Staging
5 Treatment
6 Response Assessment
7 Nodal Disease
7.1 The Common Sites
7.2 The Uncommon Sites
8 Extranodal Disease
8.1 Waldeyer’s Ring and the Upper Aerodigestive Tract
8.1.1 Nasopharynx
8.1.2 Tonsillar Fossa
8.1.3 Base of Tongue
8.1.4 Larynx
8.2 Orbit
8.2.1 Conjunctiva
8.2.2 Intra-orbital Lymphoma
8.2.3 Lacrimal Gland
8.3 Salivary Glands
8.3.1 Parotid Gland
8.4 Sinonasal Cavities
8.5 Thyroid
8.6 Bone
8.6.1 Primary Lymphoma of Bone
8.6.2 Multiple Myeloma (Kahlers’ Disease)
8.6.3 Extramedullary Plasmacytoma
8.7 Skin
9 Conclusion
References
Positron Emission Tomography in Head and Neck Cancer
1 Introduction
2 Clinical Applications
2.1 Pretreatment
2.1.1 Primary Tumor Staging
2.1.2 Nodal Staging
2.1.3 Detection of Distant Metastasis and Second Primary Tumors
2.1.4 Detection of Unknown Primary Tumors
2.2 Treatment Planning
2.3 Treatment Surveillance
2.3.1 Posttreatment Evaluation of the Primary Site
2.3.2 Posttreatment Evaluation of Nodal Disease
2.3.3 Posttreatment Evaluation of Distant Metastases
2.4 Special Considerations for Some Histological Tumor Types
2.4.1 Salivary Gland Tumors
2.4.2 Bone Lesions
2.4.3 Neuroendocrine Tumors
References
Use of Imaging in Radiotherapy for Head and Neck Cancer
1 Introduction
2 General Principles of Radiotherapy for Head and Neck Cancer
2.1 Evolution of Treatment Fields
2.2 Photon Versus Proton Therapy
3 Overview of Imaging Modalities Used in Radiotherapy
3.1 CT
3.1.1 Use
3.1.2 Advantages
3.1.3 Limitations
3.2 MRI
3.2.1 Use
3.2.2 Advantages
3.2.3 Limitations
3.3 PET
3.3.1 Use
3.3.2 Advantages
3.3.3 Limitations
4 Applications of Imaging Data in Radiation Oncology
4.1 Diagnosis and Staging
4.2 Radiotherapy Planning
4.2.1 Anatomic Imaging
4.2.2 Functional Imaging
4.2.2.1 Dose Painting
4.2.2.2 Proliferation
4.2.2.3 Hypoxia
4.2.2.4 Apoptosis
4.2.2.5 Receptor Status
4.3 Treatment Verification
4.4 Response Prediction
4.5 Follow-Up
5 Conclusion and Challenges
References

Citation preview

Medical Radiology · Diagnostic Imaging Series Editors: Hans-Ulrich Kauczor · Paul M. Parizel · Wilfred C.G. Peh

Robert Hermans   Editor

Head and Neck Cancer Imaging Third Edition

Medical Radiology Diagnostic Imaging Series Editors Hans-Ulrich Kauczor Paul M. Parizel Wilfred C. G. Peh

The book series Medical Radiology – Diagnostic Imaging provides accurate and up-to-date overviews about the latest advances in the rapidly evolving field of diagnostic imaging and interventional radiology. Each volume is conceived as a practical and clinically useful reference book and is developed under the direction of an experienced editor, who is a world-renowned specialist in the field. Book chapters are written by expert authors in the field and are richly illustrated with high quality figures, tables and graphs. Editors and authors are committed to provide detailed and coherent information in a readily accessible and easy-to-understand format, directly applicable to daily practice. Medical Radiology  – Diagnostic Imaging covers all organ systems and addresses all modern imaging techniques and image-guided treatment modalities, as well as hot topics in management, workflow, and quality and safety issues in radiology and imaging. The judicious choice of relevant topics, the careful selection of expert editors and authors, and the emphasis on providing practically useful information, contribute to the wide appeal and ongoing success of the series. The series is indexed in Scopus. For further volumes: http://www.springer.com/series/4354

Robert Hermans Editor

Head and Neck Cancer Imaging Third Edition

Editor Robert Hermans Department of Radiology KU Leuven - University Hospitals Leuven Leuven Belgium

ISSN 0942-5373     ISSN 2197-4187 (electronic) Medical Radiology ISBN 978-3-030-64734-6    ISBN 978-3-030-64735-3 (eBook) https://doi.org/10.1007/978-3-030-64735-3 © Springer Nature Switzerland AG 2006, 2012, 2021 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

To Isabelle, our children and grandchild Bob Hermans

Preface

The head and neck is a region of considerable anatomical and functional complexity, making the accurate staging of a head and neck neoplasm a challenging task. Many neoplasms in this region originate from the mucosa and are detectable by the clinician. However, the submucosal tumor extension, as well as the possible regional and distant disease spread cannot be completely assessed based on the clinical examination alone. Modern imaging modalities visualize the head and neck structures to an unprecedented level of detail. If carefully performed and interpreted, these techniques allow a comprehensive evaluation of the extent of head and neck neoplasms. The radiologist is an important member of the multidisciplinary team managing head and neck cancer patients. The value of imaging techniques in treatment choice, monitoring tumor response, and following-up patients after treatment is firmly established. Ongoing research continues to enforce the impact of imaging in oncologic patient care. The purpose of this book is to provide a comprehensive review of state-of-­ the-art head and neck cancer imaging. Several distinguished head and neck radiologists have contributed to this book, fully covering the field of advanced imaging in the head and neck cancer patient. Compared to the previous edition of this book, several chapters have been rewritten, incorporating recent insights and knowledge. In other chapters, the role of metabolic and functional imaging is included in more detail. Clinical-diagnostic techniques, as well as therapeutic strategies, also have changed significantly over the past years; in this regard, I would like to thank my colleague-clinicians from Leuven who contributed to this book. Care has been taken to situate the role of imaging within these developments. The ultimate goal of all medical actions is to provide our patients with the best possible therapy for their health problems. Hopefully this book contributes to achieving this purpose. Leuven, Belgium

Robert Hermans

vii

Contents

 Epidemiology, Risk Factors, Pathology, and Natural History of Head and Neck Neoplasms����������������������������������������������������   1 Vincent Vander Poorten Clinical and Endoscopic Examination of the Head and Neck ������������  21 Pierre Delaere and Jeroen Meulemans Imaging Techniques ��������������������������������������������������������������������������������  37 Robert Hermans, Frederik De Keyzer, Vincent Vandecaveye, and Laurens Carp Laryngeal Neoplasms������������������������������������������������������������������������������  65 Robert Hermans  Neoplasms of the Hypopharynx and Proximal Esophagus������������������ 115 Ilona M. Schmalfuss Neoplasms of the Oral Cavity ���������������������������������������������������������������� 145 Anouk van der Gijp and Frank Pameijer  Neoplasms of the Oropharynx���������������������������������������������������������������� 173 Robert Hermans Nasopharyngeal Neoplasms�������������������������������������������������������������������� 191 Julian Goh and Amit Karandikar Parapharyngeal Space Neoplasms���������������������������������������������������������� 237 Robert Hermans Malignant Lesions of the Masticator Space������������������������������������������ 253 Bela Purohit and Robert Hermans  Neoplasms of the Sinonasal Cavities������������������������������������������������������ 283 Davide Farina, Davide Lombardi, Giovanni Palumbo, and Marco Ravanelli  Parotid Gland and Other Salivary Glands Tumors������������������������������ 319 Frédérique Dubrulle, Ophélie Guillaud, Mathieu Nobile, and Christophe Jandeaux  Malignant Lesion of the Central and Posterior Skull Base������������������ 353 Ilona M. Schmalfuss

ix

x

Thyroid and Parathyroid Neoplasms ���������������������������������������������������� 387 Steve Colley Neck Nodal Disease���������������������������������������������������������������������������������� 405 Kunwar S. S. Bhatia and Ann D. King Neck Lymphoma�������������������������������������������������������������������������������������� 441 Frank A. Pameijer and Rick L. M. Haas Positron Emission Tomography in Head and Neck Cancer ���������������� 467 Ilona M. Schmalfuss  Use of Imaging in Radiotherapy for Head and Neck Cancer�������������� 495 Sandra Nuyts and Sarah Deschuymer

Contents

Epidemiology, Risk Factors, Pathology, and Natural History of Head and Neck Neoplasms Vincent Vander Poorten

Contents  pidemiology: Frequency Measures and E Risk Factors  1.1  Frequency Measure: Incidence  1.2  Risk Factors for the Development of Head and Neck Malignancies 

Abstract

1

 2  2  3

 athology and Natural History of P Frequent Benign and Malignant Head and Neck Neoplasms  2.1  Epithelial Neoplasms of the Mucous Membranes  2.2  Glandular Neoplasms 

 6  9

References 

 18

2

 6

V. Vander Poorten (*) Otorhinolaryngology, Head and Neck Surgery and Department of Oncology, Section Head and Neck Oncology, University Hospitals Leuven, KU Leuven, Leuven, Belgium e-mail: [email protected]

This introductory chapter sets the scene for the book by defining the complex domain in which the head and neck radiologist is expected to make his diagnosis. The epidemiology of epithelial and non-epithelial head and neck tumours is discussed representing up-to-­ date frequency measures. The known risk factors are subsequently reported. After that the clinical and pathological specifics of the most frequent tumours are presented, along with the expected natural history, so that the head and neck radiologist is aware of the different stages of the disease and the radiological “snapshots” that can result from imaging at different points in the evolution of the disease. Both macroscopic and microscopic aspects are illustrated by to-the-point clinical and light-microscopical pictures.

Head and neck cancer can be divided in two major groups. The largest group, the epithelial malignancies of the mucosal membranes of the upper aerodigestive tract, is called head and neck squamous cell carcinoma (HNSCC) and accounts for more than 90% of all head and neck neoplasms (Bray et  al. 2018). The second, smaller group, are the “glandular neoplasms”, arising in the thyroid and in the salivary glands.

Med Radiol Diagn Imaging (2020) https://doi.org/10.1007/174_2020_221, © Springer Nature Switzerland AG Published Online: 23 April 2020

1

V. Vander Poorten

2

Skin cancer is considered a separate entity. Non-melanoma head and neck skin cancer includes mainly squamous cell carcinoma and basal cell carcinoma, the latter being 3–4 times more frequent than the former (Lomas et  al. 2012). Infrequent head and neck neoplasia includes localized lymphoma, soft tissue and bone sarcoma, and neuroectodermal tissue tumours (paraganglioma, olfactory neuroblastoma, neuroendocrine carcinoma, ­ malignant melanoma). For information on these types we refer the reader to the specific head and neck oncology literature. In this introductory chapter the first paragraph deals with epidemiology and risk factors of head and neck neoplasms. An overview of the pathology and natural history of the most frequent benign and malignant head and neck neoplasms is outlined in the second paragraph.

1

Epidemiology: Frequency Measures and Risk Factors

1.1

Frequency Measure: Incidence

Head and neck cancer, excluding skin cancer and Hodgkin and non-Hodgkin lymphoma localized in the head and neck, is the sixth most frequent cancer worldwide. The 2018 world incidence of epithelial malignancies of the mucous membranes is 834,860 new patients per year, and these are 177,422 laryngeal cancer patients, 354,864 patients with cancer of the lip and oral cavity, and 302,574 patients with pharyngeal cancer (nasopharynx: n  =  129,079; oropharynx n = 92,887; hypopharynx n = 80,608). Thus, in 2018, 7.7% of the global incidence of cancer could be attributed to these neoplasms. Together these cancers account for 431,131 deaths yearly (Bray et al. 2018). Additionally, in 2018, head and neck oncologists treated 567,233 new patients with thyroid cancer (3.1% of the global cancer incidence) and 52,799 new patients with salivary gland cancer (0.3% of the global cancer incidence). Although thyroid carcinoma is thus ten times more frequent than salivary gland cancer, relatively much more

patients with thyroid cancer are cured. This is reflected in the reported mortality due to thyroid cancer of 41,071 patients in 2018, as compared to a salivary gland cancer mortality of 22,176 (Bray et al. 2018). Comparing the two largest groups, HNSCC and thyroid cancer, a definite gender difference in incidence is apparent. As an example, the most recent world incidence of laryngeal SCC shows a male-female ratio of 7/1, whereas for incidence of thyroid cancer, the odds are opposite with a ratio of 1/3.5 (Bray et al. 2018). The incidence of thyroid cancer has been steadily increasing in the last 40  years with a factor 2.3, mainly due to a rise in papillary thyroid cancer, while the incidence of other types remained unchanged. This rise is mainly due to better detection methods and awareness, but also to a true rise in incidence as reflected by an increased incidence of large tumours and tumours displaying extrathyroidal extension (Sipos and Mazzaferri 2010; Morris and Myssiorek 2010). The incidence of salivary gland cancer is at the subpercentual level when looking at cancer in general, but is responsible for between 1% and 3% of head and neck cancer incidence (Vander Poorten et al. 2012). There is an important geographical variation in head and neck cancer incidence (Lambert et al. 2011; Bray et al. 2013). A specifically high incidence is observed in much of Southern Asia, Australia, Brazil, Southern Africa, and parts of Central and Southern Europe. Nasopharyngeal cancer typically has a high incidence in southern China. Oral cavity cancer is the most frequent cancer in India. The incidence of hypopharyngeal cancer is typically very high in Northern France (10/100,000 males per year) as compared to, e.g. the United States of America (2/100,000 males per year). The incidence of laryngeal cancer in Northern Spain (20/100,000/year) is about 200 times as high as compared to certain regions in China (0.1/100,000/year). Besides probable differences in genetic susceptibility, a different prevalence of strong risk factors (e.g. alcoholic beverages drinking, smoking habits) undoubtedly explains these differences. In the same way can observed differences in incidence among races (higher incidence in African versus

Epidemiology, Risk Factors, Pathology, and Natural History of Head and Neck Neoplasms

3

Caucasian Americans), and among men and women, be largely attributed to differences in risk factor exposure (DeSantis et al. 2013).

1.2

Risk Factors for the Development of Head and Neck Malignancies

1.2.1 R  isk Factors for Development of HNSCC The most important risk factor is chronic use of tobacco (smoking and smokeless such as betel quid chewing) and alcohol (Fig.  1). The reason why these factors are so important is twofold: a strong association with the disease on the one hand, and a high prevalence among the population on the other. They are two independent risk factors that, already a long time ago, have been shown clearly to act in a multiplicative way when combined. Figure  2 shows that a 5.8 times increased risk for development of oral and pharyngeal cancer is observed in non-smokers who use 30 or more drinks per week, a 7.4 times increased risk in smokers not using alcohol with a history of 40 or more pack-years (smoking 20 cigarettes per day during 40 years), whereas the person combining the two has a 38 times increased risk (Blot et  al. 1988; Hashibe et  al. 2009; Giraldi et al. 2017). Conversely, after cessation of the use of tobacco, the risk of oral mucosal dysplasia and cancer falls to the level in the population that never smoked after 15–20  years (Morse et  al. 1996; Marron et  al. 2010). A pooled analysis based on over 11,000 cases and 16,000 controls shows that approximately 72% of HNSCC cancers are attributable to these two exposures, ranging from 64% for oral cavity cancer, over 72% for pharyngeal cancer, to 89% for laryngeal cancer. The strong interaction, Blot et al. already described in 1988 between the two exposures, was again confirmed in 2007 (Hashibe et al. 2007), and again in 2011 where a large case control study quantified that tobacco and alcohol together explain 73% of the burden of cancers of the upper aerodigestive tract (UADT). In these, 29% was attributable to smoking alone, less than

Fig. 1  Smoking is the most prevalent and most powerful risk factor for the development of HNSCC.  A doubled incidence of Warthin’s tumour of the parotid gland has also been observed

1% due to alcohol alone, and 44% by the combined effect of tobacco and alcohol. The latter “toxic duo”, tobacco and alcohol combined, was mainly responsible for hypopharyngeal/laryngeal cancer (Population Attributable Risk, PAR of 85%) than oropharyngeal (PAR = 74%), oesophageal (PAR = 67%), and oral cancer (PAR = 61%) (Anantharaman et al. 2011). The carcinogens in tobacco are nitrosamines, polycyclic aromatic hydrocarbons, and aldehydes. Nitrosamines are alkylating agents that induce mutational events. Alcohol acts as a solvent and thus enhances permeability of the mucosa for the toxic substances in tobacco. A direct effect of alcohol is ascribed to mucosal enzymatic formation (alcohol dehydrogenase) of the carcinogenic acetaldehyde. This was recently supported by the finding that individuals homozygous for the ∗2 allele of aldehyde ­dehydrogenase 2 (ALDH2) that do not support ­alcohol intake because of their inability to metabolize acetaldehyde have a significantly reduced

V. Vander Poorten

4

Relative Risk for Oropharyngeal Cancer

40 35 >30 drk/wk

30 25

15-29 drk/wk

20

5-14 drk/wk

15 1-4 drk/wk

10 5 0

3 cm but 6 cm in greatest dimension and without signs of ENE Metastasis in any lymph node with clinical ENE

ENE extranodal extension Table 2  cN classification of lymph node metastasis from HPV-mediated (P16+) squamous cell carcinoma of the oropharynx (eighth edition TNM classification. Brierley et al. 2017, Amin et al. 2017) Nx N0 N1 N2 N3

Regional lymph nodes cannot be assessed No regional lymph node metastasis Metastasis in single or multiple ipsilateral lymph nodes, all ≤6 cm in greatest dimension Metastasis in bilateral or contralateral lymph nodes, all ≤6 cm in greatest dimension Metastasis in a lymph node >6 cm in greatest dimension

FDG-PET scan, fails to unveil a primary tumour, then the diagnosis of metastatic carcinoma to a cervical lymph node from an unknown primary (carcinoma of unknown primary or CUP) is established.

3  Nose and Paranasal Sinuses The nasal cavity is the beginning of the upper airway and is divided in the midline by the nasal septum. Laterally, the nasal cavity contains the

Clinical and Endoscopic Examination of the Head and Neck

nasal conchae, the inferior concha being part of the nasal cavity, and the superior and middle­ ­conchae being composite parts of the ethmoid complex. The nasal cavity is surrounded by air containing bony spaces called paranasal sinuses, the largest of which, the maxillary antrum, is present on each side. The ethmoid air cells occupy the superior aspect of the nasal cavity, and separate it from the anterior skull base at the level of the cribriform plate. Superoanteriorly, the frontal sinus contained within the frontal bone forms a biloculated pneumatic space. The sphenoid sinus at the superoposterior part of the nasal cavity is located at the roof of the nasopharynx. The adult ethmoid sinus is narrowest anteriorly in a section known as the ostiomeatal complex and this is the site of drainage of the maxillary and frontal sinuses (Fig. 3). Since all of the paranasal sinuses are contained within bony spaces, primary tumours of epithelial origin seldom produce symptoms until they are of significant dimensions, causing obstruction, or

25

until they have broken through the bony confines of the involved sinus cavity. Tumours of the nasal cavity often produce symptoms of nasal obstruction, epistaxis, or obstructive pansinusitis early during the course of the disease. Unilateral epistaxis, obstruction, or sinusitis should raise the index of suspicion regarding the possibility of a neoplastic process. Tumours of the maxillary antrum may present with symptoms of obstructive maxillary sinusitis. Swelling of the upper gum or loose teeth may be the first manifestation of a malignant tumour of the maxillary antrum. Locally advanced tumours may present with anaesthesia of the skin of the cheek and upper lip, diplopia, proptosis, nasal obstruction, epistaxis, a mass in the hard palate or upper gum, or a soft tissue mass in the upper gingivobuccal sulcus. Advanced tumours may present with trismus and visible or palpable fullness of the cheek. Trismus usually is a sign of pterygoid musculature invasion. Epistaxis may be the first manifestation of tumours of the ethmoid or frontal sinus. This may be accompanied by Frontal sinus

Ethmoid air cells

Ostiomeatal complex

Uncinate process

Middle concha Inferior concha

Maxillary antrum

Fig. 3 Coronal section through maxillofacial region, showing proximity of orbit and anterior cranial fossa to nasal cavity and paranasal sinuses. Disease of the sinuses

and nasal cavity may spread directly into adjacent structures with catastrophic results

P. Delaere and J. Meulemans

26

Fig. 4  The rigid endoscope allows for detailed examination of the nasal cavity. The scope can be rotated laterally under the middle turbinate into the posterior aspect of the

middle meatus (asterisk). An excellent view of the middle turbinate, uncinate process, and surrounding mucosa can be obtained

frontal headaches or diplopia. Occasionally anosmia may be present in patients with esthesioneuroblastoma. Anaesthesia in the distribution of the fifth cranial nerve or paralysis of the third, fourth, or sixth cranial nerve may be the first manifestation of a primary tumour of the sphenoid sinus. Although sinonasal malignancy is rare, persistent nasal symptoms should always be investigated, particularly if unilateral. Tumours of the nasal cavity and paranasal sinuses are the most challenging to stage. Endoscopic evaluation of the nasal cavity is crucial in accurate clinical assessment of an intranasal lesion. Fiberoptic flexible endoscopy provides adequate visualization of the lower half of the nasal cavity. Therefore, lesions presenting in the region of the inferior turbinate, middle meatus, and the lower half of the nasal septum can be easily visualized by office endoscopy. Rigid endoscopic evaluation with telescopes generally requires adequate topical anaesthesia as well as shrinkage of the mucosal surfaces of the interior of the nasal cavity with the use of topical cocaine solution. A set of 0°, 30°, and 70° telescopes should be available for appropriate evaluation of the interior of the nasal cavity (Fig. 4). Diagnostic nasal endoscopy allows the characterization of intranasal anatomy and identification of pathology not otherwise visible by traditional diagnostic techniques, such as the use of a headlight, speculum, and mirror (Bolder and Kennedy 1992; Levine 1990).

4  Nasopharynx The nasopharynx is the portion of the pharynx bounded superiorly by the skull base and the sphenoid and laterally by the paired tori of the Eustachian tubes, with the Rosenmüller’s fossa. Anteriorly the posterior choanae form the limit of the space, and inferiorly an artificial line drawn at the level of the hard palate delimits the nasopharynx from the oropharynx. Presenting symptoms of nasopharyngeal cancer may include a neck mass, epistaxis, nasal obstruction, otalgia, decreased hearing, or cranial neuropathies. Serous otitis media may occur due to Eustachian tube obstruction. Cranial nerve VI (abducens nerve) is most frequently affected but multiple cranial nerves may be involved. Nasopharyngeal carcinoma has a tendency for early lymphatic spread. Approximately 85% of patients present with cervical adenopathies and 50% have bilateral neck involvement (Lindberg 1972). The most commonly involved regions include retropharyngeal lymph nodes (of Rouvier) (69%) and level II lymph nodes (70%) (Ho et  al. 2012). Although the lateral retropharyngeal lymph node is the first lymphatic filter, it is not palpable upon clinical examination. As a result, the common first palpable node is the jugulodigastric and/or apical node underneath the sternocleidomastoid muscle (in level II), which are in fact second echelon nodes. Nasoendoscopy

Clinical and Endoscopic Examination of the Head and Neck

27 Sphenoid sinus Torus tubarius Opening of eustachian tube Rosenmuller’s fossa

Fig. 5  Examination of nasopharynx with flexible scope

(Fig. 5) using the flexible scope gives a good view of the nasal floor, the walls of the nasopharynx, and the fossa of Rosenmüller. Nasopharyngeal tumours in any quadrant including the fossa of Rosenmüller can be seen and accurately biopsied. For the nasopharynx, also rigid 0° and 30° sinus endoscopes can be similarly used in the clinical setting. Under anaesthesia, should this be necessary, these are the scopes of choice for visual assessment and biopsy. Evidence of lower cranial nerve deficits may be apparent from palatal or glossal paralysis and atrophy. A full evaluation of the remaining cranial nerves should include assessment of the vision and microscopic examination of the tympanic membranes should be performed to exclude serous otitis media.

5  Oral Cavity The oral cavity extends from the vermilion borders of the lips to the junction of the hard and soft palates superiorly and to the line of the circumvallate papillae inferiorly. Within this area are the lips, buccal mucosa, alveolar ridges with teeth and gingiva, retromolar trigone, floor of mouth, anterior two-thirds of the tongue, and hard palate (Fig. 6). All mucosal surfaces of the mouth require thorough and systematic examination. The oral cavity is lined by a mucous membrane which is a non-keratinizing stratified squamous epithelium and is therefore pink. It contains taste buds and many minor salivary glands. All mucosal surfaces should be examined using tongue blades under optimal lighting conditions.

28 Fig. 6  Oral cavity and oropharynx. The posterior limits of the oral cavity are the anterior tonsillar pillars, the junction of the anterior two-thirds and posterior one-third of the tongue (i.e. the circumvallate papillae) and the junction of the hard and soft palate. The soft palate and the tonsil are therefore part of the oropharynx. Carcinoma of the anterior two-thirds of the tongue is the most frequent site for a mouth cancer and the lateral border (1) is the most common location. Carcinoma of the floor of the mouth most commonly occurs anteriorly either in the midline or more usually to one side of the midline (2). Carcinoma of the oropharynx most commonly occurs in the slit between tonsil and base of tongue, at the level of the anterior tonsillar pillar (3)

The clinical features of the primary tumours arising in the mucosal surface of the oral cavity are variable. The tumour may be ulcerative, exophytic or endophytic, or it may be a superficial proliferative lesion. Most patients with a mouth cancer present with a painful ulcer. Squamous carcinomas with excessive keratin production and verrucous carcinomas present as white heaped-up keratotic lesions with varying degrees of keratin debris on the surface. Bleeding from the surface of the lesion is a characteristic for malignancy and should immediately raise the suspicion for a neoplastic process. Endophytic lesions have a very small surface component but have a substantial amount of soft tissue involvement beneath the surface. Oral salivary tumours may present as a nodule, a non-ulcerative swelling, or more usually as

P. Delaere and J. Meulemans

Circumvallate papillae Foramen caecum

an ulcerative lesion. Metastatic tumours may also present as submucosal masses. Mucosal melanoma shows characteristic pigmentation. Macroscopic lesions should be evaluated for mobility and tenderness and be palpated with the gloved finger to detect submucosal spread. This is particularly important in tongue lesions extending posteriorly into the posterior third and tongue base. The distance from the tumour to the mandible and the mobility of the lesion in relation to the mandible are critical elements in determining the management of perimandibular cancers. Palpation of the neck is of course essential in the assessment of a patient with mouth cancer. The presence of neck nodal disease influences the method of treatment (therapeutic neck dissection instead of prophylactic neck dissection), and also prognosis is determined by the presence of

Clinical and Endoscopic Examination of the Head and Neck

­ etastatic nodes. Full examination of the neck m must be carried out to detect any lymph node metastases and each level must be carefully palpated, particularly the upper and middle deep cervical nodes deep to the sternocleidomastoid, from behind the patient, using the tips of the fingers. Carcinoma of the oral tongue has the greatest propensity for metastasis to the neck among all oral cancers. The primary echelon of drainage is level II but other levels may also be involved.

6  Oropharynx The oropharynx is that part of the pharynx which extends from the level of the hard palate above to the hyoid bone below. The anterior wall of the oropharynx is formed by the base or posterior third of the tongue bounded anteriorly by the v-shaped line of circumvallate papillae (Fig. 6). When present, the initial symptoms of oropharyngeal cancer are often vague and non-specific, leading to a delay in diagnosis. Consequently, the overwhelming majority of patients present with locally advanced tumours. Presenting symptoms may include sore throat, foreign-body sensation in the throat, altered voice, or referred pain to the ear that is mediated through the glossopharyngeal and vagus nerves. Over two-thirds of patients present with a neck lump. As the tumour grows and infiltrates locally, it may cause progressive impairment of tongue movement which affects speech and swallowing. Most tumours of the oropharynx can be easily seen with good lighting, but those originating in the lower part of the oropharynx and tongue base are best viewed with a fiberoptic nasopharyngeal endoscope, which has greatly enhanced the ease of examination of these tumours, particularly in assessing the lower extent of the tumour and also the superior extent if the nasopharynx is involved. The patient should be asked to protrude the tongue, to rule out injury to the hypoglossal nerve. Trismus is a sign of invasion of the masticator space. Sensory and motor function should be assessed, particularly mobility of the tongue as well as fixation.

29

As the extent of involvement is often underestimated on inspection, bimanual palpation of the tumour must be undertaken in all patients. Careful palpation should be carried out to estimate the extent of infiltration; but when this examination is limited by patient tolerance, thorough palpation under general anaesthesia is advisable. Advanced tumours that cause trismus may also be better assessed under a general anaesthesia. A detailed examination and biopsy under general anaesthesia may even be the only accurate method of assessing the extent of some oropharyngeal tumours such as those of the tongue base that may be in a submucosal location. Examination of the neck must be carried out systematically and each level must be carefully palpated to detect lymph node enlargement or deep invasion of the tumour. Nodal metastases from squamous cell carcinomas are typically hard and when small are generally mobile. As they enlarge, those in the deep cervical chain initially become attached to the structures in the carotid sheath and the overlying sternocleidomastoid muscle with limitation in vertical mobility, but later become attached to deeper structures in the prevertebral region with absolute fixation. Lymphomas on the other hand have a rubbery consistency and are generally larger and multiple with matting together of adjacent nodes. Cystic degeneration in a metastatic jugulodigastric node from a squamous carcinoma of the oropharynx, which is quite typical for HPV-mediated (p16+) oropharyngeal cancer, may have a similar presentation to a brachial cyst but the latter is a far less likely diagnosis in the older patient.

7  Larynx The larynx communicates with the oropharynx above and the trachea below. Posteriorly it is partly surrounded by the hypopharynx. It may be functionally divided into three important areas. The supraglottis contains epiglottis, aryepiglottic folds, arytenoids, and false cords and includes the

P. Delaere and J. Meulemans

30

a

Epiglottis Ventricle True vocal cord Cricoid cartilage Thyroid cartilage

b

c Epiglottis

False vocal fold True vocal fold

Apex pyriform sinus

Fig. 7  Indirect laryngoscopy with Hopkins rod telescope. (a) Sagittal view. (b) View during quiet breathing. The arytenoid cartilages (1) articulate with facets on the superior surface of the posterior arch of the cricoid cartilage (2). A small mass of cartilage, the corniculate cartilage (3), usually articulates with the apex of the arytenoid and is located within the inferomedial part of the aryepiglottic fold (4). In the midline the mucosa forms a shallow notch between the two corniculate cartilages, known as the pos-

terior commissure (5). On the lateral aspect of the corniculate cartilages, within the aryepiglottic folds, are the cuneiform cartilages (6). During laryngoscopy, the corniculate and cuneiform cartilages appear as small paired swellings in the aryepiglottic folds lying on either side of the posterior commissure. (c) View during phonation. The aryepiglottic folds (1) define the anteromedial border of the pyriform fossae (2)

laryngeal ventricle. The glottis includes the vocal cords and anterior commissure and posterior commissure. The subglottis is limited by the undersurface of the true cords to the inferior margin of the cricoid cartilage (Fig. 7a).

Patients with primary tumours of the larynx usually present with complaints of hoarseness of voice, discomfort in the throat, dysphagia, odynophagia, sensation of something stuck in ­ the throat (globus pharyngeus), occasional

Clinical and Endoscopic Examination of the Head and Neck

r­espiratory obstruction resulting in stridor, haemoptysis, or with referred pain in the ipsilateral ear. Hoarseness is an early symptom of glottic cancer but may be seen later in advanced supraglottic or subglottic tumours indicating spread to the vocal cord, arytenoid, or cricoarytenoid joint with resulting hemilaryngeal fixation. Submucosal spread within the paraglottic space can occur from these sites to produce hoarseness without mucosal irregularity. Dyspnoea and stridor occur with bulky supraglottic tumours or in the presence of vocal cord fixation. In most instances, the diagnosis is made by a thorough clinical examination, which includes rigid or flexible endoscopic examination of the larynx for adequate assessment of the surface extent of the primary tumour and mobility of the vocal cords. Examination must be carried out carefully to identify the possible spread of tumour beyond the larynx either directly or by metastasis to the regional lymph nodes. A neck mass usually indicates lymphatic metastasis but may result from direct extension of the tumour into the soft tissues of the neck. The most frequent site of secondary deposits is the ipsilateral deep cervical chain, usually in the upper/middle region (level II, III). Glottic tumours rarely metastasize, while deposits in the lymph nodes are more frequent from supraglottic lesions. Examination must include an assessment of the number, mobility, and level of the lymph nodes. Some anterior swelling of the larynx, by widening or by penetration of tumour through the cricothyroid membrane, may be felt. The use of the 70° or 90° Hopkins rod telescope (Fig.  7a) allows a high resolution and magnified view of the larynx. It allows assessment of vocal cord function, high-quality photography, and is the ideal instrument for videostroboscopy of the larynx. The clinical appearance of a normal larynx is shown in Fig. 7b. This view of the normal larynx provides adequate visualization of all the anatomic sites of the supraglottic and glottic larynx as well as the hypopharynx. The dynamic function of the larynx should also be observed and recorded by asking the patient to phonate. During phonation, the vocal cords adduct while the pyriform

31

sinuses are opening up, revealing their apices (Fig. 7c). Stroboscopy is useful for the differentiation of functional from anatomical defects (Sercarz et al. 1992) and has been employed in the early detection of glottic cancer. In the latter setting, preservation of the mucosal wave suggests that a lesion is not invasive (Zhao 1992). Technological advance is producing increasingly smaller diameter fiberoptic endoscopes for examination of the human body. The flexible nasendoscope can be used to examine the postnasal space, pharynx, and larynx, down to the level of the vocal cords. Flexible nasolaryngoscopy (Fig. 8) is generally carried out in a normal anatomical position and during normal respiration, unlike the rather distorted position achieved by indirect laryngoscopy or the use of the Hopkins rods. Additionally, flexible endoscopy can be used to directly observe the pharyngeal phase of swallowing, giving complementary information to that obtained by videofluoroscopy. Test swallows of milk or coloured food can be examined during this fiberoptic endoscopic examination of swallowing (FEES) (Logemann 1983). Nowadays, flexible nasolaryngoscopes with NBI (Narrow band imaging) have an important clinical value in detecting superficial laryngeal lesions. Using only two narrow wavebands of light, NBI highlights both the mucosal surface and the underlying intraepithelial microvasculature, which helps determining the nature of the lesion and aids in earlier detection of laryngeal (pre)malignancies when compared to conventional white light examination (Ni and Wang 2016). Direct laryngoscopy under general anaesthesia remains the “gold standard” to assess mucosal lesions of the larynx and pharynx (Phelps 1992; Parker 1992), and more often enables adequate biopsies to be sampled than with flexible techniques (Ritchie et al. 1993). If a tumour is detected, its limits in all directions should be determined by both sight and palpation. The introduction of the operating microscope has facilitated detailed examination of the larynx (Kleinsasser 1965) (Fig. 9). Use of various telescopes (0°, 30°, 70°) provides an excellent and detailed view of the lesion.

32

P. Delaere and J. Meulemans

Fig. 8  Flexible laryngoscopy. Fiberoptic laryngeal nasendoscopy provides a clear image of the larynx, laryngopharynx, and base of tongue

Fig. 9  The arrangement for stable microlaryngoscopy. After placement of the laryngoscope, the laryngostat is attached and the tension tightened until the view is just adequate. The microscope is then brought into position and focused

Clinical examination is limited by the fact that certain areas of the larynx are inaccessible to both visualization and palpation; nevertheless,

involvement of these structures has an important bearing on staging as well as on management. Information from radiological imaging and

Clinical and Endoscopic Examination of the Head and Neck

o­ perative endoscopy must be utilized in conjunction with physical findings to obtain an accurate ­pretreatment TNM staging record. Particularly supraglottic tumours are frequently understaged because the pre-epiglottic and paraglottic spaces cannot be assessed clinically.

8  Hypopharynx and Cervical Oesophagus The hypopharynx links the oropharynx superiorly to the larynx and oesophagus below. Its boundaries are roughly the hyoid and valleculae above and the cricoid below. Common sites for squamous cell cancers are the pyriform sinuses, the posterior pharyngeal wall, and the postcricoid space. Patients with primary tumours of the hypopharynx usually present with the complaints of discomfort in the throat, dysphagia, odynophagia, sensation of something stuck in the throat (globus pharyngeus), referred pain in the ipsilateral ear, haemoptysis, hoarseness of voice, or shortness of breath. In most instances, diagnosis is made by a thorough clinical examination including rigid telescopic or fiberoptic nasolaryngopharyngoscopic examination of the hypopharynx and larynx. Again, fiberoptic examination of the hypopharynx using NBI is a valuable technique for the detection of early hyopharyngeal tumours and helps in the assessment of its mucosal margins (Ni and Wang 2016). While clinical examination permits the diagnosis of a primary tumour of the hypopharynx, direct laryngoscopy and oesphagoscopy under general anaesthesia are essential for accurate assessment of the tumour extent and to obtain a biopsy for histologic diagnosis. The important features to be assessed during endoscopic examination under anaesthesia include the site of origin of the primary tumour, and its local extension to the other sites within the hypopharynx and adjacent regions. In patients with a malignant tumour of the upper respiratory or upper digestive tract, it is advisable to perform flexible oesophagogastroduodenoscopy; the detection rate of a synchronous primary tumour in the upper digestive tract is about 3–13% (Levine and Nielson 1992).

33

9  Salivary Glands The parotid glands are located in close proximity to the cartilage of the external auditory canal. Anteriorly the gland abuts both the lateral and posterior border of the ramus of the mandible and the overlying masseter muscle, while inferiorly it rests medially on the posterior belly of the digastric muscle, as well as the sternocleidomastoid muscle laterally. Medially the parotid is adjacent to the parapharyngeal space, while superiorly it reaches the arch of the zygoma. The facial nerve courses through the parotid gland. The parotid gland is arbitrarily divided into a ‘superficial’ and ‘deep’ lobe by the plane of the facial nerve. Numerous lymph nodes are localized within, and adjacent to, the capsule of the parotid gland, serving as the first echelon drainage for the temporal scalp, portions of the cheek, the pinna, and the external auditory canal. For this reason, the parotid gland may harbour metastatic cutaneous malignancy from these sites. The submandibular glands are located in the anterior triangle of the neck, and are bounded superiorly and laterally by the body of the mandible. The mylohyoid muscle is located anterior to the gland, while the hyoglossus muscle lies medial to the gland. The submandibular (Wharton’s) duct exits the gland medial to the mylohyoid muscle, and then courses anteriorly and superiorly into the anterior floor of mouth (Fig. 10). Located beneath the mucosa of the floor of the mouth, the small sublingual glands drain directly into the oral cavity through numerous small ducts. The majority of neoplastic lesions of salivary glands appear as a lump without other symptoms. Swellings in the retromandibular sulcus, the immediate preauricular region, and over the masseter are, in most cases, of parotid gland origin. Although about 10% of parotid gland tumours arise medial to the plane of the facial nerve in the deep ‘lobe’ of the gland, more than three-fourths of these deep lobe tumours will present as a typical parotid mass. In the parotid gland, pleomorphic adenomas present as round, firm, reasonable well-­ demarcated tumours, with a tendency to

P. Delaere and J. Meulemans

34

Stenon’s duct

Parotid gland

Sublingual gland

Submandibular gland

Wharton’s duct

Fig. 10  Anatomic relations of the parotid, submandibular, and sublingual salivary gland

n­ odularity as they grow. Their site of election is between the ascending ramus of the mandible anteriorly, and the mastoid process and sternocleidomastoid posteriorly, usually in the tail of the gland. Occasionally they arise in the immediate preauricular region, where they tend to be small. Wharton’s tumours lie almost invariably in the lower pole of the gland, are ovoid in shape, and vary in consistency between soft and firm, depending on whether or not they have been exposed to previous inflammation; these tumours can occur bilaterally. Weakness or paralysis of the facial nerve in a previously untreated patient usually indicates that a tumour is malignant (Spiro et  al. 1975; Borthune et al. 1986). Careful assessment should be made of the facial nerve and the nerves traversing the nearby carotid space (cranial nerve IX and XII) if a deep lobe or parapharyngeal space tumour is suspected.

It is often difficult to distinguish between a tumour arising within the submandibular gland or an enlarged node close to the gland or on its outer surface. Bimanual palpation is essential to differentiate between the two, since a node lying on the outer surface of the salivary gland is unlikely to be palpated by a finger in the mouth, whereas a tumour of the gland itself is more readily compressible bimanually. Pleomorphic adenomas of the submandibular gland are usually large, quite hard, and nodular, but may be confused with a slowly growing malignancy such as an adenoid cystic carcinoma. Submandibular gland neoplasms also need evaluation of the lingual and hypoglossal nerves. The assessment of intraoral minor salivary gland neoplasms depends on the location of the tumour. Palatal lesions are the most common, usually giving the appearance of being fixed, whether they are benign or malignant, because of the tight

Clinical and Endoscopic Examination of the Head and Neck

adherence of the mucous membrane to bone. Tumours of the hard or soft palate are often fusiform, firm to hard in consistency, and nodular. Again, the distinction between mixed tumour and adenoid cystic carcinoma may be difficult to make. A salivary gland tumour arising from the deep lobe of the parotid gland, or from a minor salivary gland in the parapharyngeal space, may cause secondary displacement of the palatotonsillar region. Swelling detectable both in the pharynx and parotid region indicates a very bulky tumour originating from within the deep lobe of the parotid gland. This parotid swelling can be visible externally, but the technique of bimanual palpation will elicit the characteristic sign of ballottement between the examining fingers, typical of masses occupying such a wide area. The absence of both a visible swelling in the parotid gland and ballottement suggests an origin exclusively in the parapharyngeal space.

10  Thyroid Gland The thyroid gland lies within the pretracheal fascia in the front of the neck, and consists of two symmetrical lobes united in the midline by an isthmus that overlies the second to fourth tracheal rings (Fig. 1). There is often a pyramidal lobe, which may extend as high as the top of the thyroid cartilage. The incidence of palpable thyroid nodules is estimated at only 4–7% of the general adult population. They occur more frequently in women and they are increasing with age (Mazzaferri et al. 1988). Slightly less than 5% of thyroid nodules are found to be malignant (Gharib and Goellner 1993). Most patients with differentiated carcinoma present with a palpable nodule in the thyroid gland of varying size, consistency, and local extent. The primary tumour may present as a solitary, well-defined, intrathyroidal discrete palpable nodule or it may manifest with diffuse involvement of the thyroid gland with or without extrathyroid extension and fixation to the structures in the central compartment of the neck, or it may present as multiple palpable nodules. The most common location of palpable metastatic lymph nodes from thyroid cancer is at levels III,

35

IV, or V in the lateral neck. Procedures commonly used for the initial evaluation of thyroid nodules are ultrasound, radionuclide imaging, and fine-needle aspiration biopsy (FNAB). Anaplastic carcinoma of the thyroid gland usually manifests in the older population with a very short history of a rapidly enlarging thyroid mass. Physical examination reveals a diffusely enlarged firm to hard ill-defined thyroid mass with significant extrathyroid extension to adjacent soft tissues. The mass appears fixed and inseparable from the laryngotracheal-­oesophageal complex. Palpation of the thyroid gland and the neck nodes needs to be supplemented with rigid or flexible endoscopic examination of the larynx with special focus on the assessment of vocal fold mobility. When reduced or absent unilateral vocal fold mobility is observed, underlying malignancy in the thyroid gland with invasion of the recurrent nerve needs to be suspected.

11  Role of Imaging Studies The clinical evaluation allows appreciating the mucosal layer of the head and neck region quite well. However, the deep extent of potentially infiltrating lesions can only be judged indirectly. Some regions, such as the base of the skull, pterygopalatine and infratemporal fossa, orbits and brain are beyond clinical evaluation, but critical management decisions have to be made based on the involvement of these structures; imaging findings are of the utmost importance in such cases. Perineural and/or perivascular spread, eventually leading to tumour progression or recurrences at distance from the primary tumour, can often only be detected by imaging. Metastatic adenopathies can be identified, sometimes still in a subclinical stage or at places not accessible for clinical examination, such as the retropharyngeal or paratracheal lymph nodes. In addition, information on extranodal tumour spread and the relation to critical structures such as the carotid arteries is necessary for determining the optimal patient management, and can be deduced from imaging studies. Imaging is needed in submucosal lesions, covered by an intact mucosa. The origin and extent

36

of such lesions is often difficult to determine based on the clinical evaluation alone. Imaging may provide important clues to the diagnosis, as representative biopsies may be difficult to obtain in deep-seated lesions. All these findings can profoundly influence the staging and management of the patient with head and neck cancer. Finally, imaging may be used to monitor tumour response and to try to detect recurrent or persistent disease before it becomes clinically evident, possibly with a better chance for successful salvage. The single most important factor in the optimal use of all this information is the mutual co-­ operation between the radiologist and the physicians in charge of patient care.

References Amin M, Edge S, Greene F, et al (2017) AJCC cancer staging manual. 8th ed. Springer, New York Brierley J, Gospodarowicz M, Wittekind C (2017) UICC TNM classification of malignant tumours. 8th ed. Wiley, Chichester Bolder WE, Kennedy DW (1992) Nasal endoscopy in the outpatient clinic. Otolaryngol Clin North Am 25:791–802 Borthune A, Kaalhus O, Vermund H (1986) Salivary gland malignant neoplasms: treatment and prognosis. Int J Radiat Oncol Biol Phys 12:747–754 Gharib H, Goellner JR (1993) Fine needle aspiration biopsy of the thyroid: an appraisal. Ann Intern Med 118:282–289 Ho FCH, Tham IWK, Earnest A, Lee KM, Lu JJ (2012) Patterns of regional lymph node metastasis of nasopharyngeal carcinoma: a meta-analysis of clinical evidence. BMC Cancer 12:98 Huang SH, O’Sullivan B (2017) Overview of the 8th edition TNM classification for head and neck cancer. Curr Treat Options Oncol 18:40 Kleinsasser O (1965) Weitere technische entwicklung und erste ergebnisse der ‘endolaryngealen microchirurgie’. Z Laryngol Rhinol Otol 44:711–727 Levine HL (1990) The office diagnosis of nasal and sinus disorders using rigid nasal endoscopy. Otolaryngol Head Neck Surg 102:370–373 Levine B, Nielson EW (1992) The justifications and controversies of panendoscopy—a review. Ear Nose Throat J 71:335–343

P. Delaere and J. Meulemans Li-Jen L et  al (2012) Detection of cervical lymph node metastasis in head and neck cancer patients with clinically N0 neck – a meta-analysis comparing different imaging modalities. BMC Cancer 12:236 Lindberg RD (1972) Distribution of cervical lymph node metatstases from squamous cell carcinoma of the upper respiratory and digestive tracts. Cancer 29:1446–1450 Logemann JA (1983) Evaluation and treatment of swallowing disorders. College Hill Press, San Diego Mazzaferri EL, de los Santos ET, Rofagha-Keyhani S (1988) Solitary thyroid nodule: diagnosis and management. Med Clin North Am 72:1177–1211 Merritt RM, Williams MF, James TH et  al (1997) Detection of cervical metastasis. A meta-analysis comparing computed tomography with physical examination. Arch Otolaryngol Head Neck Surg 123:149–152 Ni XG, Wang GQ (2016) The role of narrow band imaging in head and neck cancers. Curr Oncol Rep 18:10 Parker R (1992) Laryngoscopy, microlaryngoscopy and laser surgery. In: McGregor IA, Howard DJ (eds) Rob and Smith’s operative surgery: head and neck, part 2, 4th edn. Butterworth, Oxford, pp 451–463 Phelps PD (1992) Carcinoma of the larynx-the role of imaging in staging and pre-treatment assessments. Clin Radiol 46:77–83 Ritchie AJ, McGuigan J, Stenvenson HM et  al (1993) Diagnostic rigid and flexible oesophagoscopy in carcinoma of the oesophagus: a comparison. Thorax 48:115–118 Sercarz JA, Berke GS, Ming Y et  al (1992) Videostroboscopy of human vocal fold paralysis. Ann Otol Rhinol Laryngol 101:567–577 Shah JP (1990) Patterns of nodal metastases from squamous carcinomas of the upper aerodigestive tract. Am J Surg 160:405–409 Spiro RH, Huvos AW, Strong EW (1975) Cancer of the parotid gland: a clinicopathologic study of 288 primary cases. Am J Surg 130:452–459 Stell PM, Maran AGD (eds) (1972) Pre-operative considerations. In: Head and neck surgery. London: William Heinnemann Medical, p 6 van den Brekel MW, Castelijns JA, Stel HV, Golding RP, Meyer CJ, Snow GB (1993) Modern imaging techniques and ultrasound-guided aspiration cytology for the assessment of neck node metastases: a prospective comparative study. Eur Arch Otorhinolaryngol 250:11–17 Watkinson JC, Johnston D, James D et al (1990) The reliability of palpation in the assessment of tumours. Clin Otolaryngol 5:405–410 Zhao R (1992) Diagnostic value of stroboscopy in early glottic carcinoma. Chung Rua Ehr Pi Yan Hou Ko Tsa Chih 27:175–176

Imaging Techniques Robert Hermans, Frederik De Keyzer, Vincent Vandecaveye, and Laurens Carp

Contents

Abstract

1

Introduction 

 37

2

Plain Radiography 

 38

3

Ultrasonography 

 38

 omputed Tomography and Magnetic C Resonance Imaging  4.1  Computed Tomography  4.2  Magnetic Resonance Imaging  4

 38  41  47

Positron Emission Tomography  Physical Aspects  Radiopharmaceuticals  Technical Aspects of FDG-PET and Integrated FDG-PET/CT in Head and Neck Cancer  5.4  PET/MRI 

 59  60

References 

 62

5 5.1  5.2  5.3 

 56  56  57

Various imaging techniques are used to investigate the presence and extent of head and neck neoplasms, including ultrasound, computed tomography, magnetic resonance imaging, and nuclear imaging techniques. To obtain the most optimal results, close attention should be paid to a correct technical execution of the imaging study. This chapter provides information on the relative advantages and disadvantages of each of the available imaging techniques, as well as on patient preparation, contrast agent or tracer injection, data acquisition, and image reconstruction, reformatting and display.

1  Introduction

R. Hermans (*) · F. De Keyzer · V. Vandecaveye Department of Radiology, University Hospitals, KU Leuven, Leuven, Belgium e-mail: [email protected] L. Carp Department of Nuclear Medicine, University Hospital Antwerp, Edegem, Belgium

Various imaging techniques are used in the evaluation of patients with head and neck cancer, before, during, and after treatment. Each of these imaging techniques has its own advantages and disadvantages. Many head and neck neoplasms arise from the mucosal lining; when a patient is referred for imaging, the histological diagnosis often was already established by biopsy. Therefore, imaging should primarily supply information on the submucosal extension of the primary tumor, including its relation to surrounding structures, as well as on the presence of regional and/or distant metastasis, or a second primary tumor.

Med Radiol Diagn Imaging (2020) https://doi.org/10.1007/174_2020_223, © Springer Nature Switzerland AG Published Online: 23 April 2020

37

38

The purpose of this chapter is to describe the various techniques available for imaging the head and neck cancer patient, and to provide general rules for their use. Specialized imaging applications are described in the following chapters where appropriate.

2  Plain Radiography In the past, a variety of conventional methods were applied to stage head and neck cancer, including soft tissue views of the neck, plain films of the facial skeleton, xeroradiography, plain film tomography, laryngography, and barium swallow. The value of these studies to stage head and neck cancer is very limited; these techniques are now replaced by cross-sectioning imaging modalities. Barium swallow remains an indispensable method in the early phase after pharyngeal surgery, to rule out or confirm the presence of fistulae. This technique is also essential in the evaluation of functional disorders (such as bolus retention, delayed passage, and aspiration) after surgery or radiotherapy.

3  Ultrasonography Ultrasonography is a widely used technique for the evaluation of the thyroid gland, neck lymph nodes, and salivary glands, as it offers visualization of these structures with high spatial resolution, at a low cost and without using ionizing radiation. Ultrasonography in combination with fine needle aspiration cytology (FNAC) is the most accurate method for neck nodal staging in most head and neck cancers (van den Brekel et  al. 1991). However, execution of this procedure is time consuming, and the obtained results are somewhat operator dependent (Takes et al. 1996). In a multicenter study where both computed tomography and ultrasound of the neck were applied for staging of head and neck cancer, the addition of ultrasound-guided FNAC did not provide significant additional value (Takes et  al. 1998). Nevertheless, ultrasound-guided FNAC

R. Hermans et al.

can be a very helpful problem-solving tool when a doubtful lymph node is visualized on CT or MRI, especially when localized outside of the electively treated region and therefore potentially affecting patient management.

4  Computed Tomography and Magnetic Resonance Imaging In most patients computed tomography (CT) or magnetic resonance imaging (MRI) is performed for pretherapeutic staging of a head and neck malignancy. Both techniques can supply the information needed by the clinician for adequate treatment planning. A common question is which of these techniques should be used in a particular patient. The most widely used technique is CT, as it has a number of important advantages over MRI: –– Wide availability. –– Relative low cost. –– Easy to execute, and this in a reproducible way. –– Short examination time, resulting in less image quality degradation caused by motion, such as swallowing and respiration. –– Superior bone detail. –– High-quality multiplane imaging. –– Easy extend of the study into the upper thoracic cavity or intracranial cavity, if needed. –– Easier interpretation, especially regarding nodal involvement (Curtin et al. 1998). However, CT also has a number of disadvantages compared to MRI: –– Relative low soft tissue contrast resolution. –– Administration of iodinated contrast agent is necessary. –– Severe image quality degradation by dental fillings or other metallic foreign objects (Fig. 1). –– Radiation exposure. The advantages of MRI over CT in the evaluation of head and neck cancer are its superior soft tissue contrast resolution, and the absence of

Imaging Techniques

a

Fig. 1  Patient suffering left-sided oral tongue cancer. (a) Axial contrast-enhanced CT image. As the image quality is severely hampered by artifacts arising from dental fillings, the primary tumor is not visible. A complementary

a

39

b

MR study was advised. (b) Gadolinium-enhanced T1-weighted image, not affected by the presence of dental fillings, clearly shows the primary tumor (arrowhead)

b

Fig. 2  Patient referred for MR study of the maxillofacial region and skull base, because of unilateral facial pain. (a) Initial MR study was nonconclusive, as artifacts caused by

fixed orthodontic material severely degrade image quality. (b) After removal of the orthodontic material by the dentist, optimal image quality was achieved

radiation exposure. Overall, the image quality is not or less hampered by the presence of dental fillings than in CT, but also MRI studies may be severely jeopardized by metallic implants

(Fig.  2). The disadvantages of MRI are mainly related to the long acquisition time, making the technique sensible to motion artifacts which cause a nondiagnostic study (Fig.  3). It is also

R. Hermans et al.

40

a

b

c

Fig. 3  Patient suffering tongue base cancer. (a, b) Axial and sagittal reformatted contrast-enhanced MDCT images clearly show tumor extent into tongue base and involvement of free epiglottic rim (arrowheads), as well as the relationship of the tumor to the pre-epiglottic space (aster-

isk). (c) For study purposes, a MR study was performed in this patient 1 day later. Because of motion artifacts, the tumor is not confidently discernible. This MR study was considered nondiagnostic

technically more challenging with MRI to properly stage both primary tumor and neck nodal disease in a single study. The lower availability of MRI, resulting in a longer waiting list, and its

higher cost should also be taken into consideration. In many institutions, CT is the preferred imaging method for evaluation of laryngeal and

Imaging Techniques

hypopharyngeal cancer, as well as of oral cavity and oropharyngeal cancer. These cancer sites constitute about 80–85% of all head and neck malignancies (excluding skin cancer and lymphoma) in Europe and the USA. In most cases, a dedicated CT study will provide all answers needed by the clinician; in such a setting, MRI is used as complementary tool to solve remaining questions. Because of its higher contrast resolution, MRI is the preferred imaging method in less common head and neck malignancies, such as nasopharyngeal cancer and sinonasal cancer. MRI is also preferred as the primary imaging tool in salivary gland cancer.

4.1  Computed Tomography CT remains one of the staples of head and neck cancer imaging. While it is not possible to define the ideal imaging protocol, as available equipment varies, the minimal requirements for an optimal diagnostic study will be outlined.

41

For all practical purposes, a single bolus technique with an injection rate of 1–2 cc/s is appropriate on modern CT machines (Keberle et  al. 2002). A total amount of 100 ml is sufficient. It is essential to wait long enough before starting the acquisition, as the contrast agents need some time to diffuse in the normal and pathologic soft tissues. Using a MDCT machine, allowing a rapid entire neck examination without gantry angulation, the scan should be started only after injection of the entire contrast volume. A subsequent saline injection at the same injection rate is recommended. The contrast agent injection protocol for evaluation of the head neck, as currently used in Leuven, is: 1.5 cc/s contrast agent up to 100 ml, followed by 30 ml saline at the same rate; image acquisition starts 80 s after initiation of the injection.

4.1.3  D  ata Acquisition and Image Reconstruction

4.1.3.1  General Comments On a lateral scout view, the area of interest is indicated. For a routine head and neck imaging 4.1.1  Patient Positioning study, images are acquired from the top of the The images are obtained with the patient supine sphenoid sinus to the lower border of the sternoand during quiet respiration. The neck should be clavicular joints. It makes sense to scan from crain slight extension. The head is aligned in the nial to caudal: this allows the contrast medium cephalocaudal axis in order to make it possible to concentration in the subclavian vein, at the side compare symmetric structures. Malposition may of injection, to drop to a similar or only slightly result in an appearance that simulates disease. higher level compared to other neck vessels, Every effort should be made to make the patient reducing artifacts at the level of the thoracic inlet. feel comfortable; this will help the patient dropWhen performing a routine study of the face, ping the shoulders to a position as low as possible sinonasal region, or skull base, images are (Wirth et  al. 2006). Correct patient positioning acquired from the top of the frontal sinus to the also helps to reduce radiation dose. submental region. The field of view (FOV) must be as small as possible, to optimize spatial resolution. The rec4.1.2  Contrast Agent Injection While evaluating a patient suffering head and ommended FOV for neck studies varies between neck cancer, a proper injection method of iodin- 16 and 22  cm, depending on the size of the ated contrast agent is crucial to obtain state-of-­ patient. The selected FOV also depends on the the-art CT images. Optimal tissue enhancement, type of pathology: in a study performed for squaallowing correct discrimination of tumoral from mous cell cancer, the posterior part of the perivernormal tissue, and a high neck vessel density tebral space not necessarily needs to be included must be realized at the same time. Several in the FOV as it is unlikely to encounter patholcontrast agent injection protocols have been ogy in that region; however, for example in skin described, some of them being fairly complicated. cancer and lymphoma, this part of the neck

42

should also be visualized, as (sub)occipital adenopathies may be present. The optimal display slice thickness for evaluation of neck structures is 2–3  mm; adjacent slices should be obtained. Somewhat thinner slices (1–2 mm) are apt for the evaluation of the facial bones, sinonasal cavities, and orbits. In laryngeal and hypopharyngeal neoplasms, it is useful to reconstruct an additional series of images coned down to the laryngohypopharyngeal region, with a FOV of about 10  cm and a slice thickness of 1–2 mm. Also the evaluation of the temporal bone requires a coned down FOV (about 8  cm), and a thin slice thickness of 0.4–1 mm. Image reconstruction is always done in a soft tissue algorithm. Additional images, reconstructed in a high-resolution (bone detail) algorithm, are always generated in sinonasal cavity, skull base, and temporal bone studies. In patients suffering neoplastic disease, the lower slices including the upper part of the lungs should also be reviewed in lung window, as unknown metastatic disease or second primary tumors may then become visible. 4.1.3.2  Dose Reduction Apart from correct patient positioning, avoiding over ranging (i.e., scanning too cranial and/or too caudal) is important to limit patient radiation exposure. Several technical features allow to substantially reduce patient dose. Tube current modulation is a very effective tool for reducing patient radiation exposure, as head and neck CT ranges from high attenuation areas, such as the skull base and shoulder regions, to low attenuation areas in the neck. Further dose reduction is possible by using iterative reconstruction and lowering kV.  Lowering kV also increases iodine contrast, allowing to reduce the amount of contrast agent needed to obtain high-quality images (Yabuuchi et al. 2018). 4.1.3.3  Multidetector Spiral CT State-of-the-art CT of the head and neck requires the use of MDCT. The rapid acquisition results in a volumetric data set, reconstructed to a stack of

R. Hermans et al.

Fig. 4  Routine head and neck CT study. Midline sagitally reformatted image from native axial MDCT images. The data acquisition extended from just above the sphenoid sinus to the thoracic inlet. From the native images, two sets of axial images are routinely reformatted for display: the first set (1), parallel to the hard palate, from the skull base to the lower margin of the mandible; the second set (2), parallel to the vocal cords (or C4–C5/C5–C6 intervertebral space), from the oral cavity to the thoracic cavity. The two image sets should be overlapping

thin and overlapping native images; this reduces partial volume averaging and motion artifacts. Furthermore, full advantage of the injected contrast agent is accomplished by optimal timing between injection and image acquisition. Although the large amount of axial native images can rapidly be reviewed on a PACS monitor, the signal/noise level of these images is relatively low. Moreover, these images may not show the anatomic structures of interest with the most optimal angulation (e.g., to judge the vocal cords). Therefore, new sets of images are reformatted from these native images. These images are routinely reformatted in the axial plane: for neck studies, adjacent 2–3 mm thick images are reformatted parallel to the hard palate from the skull base to the oral cavity, and parallel to the vocal cords from the oral cavity to the thoracic inlet (Fig.  4). Also coronal and sagittal images

Imaging Techniques

a

43

b

Fig. 5 (a) Axial MDCT image (3 mm thick) shows lymph node on right side, appearing centrally hypodense (arrowhead): central necrosis or partial volume averaging of fatty nodal hilum? (b) Additionally, a thinner (1.5  mm) reformatting was made through this lymph node in the

coronal plane. On this section, the hypodense nodal region shows fat density and communicates with the outer nodal border: fatty metaplasia in nodal hilum. Normal lymph node

are routinely reformatted from the native axial images. Reformatting in other planes and/or with a thinner slice thickness is done according to the organ of interest (see above) and/or the findings on the axial images (Fig. 5). The data acquisition with MDCT is usually done with zero gantry tilt. However, in some patients, this causes problems at the level of the oral cavity when dental fillings are present. Also, in patients with short necks or a high position of the shoulders, the image quality may be suboptimal at the level of the larynx due to artifacts arising from the shoulder girdle. To avoid these problems, some head and neck radiologists use gantry tilting in MDCT, although this makes it impossible to obtain reformatted images in the coronal or sagittal plane at the level of the oral cavity. An alternative is to perform a complete head and neck study without gantry tilt, and acquire additional images at the level of the oral cavity and oropharynx, if dental filling artifacts are present, with a tilted gantry (Fig.  6). Yet another solution is to obtain additional images with the mouth widely opened; this may bring the pathology out of the dental filling artifacts (Fig. 7) (Henrot et al. 2003).

More technical details on the MDCT parameters for head and neck cancer, currently used in the University Hospitals of Leuven, are summarized in Tables 1 and 2. 4.1.3.4  Dual Energy CT Although currently not yet widely used, the application of dual energy CT (DECT) may offer several advantages over single energy MDCT in the evaluation of head and neck cancer patients. The better contrast between tumor and normal tissue helps in detecting subtle neoplastic disease and makes the tumor borders more conspicuous. The detection of pathological lymph nodes may be improved, as well as early cartilage and bone invasion. Furthermore, reduction of metal artifacts (e.g., from dental restoration material) and lowering patient radiation exposure seem to be feasible (Roele et al. 2017; Forghani et al. 2017).

4.1.4  Dynamic Maneuvers The data acquisition is routinely performed while the patient continues breathing. Dynamic maneuvers during scanning can improve the visualization of particular anatomic structures. During prolonged phonation of [i], arytenoid mobility

R. Hermans et al.

44

a

c

b

d

Fig. 6  Axial MDCT image at the level of oral cavity and oropharynx, obtained with a zero degree gantry tilt (a), is severely degraded by artifacts arising from dental fillings. By obtaining additional images with angulation of the

gantry around the dental fillings (b), optimal image quality could be achieved at this level (c, d). (Courtesy of Ilona Schmalfuss, MD, Gainesville, FL, USA)

can be judged and a better visualization of the laryngeal ventricle can be achieved; the slight distention of the pyriform sinuses may also allow better delineation of the aryepiglottic folds (Lell et  al. 2004). A modified Valsalva maneuver

(blowing air against closed lips, puffing out the cheeks) produces substantial dilatation of the hypopharynx, allowing better visualization of the pyriform sinuses, including the postcricoid region (Robert et al. 1993) (Fig. 8). This modified

Imaging Techniques

a

45

b

c d

Fig. 7  Patient suffering cancer of the lateral tongue edge. (a, b) Sagittal and axial reformatted CT image, obtained after a standard MDCT acquisition with closed mouth (a). This study did not show the extent of the primary tumor because of dental filling artifacts (b). Tilting the gantry would not have solved the problem in this patient, as the

dental fillings are close to the tumor. (c, d) Because of a contraindication for MRI, an additional CT study was performed with the mouth of the patient widely opened (c). This brings the region of the tongue tumor (d, arrowheads) out of the dental filling artifacts

Valsalva maneuver may also be of use in the evaluation of oral cavity tumors, as the inner cheek walls and gingivobuccal sulci become better visible. The success rate of these dynamic maneuvers is variable, especially when an incremental CT

technique is used, and is strongly depending on the cooperation of the patient. Dynamic maneuvers are mainly helpful in showing superficial tumor spread, while the purpose of imaging is describing deep tumor extent. Also, abnormal mobility of the vocal cord is

R. Hermans et al.

46 Table 1  MDCT data acquisition and native image reconstruction parameters Necka Faceb Sinonasal cavitiesb

Temporal boneb Skull baseb

128-row 128 × 0.6 mm 0.8 1 s 120 230c 1 mm 0.5 mm 16 × 0.3 mm 1 1 s 120 250 0.4 mm 0.2 mm

Collimation Feed/pitch Rotation time kV mAseff Sliceeff Slice interval Collimation Feed/pitch Rotation time kV mAseff Sliceeff Slice interval

192-row 192 × 0.6 mm 0.8 1 120 190 1 0.5 64 × 0.3 mm 0.85 1 s 110 150 0.4 mm 0.2 mm

mAseff effective mAs, sliceeff effective slice thickness a Soft tissue algorithm b Both soft tissue and bone detail algorithm in tumoral pathology c Effectively used mAs may be lower (determined by automatic exposure control system)

Table 2  MDCT image reformatting Neck Face, sinonasal cavities, skull base (soft tissue detail) Temporal bone, skull base (bone detail)

Slice thickness 3 mm 2 mm

Slice interval Image plane 3 mm Axial + coronal + sagittal 2 mm Axial + coronal + sagittal

0.4 mm

0.4 mm

a

Fig. 8  Contrast-enhanced single-slice spiral CT images in a patient suffering cancer of the pyriform sinus. (a) Axial image during quiet breathing shows subtle soft tissue thickening in the apex of the right pyriform sinus (arrow); some evidence of subtle infiltration or displace-

Axial + coronal (sagittal if needed)

b

ment of the paraglottic space fat is present (arrowhead). (b) Axial image obtained during modified Valsalva maneuver. The right pyriform sinus expands somewhat less than the opposite one; the mucosal irregularity produced by the cancer is now better visible (arrowheads)

Imaging Techniques

a

47

b

Fig. 9 (a) Axial contrast-enhanced MDCT image. Large floor of the mouth cancer, massively invading the right side of the mandible (arrows). Several small but inhomog-

enously enhancing adenopathies are visible on both sides of the neck (arrowheads). (b) Extent of mandibular bone destruction can easily be appreciated on 3D reformatting

more accurately seen during clinical examination than on dynamic imaging studies. Therefore, the added value of acquiring images during a dynamic maneuver in staging head and neck neoplasms is, on average, limited.

4.2  Magnetic Resonance Imaging

4.1.5  Three-Dimensional Image Reformatting Three-dimensional (3D) display of the data set is most often done to evaluate the bony structures of the maxillofacial skeleton in congenital abnormalities or traumatic lesions. Meaningful 3D display of the often subtle osteolytic changes seen in head and neck malignancies is rarely possible. However, in some cases of extensive bone destruction, 3D displays are helpful for the surgeon in planning bone resection (Fig. 9). Virtual endoscopy of the larynx and hypopharynx has been studied; otolaryngologists rank such 3D images as more beneficial than radiologists, usually in bulky masses that precluded definitive direct endoscopic evaluation (Silverman et al. 1995). This technique does not show the adjacent soft tissues, and its clinical role is not exactly defined (Magnano et al. 2005).

MRI of the head and neck can be performed on all clinically available field strengths. However, at comparable measuring time the high-field (≥3 T) machines provide a better signal-to-noise ratio and/or a higher spatial resolution. Therefore, it is currently the consensus that 3 T is preferred for all clinical head and neck examinations, unless the patient has metallic implants near the imaging site. In the latter case, the susceptibility artifacts that are more pronounced on higher field strengths may obscure the pathology.

4.2.1  Patient Positioning Similar to CT, the image acquisition is performed with the patient in supine position, and during quiet respiration. The head and neck should be aligned and symmetrically positioned. Every effort should be made to make the patient feel as comfortable as possible. The patient should be instructed not to move during the examination, and to try not to cough during the image acquisition. The patient should not be prohibited to swallow, as this is hardly feasible in clinical

48

practice as the imaging sequences take several minutes each. Slight lateral head fixation is allowed, but should not be too rigid to avoid causing patient discomfort.

R. Hermans et al.

Conversely, the high signal intensity of fat and bone marrow on a plain T1-weighted SE or TSE sequence is often very helpful to determine tumor extent, as it contrasts clearly with the low signal intensity of most tumors (Widmann et al. 2017). 4.2.2  Coils Repetition of this sequence after injection of The choice of the receiver coil is dictated by the gadolinium-­DTPA, and comparison with the pre-­ localization of the disease process. If the tumor is injection sequence, allows to determine the areas localized in the oral cavity or infrahyoid part of of contrast enhancement and to differentiate the neck, the neck coil should be used. When the these areas from fat. patient suffers neoplastic disease at the level of A fat-saturated T1-weighted SE sequence the skull base, sinonasal cavities, face, parotid after injection of gadolinium-DTPA may be helpglands, or nasopharynx, the head coil should be ful, as the contrast between enhancing tissue and selected. A disadvantage of using a single receive fat is increased, but at the cost of some more coil however is the inability to cover the entire ­susceptibility artifacts and a substantially longer neck for evaluation of nodal stations. acquisition time. This may be of particular importance in nasoDepending on the investigated region, a slice pharyngeal, tongue, and oropharyngeal cancer, thickness of 3–4  mm is optimal, with an interwhich are frequently associated with adenopa- slice gap of 0–50%. The FOV is similar to what is thies throughout the neck. Modern machines usu- described above for CT.  The imaging matrix ally allow imaging the entire head and neck with should be at least 256  ×  256, but more often a an integrated head-neck coil of 12 or more coil base matrix of 384 or 512 is advocated, espeelements. Some older machines still have a cially for lesions in and around the skull base and ­combination of a standard head coil and a dedi- sinonasal cavities. cated neck coil. These setups allow for adequate The plane of section is chosen according to the anatomical imaging, but are usually not advo- localization of the disease process. For most neck cated for functional imaging. lesions, it is appropriate to start with a T2- and T1-weighted sequence in the axial plane, and to 4.2.3  Standard Sequences continue with a gadolinium-enhanced axial, corAfter obtaining scout images, a standard exami- onal, and sagittal T1-weighted sequence. In gennation of the head and neck should start with a eral, the axial plane should be, similar as for CT, T2-weighted turbo spin echo (TSE) sequence. parallel with the hard palate when dealing with Compared to a conventional T2-weighted spin suprahyoid pathology, and parallel with the vocal echo (SE) sequence, a TSE sequence takes less cords when dealing with infrahyoid pathology. In time to perform, reducing motion artifacts and naso-ethmoidal neoplasms, it may be more useful improving image quality. However, the high sig- to start the study with a coronal T2- and nal intensity of fat on a T2-weighted TSE T1-weighted sequence, in order to better evaluate sequence can be a disadvantage, as this may potential spread to the anterior cranial fossa. reduce the contrast between a tumor and the surThe use of very fast imaging such as single-­ rounding tissues. The contrast can be improved shot techniques is usually not recommended. by performing the sequence with either a chemi- Single-shot (gradient-echo) techniques have in cal shift-moderated fat suppression technique general a lower signal-to-noise ratio and are very (chemical shift-selective fat suppression, water-­ sensitive to magnetic field inhomogeneities (susselective excitation, or Dixon approach ceptibility effects). Also, single-shot sequences (Gaddikeri et al. 2018)) or by applying an addi- often yield a somewhat blurred image, impairing tional inversion recovery preparation pulse with a visualization of thin structures or making short inversion time (SPIR, SPAIR, STIR). accurate delineations. In case of uncooperative

Imaging Techniques

49

patients or when fast scan time is absolutely required, patient positioning and shimming should be performed with the utmost precision. If patient movement is within limits, a segmented approach, probing the entire k-space into several separate acquisitions and thereby reducing the echo-train length, is preferable to minimize the abovementioned drawbacks. The use of parallel imaging grants the benefit of a reduced scan time with the same image quality, or a comparable scan time but with a better image quality. In practice, this means that patient movement artifacts can be reduced and signal-to-noise ratio increased, making parallel imaging an ideal addition for comprehensive head and neck MR imaging. Example parameters for head and neck MRI are listed in Table 3A, 3B.

4.2.4  Contrast Agents In MR studies for head and neck neoplasms, obtaining sequences before and after injection of gadolinium-DTPA (at a dose of 0.1–0.2 mmol/kg body weight) is mandatory. Most neoplasms will show increased signal intensity after contrast agent injection. This will usually increase the contrast between the tumor and the surrounding lesions. However, tumors infiltrating bone marrow may become less well visible after contrast injection, as their signal intensity may become similar to that of the surrounding bone marrow; this problem can be solved by obtaining a fat-­ suppressed sequence (see above). Tumor necrosis becomes better visible after injection of gadolinium; this is of particular importance in staging the neck nodes.

Table 3A  Example parameters for full head and neck examination at 1.5 T with combination of head and neck coils

Parameter Sequence type

Pre-contrast T2-weighted imaging T2-weighted Turbospin-echo (TSE) 203 × 250 291 × 512 0.70 × 0.49 48 4

Pre-contrast diffusion-weighed imaging Spin-echo echoplanar imaging (SE-EPI) 203 × 250 104 × 128 1.95 × 1.95 48 4

0.4 3080/106 Anteroposterior

225 × 300 134 × 256 1.68 × 1.17 48 4.4

Pre- and Post-contrast T1-weighted imaging T1-weighted Turbospin-echo (TSE) 203 × 250 333 × 512 0.61 × 0.5 30 4

0.4 7400/84 Anteroposterior

0 4.3/1.6 Anteroposterior

0 532/8.3 Anteroposterior

2 0.875 136

3 0.75 1502

1 0.75 350

2 1 195

19

104

–

7

–

–

–

–

–

0, 50, 100, 500, 750, 1,000 Fatsat 6:03

–

–

– 3:29 for 25 measurements

– (If needed) 5:11

Field of view (mm) Matrix Pixel size (mm) Number of slices Slice thickness (mm) Interslice gap (mm) TR/TE (ms) Phase encoding direction Averages Phase partial Fourier Bandwidth (Hz/ pixel) Turbo factor/echo train length Parallel imaging factor B-values (s/mm2) Saturation Scan time (min:s)

– 5:42

Dynamic contrastenhanced imaging 3D spoiled gradient echo

50

R. Hermans et al.

Table 3B  Example parameters for full head and neck examination at 3 T with integrated head and neck coil Pre-contrast Pre-contrast diffusion- Dynamic contrastPre- and Post-contrast Parameter T2-weighted imaging weighed imaging enhanced imaging T1-weighted imaging 3D spoiled gradient T1-weighted Spin-echo Sequence type T2-weighted echo Turbospin-echo echoplanar imaging Turbospin-echo (TSE) (SE-EPI) (TSE) Field of view (cm) 240 × 220 228 × 190 294 × 395 240 × 220 Matrix 377 × 440 116 × 100 248 × 360 303 × 400 Pixel size (mm) 0.64 × 0.50 1.96 × 1.90 1.19 × 1.10 0.79 × 0.55 Number of slices 54 2 stacks of 28 162 55 Slice thickness (mm) 3.5 4 1.3 3.5 Interslice gap (mm) 0.4 0.4 0 0.35 TR/TE (ms) 4476/90 5045/64 3.4/1.65 666/16 Phase encoding Anteroposterior Anteroposterior Anteroposterior Anteroposterior direction Averages 2 2 1 2 Phase partial Fourier 1 0.689 0.625 1 Bandwidth (Hz/ 194 2896 1206 196 pixel) Turbo factor/echo 24 61 – 7 train length Parallel imaging 2 2 2 3 factor B-values (s/mm2) – 0, 50, 100, 500, 750, – – 1,000 – (If needed) Saturation – Short tau inversion Spectral selection recovery (STIR) attenuated inversion recovery (SPAIR) Scan time (min:s) 3:48 2 × 2:47 3:15 for 20 3:39 measurements

4.2.5  Additional MRI Techniques Although defining the extent of the primary tumor is often possible based on anatomical criteria, identification of small nodal metastases remains challenging. Also the distinction of post-­ therapeutic tissue changes from residual tumor may be difficult. Since 2000, a lot of progress has been made in the application of functional MRI in head and neck imaging (Vandecaveye et  al. 2010; King and Thoeny 2016). While the clinical application of MR spectroscopy (MRS) remains challenging, dynamic contrast-­ enhanced MRI (DCE-MRI) (Kabadi et  al. 2018; Noij et  al. 2015) and diffusion-­ weighted MRI (DWI) (Driessen et  al. 2015; Connolly and Srinivasan 2018) can now be used as complementary imaging tools in the locoregional imaging evaluation of head and neck cancer.

4.2.5.1  Dynamic Contrast-Enhanced Magnetic Resonance Imaging The tumoral vascular network is substantially different from a morphological and functional point of view compared to normal blood vessels, resulting in a heterogeneous blood flow, with increased capillary permeability (Carmeliet and Jain 2000). DCE-MRI uses serial imaging with high temporal resolution over a lesion or anatomical area prior to, during, and after the bolus injection of a gadolinium-based contrast agent to depict the perfusion properties of tumoral lesions. For head and neck imaging, a T1-weighted gradient-echo sequence is generally used. Compared with ­T2∗-based dynamic susceptibility contrast MRI (DSC-MRI), DCE-MRI requires less contrast agent and is far less prone to susceptibility artifacts. Most importantly, the signal intensity

Imaging Techniques

changes caused by the contrast agent are much slower than in DSC-MRI, and therefore a lower time resolution is allowed (around 4–8 s per volume), providing the opportunity to obtain larger anatomical coverage and higher spatial resolution images, while retaining perfusion information (Fig. 10). In both normal and tumoral tissue, the injected contrast will leak from the vessels into the interstitial space with a variable rate and will start to wash out when the interstitial concentration exceeds the intravascular concentration. The main parameters influencing this rate of interstitial contrast leakage are contrast inflow into the tissue, vessel wall permeability, and the total vessel surface area (Padhani 2003). The signal intensity curve, obtained from the consecutive sequences over time, holds information about tumor perfusion, tracer uptake, and blood volume. After the contrast injection, the enhancement pattern will typically consist of three phases: the upslope or rapid arterial enhancement, the point of maximum enhancement, and subsequent washout. The dynamic signal enhancement on T1-weighted DCE-MRI can be assessed by semiquantitative evaluation of the signal intensity curve or by quantification of the change of contrast-agent concentration using pharmacokinetic modeling techniques. Semiquantitative parameters evaluate discrete points of the signal intensity curve while ignoring the information in the rest of the curve; these include the maximal contrast-enhancement, time to maximal contrast-enhancement (time to peak), and the speed of arterial contrast-enhancement (initial slope) (Fig. 10). Their main advantage is their robustness and reproducibility, facilitating their use in clinical routine, although it should be kept in mind that semiquantitative parameters do not reflect absolute contrast medium concentration in tissues and may show variability because of scanner settings and differences in interpatient cardiovascular physiology (Padhani 2003). Therefore, in order to decrease intra- and interpatient variability during treatment follow-up, normalization of contrast-uptake may be done by analyzing contrast-enhancement in a feeding artery and normalizing the tissue measurements

51

to this arterial input function (AIF) (Port et  al. 2001). For quantitative analysis, the entire contrast-­ time-­curve is fitted by a curve model based on biological assumptions, such as blood volume, blood flow, or permeability. Quantitative parameters that are investigated include the volume transfer constant of the contrast agent (Ktrans) and the rate constant (kep), which are calculated based on a two compartment model correlating the tissue tracer concentration to the difference between arterial plasma (Cp) and interstitial fluid (Ct) ­concentrations (Tofts and Kermode 1991). These parameters have the advantage that they show closer correlation to underlying biologic processes of the vasculature, such as the permeability surface area, blood flow, and relative blood vessel volume. However, quantified models are more complex to analyze, less robust, and more susceptible to artifacting, making them more difficult to use in clinical routine. An intermediate solution to quantify perfusion is the use of the initial area under the signal intensity curve (IAUC) or contrast medium concentration curve (IAUGC) (Padhani 2003). 4.2.5.2  Diffusion-Weighted Magnetic Resonance Imaging DWI allows for tissue characterization based on changes in the random movement of tissue water molecules (Le Bihan et al. 1988). In solid malignant tumors, the high cellular density, intact cellular membranes, and diminished extravascular extracellular space (EES) restrict the random movement of water (Fig. 10). On the contrary, in necrosis and inflammation, the low cellular density and increased EES facilitate the random water movement. Spin-echo EPI-based DWI (SE-EPI-DWI) is most frequently used for imaging of head and neck cancer. The major advantage of an EPI readout sequence is the inherent rapidity as it is a single-shot sequence without the need for refocusing radiofrequency pulses. This allows scanning relatively large volumes with multiple b-values in a short time period, making the technique suitable for the simultaneous evaluation of

R. Hermans et al.

52

a

c

Fig. 10  Patient with persistent pain in the right oropharyngeal area, 4 months after end of chemoradiotherapy for a base of tongue cancer. T1-weighted (a), T2-weighted (b), and contrast-enhanced T1-weighted spin echo sequence (c) show some soft tissue thickening in the right glossotonsillar sulcus and anterior tonsillar pillar (arrow). B1000 diffusion-weighted image (d) shows hyperinten-

b

d

sity in this area, while the native dynamic contrast-­ enhanced image (e) and corresponding perfusion map (initial slope, f) indicate hypervascularity (arrows). These findings are suspect for persistent or recurrent tumor. Histopathology after surgical resection confirmed presence of squamous cell carcinoma

Imaging Techniques

e

53

f

Fig. 10 (continued)

the primary tumor site and all nodal stations in the head and neck (Vandecaveye et al. 2010). As a drawback, SE-EPI-DWI is highly sensitive to susceptibility artifacts, possibly reducing diagnostic quality. In areas with strong susceptibility effects, it is advocated to use a STIR-based, rather than a spectral, fat suppression technique for the SE-EPI-DWI, as this strongly reduces the detrimental failed fat suppression artifacts. In exceptional cases, single-shot turbo spin-echo (SS-TSE)-DWI may be considered as an alternative sequence for EPI-DWI, as this sequence minimizes susceptibility-induced artifacts (De Foer et al. 2008). However, because of the inherent need of 180° spin-echo pulses, SS-TSE-DWI is limited by a high specific absorption rate (SAR), unfavorable SNR, and a very long acquisition time. Therefore, SS-TSE-DWI is only ­useful for evaluation of a limited anatomical area, where susceptibility artifacts make the application of SE-EPI-DWI nearly impossible. Because of the sensitivity of DWI to artifacts, a number of technical optimizations are pivotal to

preserve diagnostic imaging quality. First, it is important that the shortest possible TE is selected, maximizing the signal-to-noise ratio, and minimizing susceptibility and fat-shift artifacts. This requires the scanner to be equipped by a strong gradient system and the use of a high bandwidth. Second, one should rather not use unrealistically high DW imaging matrices, as this increases the echo-train length, introducing T2∗ and susceptibility artifacts. In most clinical settings, an imaging matrix of 128 is sufficient, although in low-susceptibility areas it can be increased up to 192. The third issue concerns positioning of the shim block, mostly in non-STIR-based fat suppression with SE-EPI-DWI at high field strengths. In those situations, the default shim can lead to strongly distorted presentations of failed fat suppression artifacts, and possibly to nondiagnostic images. To avoid this, a manual shim could be attempted, which should be tailored around the total head and neck, excluding large areas of movement and air-tissue boundaries (Fig.  11). Currently, no uniform guidelines exist to optimize

R. Hermans et al.

54

a

b

Fig. 11  Localizer image in the coronal (a) and sagittal plane (b), obtained on a 3T system. The imaging volumes (red boxes) are divided into two sub-volumes, respectively, extending from the skull base to the level of the

hyoid and from the hyoid to the thoracic inlet. Shim boxes (green box and orange box) have been adapted in the phase encoding direction, excluding the air around the head and neck

shimming, as optimal shim approaches are scanner-, field strength, and patient-dependent, but usually the shim should be tailored around the total head and neck, excluding large areas of aircontaining or moving structures. Finally, when DWI needs to cover the entire head and neck, the imaging volume can be subdivided in two separate but spatially linked scan volumes: one extending from the skull base down to the level of the hyoid, and the second one extending further down to the aortic arch. This allows closer positioning of each imaging stack to the isocenter of the magnet, minimizing geometric distortion and failed fat suppression at the slices further away from the isocenter, which is especially useful in higher field strengths, when a spectral-based fat suppression is used. DWI can be evaluated in a qualitative and quantitative way. Qualitative analysis using DWI at a single high b-value (usually b = 1000 s/mm2) offers high sensitivity for detection of tumoral disease but has relatively low specificity (Takahara et al. 2004). Such images can be used for the rapid detection of potentially tumoral localizations. However, DWI-based tissue characterization, especially of lymph nodes, often

requires additional calculation of the apparent diffusion coefficient (ADC) (Fig. 12). In head and neck DWI, at least 3 b-values ranging from 0 to 1000 s/mm2 are preferentially used. The large number of b-values improves the accuracy of the ADC calculation, by minimizing the influence of movement and noise propagation. Also, in highly vascular tumors, the addition of nonzero low b-values can be used to eliminate vascular contributions that may falsely increase ADC values by an additional but variable influence of microperfusion (Thoeny et al. 2005). If complete separation of diffusion and microperfusion is preferred, this is usually performed using the intravoxel incoherent motion (IVIM) approach. IVIM is a 4-parameter model that includes estimations of the slow or true diffusion coefficient (strongly dependent on underlying cellularity), the fast diffusion coefficient (maximally based on microcirculation), and the “perfusion fraction” (relative amount of perfusion effects, between 0 and 1) (Noij et al. 2017). A DWI sequence with 6 b-values and 4  mm slice thickness typically takes around 5  min on a 1.5  T system, and around 3 min on a 3 T system.

Imaging Techniques

a

55

b

d c

e

Fig. 12  Patient suffering floor of the mouth cancer. (a) Sagittal contrast-enhanced MDCT image shows primary tumor at the junction of the floor of the mouth and oral tongue (arrowheads). (b) Axial MDCT image. Slightly enlarged lymph node (minimal axial diameter 12 mm) in the submandibular region (arrow): suspect for metastatic adenopathy. (c–e) Axial diffusion-weighted MR images.

Compared to the b = 0 image (c), the signal clearly reduced in the adenopathy (arrow) on the b = 1000 image (d), indicating easy diffusion of water protons. The ADC-­map (e) shows a relative high signal (ADC  >  0.00130  mm2/s). These findings are consistent with a benign adenopathy. After neck dissection, histologically no tumor was found in this lymph node

R. Hermans et al.

56

5  Positron Emission Tomography 5.1  Physical Aspects positron

Positron emission tomography (PET) allows e+ evaluation of the biodistribution of small amounts 511 keV photon of positron-emitting radiopharmaceuticals and is considered the most sensitive and specific technique for in vivo imaging of metabolic pathways electron and receptor–ligand interactions in the tissues of emen (Jones 1996). PET uses radioisotopes of natural elements, such as oxygen-15, carbon-11, 511 keV nitrogen-13, and fluorine-18, which can be used photon for labeling of most biomolecules, without altering their biochemical properties. Positron-emitting isotopes decay by emission of a positron, which is the positively charged Fig. 13  After the decay of a positron-emitting radioisoantiparticle of the electron. Positrons are formed tope, the positron annihilates with an electron, resulting in during the decay of nuclides that have a large the creation of two annihilation photons, each having an number of protons in their nuclei compared with energy of 511 keV. In a PET camera, a large number of detectors are installed in a ring-shaped pattern around the the number of neutrons, which makes them patient, enabling the simultaneous detection of the unstable. After its emission, the positron travels a opposed photons within a narrow time-window, the so-­ short distance until it annihilates with an elec- called “coincidence detection” tron. In this process, the positron and electron masses are converted into two photons that travel resolution has been achieved with the introducapart in nearly opposite directions (180°) with an tion of smaller detector elements (Townsend energy of 511  keV each (Fig.  13). PET tomo- 2008), enabling the detection of smaller lesions, graphs are designed to detect these photon pairs particularly important in patients with head and and to reconstruct tomographic images of the neck cancer. However, the physics of PET impose regional tracer distribution. Detector pairs are certain well-known limitations on the spatial, installed in a ring-shaped pattern for the detec- temporal, and contrast resolution that can be tion of the opposed photons within a narrow attained in a particular situation. Both the range time-window, enabling the so-called “coinci- the positron travels until annihilation and the dence detection.” Because of this “electronic” residual momentum of the positron–electron pair collimation, PET does not require collimators, as degrade the spatial resolution and will limit the they are necessary in single photon measure- resolution of future whole body PET cameras to ments with gamma cameras. This results in a sig- approximately 2  mm. Other technical achievenificantly higher sensitivity and a relatively high ments contributing to image quality include the spatial resolution of 4–6 mm of PET compared to introduction of iterative reconstruction algoSPECT. rithms and the technology of time-of-flight (TOF) In the past decades significant improvements reconstruction. More recently conventional phohave been made in both hardware and software of tomultiplier tubes have been replaced by solid-­ PET cameras. The sensitivity of PET scanners state photodetectors providing digital photon has been improved with the introduction of new counting, resulting in improved timing and spaPET scintillators, the switch from two-­ tial resolution or improved sensitivity, depending dimensional (2D) to three-dimensional (3D) sys- on the technical design of the digital coupling tems, and the increase of the axial coverage by which differs between vendors (van der Vos et al. adding extra detector rings. An increase in spatial 2017). A recent comparison of image quality and

Imaging Techniques

57

lesion detection between digital and analog PET/ CT has shown that digital PET/CT offers improved image quality and lesion detection capability over analog PET/CT in oncological patients, and was even better for sub-centimeter lesions, especially in organs with heterogeneous appearance and background activity such as the liver (López-Mora et al. 2019). Further improvements that can be expected include the increase of TOF timing resolution and the introduction of PET scanners with a very large axial field of view. Both improvements may offer a better image quality or can be used to shorten the scanning time or to reduce radiation dose while keeping the same image quality (van der Vos et al. 2017).

5.2  Radiopharmaceuticals 5.2.1  Imaging of Glucose Metabolism: 18 Fluorodeoxyglucose The tracer most commonly used worldwide is Fluorine-18-labeled 2-fluoro-2-deoxy-d-glucose, [18F]FDG. This is a d-glucose molecule in which

glucose transporter glucose glucose

hexokinase X phosphatase

glucose-6-PO4

TUMOR CELL

Glycolysis

X

Fig. 14  Uptake kinetics of glucose and FDG in a tumor cell. Glucose transporters facilitate FDG uptake in tumor cells and hexokinase subsequently phosphorylates FDG. The phosphorylated product FDG-6-phosphate is not metabolized in the glycolytic pathway and the activity of glucose-6-­ phosphatase in tumor tissue is low, resulting in the metabolic trapping of FDG-6-phosphate in the tumor cell

a hydroxyl group in the 2-position is replaced by a positron-emitting 18F.  Several decades ago, Warburg et  al. described the higher rate of glucose metabolism in cancer cells compared to nonmalignant tissue (Warburg 1956). After malignant transformation, there is an increased expression of epithelial glucose transporter proteins and an upregulation of hexokinase activity. After intravenous injection, FDG distributes throughout the body in proportion to glucose metabolism of tissues. Glucose transporters facilitate FDG uptake in tumor cells and hexokinase subsequently phosphorylates FDG to FDG-6-­ phosphate. FDG-6-phosphate is not metabolized in the glycolytic pathway because it lacks a hydroxyl group in position-2. As most tumor cells do not contain significant amounts of glucose-­ 6-phosphatase, FDG-6-phosphate will accumulate in the cell, resulting in the so-called “metabolic trapping” (Fig.  14) (Pauwels et  al. 1998). It has been demonstrated that increased FDG uptake by malignant cells, although a function of the proliferative activity, is mainly related to the number of viable tumor cells (Higashi et  al.

hexokinase 18FDG

glucose transporter

18FDG

X phosphatase

18FDG-6-PO 4

58

R. Hermans et al.

1993). However, some benign tumors can also consume considerable amounts of glucose, as can be seen in Warthin tumors of the parotid gland. In addition, inflammatory tissue may exhibit an increased FDG uptake due to glycolytic activity in neutrophils, eosinophils, macrophages, and proliferating fibroblasts, to a degree sometimes more marked than in malignant cells, resulting in a false positive FDG signal (Kubota et al. 1992). This can be a problem in the field of therapy monitoring, where posttreatment inflammation may raise the overall FDG uptake, causing an underestimation of treatment effectivity and a decrease in the specificity of FDG-PET (Schöder et  al. 2009). For this reason, FDG-PET after radiotherapy should not be performed before 10–12 weeks after the end of treatment of head and neck tumors. Despite these limitations, the role of FDG-PET in staging and follow-up of head and neck tumors is well established. Alternatively, some more tumor-specific tracers have been proposed, which should be less sensitive to inflammatory conditions.

5.2.3  I maging of Amino Acid Metabolism: 18FET and 11C-MET Labeled amino acid analogs like 18F-fluoro-­ethyltyrosine (18FET) and 11C-methionine (11C-MET) have been developed for tumor detection. Increased amino acid metabolism results in an accumulation of amino acids in tumor cells. Compared to FDG, the uptake of amino acids in inflammatory cells is lower, resulting in a higher specificity, confirmed in several studies on the use of FET in head and neck cancer. Unfortunately this was at the cost of a lower uptake and sensitivity, making FET unsuitable for accurate staging (Pauleit et  al. 2006) Nevertheless, the higher specificity makes FET-­PET an interesting tool for treatment monitoring. In contrast with FET, MET had the disadvantage of the short half-life of 11C and a high uptake in the salivary glands which may interfere with the detection of tumors or lymph node metastases near these glands. More research is needed to define the exact role of these tracers in the follow-­up of head and neck cancer.

5.2.2  Imaging of Tumor Proliferation: 18 Fluorothymidine 18 F-labeled fluorothymidine, 18FLT, is utilized in oncology as a marker of cellular proliferation. Thymidine is a native nucleoside, which is used by cells for DNA replication. FLT enters the cell, undergoes phosphorylation by thymidine kinase 1, and is being accumulated in the cytosol instead of being trapped into DNA. Thymidine kinase 1 activity is high in proliferating cells and low in quiescent cells. FLT shows less accumulation in inflammatory tissues and shows high accumulation in squamous cell carcinomas of the head and neck. For head and neck cancer staging, however, there seems to be no additional value of FLT compared to FDG (Troost et  al. 2007). More interesting are the results obtained in the prediction and monitoring of early tumor response to radiotherapy, with a decline in uptake already after 1 week of radiotherapy, preceding volumetric tumor response (Troost et al. 2010). However the sensitivity to detect residual tumor after radiotherapy is lower in comparison with FDG (Been et al. 2009).

5.2.4  Imaging of Hypoxia Tumor hypoxia has been associated with a poor response to radiotherapy in head and neck cancer. In contrast with invasive measurements of hypoxia, noninvasive PET imaging provides geographic mapping of hypoxia inside the whole tumor allowing reaching areas not accessible to invasive procedures. This knowledge about the presence and extent of hypoxic areas may be used for dose-painting in radiotherapy planning, where a higher dose may be administered to the more hypoxic parts of the tumor. The most commonly used PET tracer for hypoxia detection is 18F-fluoromisonidazole (18FMISO), belonging to the group of 2-­nitroimidazoles. These compounds are trapped inside the hypoxic cells, which is the basis for their use to measure hypoxia. FMISO uptake has appeared to be an independent prognostic factor in head and neck cancer (Rajendran et al. 2006). However, FMISO has a relatively low tumor-to-­ background ratio due to its relatively high lipophilicity and a slow clearance from normoxic tissues, resulting in reduced image quality. For

Imaging Techniques

this reason several other hypoxia specific tracers with more favorable imaging characteristics have been developed. 18F-fluoroazomycin arabinoside (18FAZA), a second-generation nitroimidazole, has a higher hydrophilicity and better clearance kinetics from blood and normoxic tissue, resulting in a higher tumor-to-background ratio. In the DAHANCA 24 trial, FAZA has demonstrated good prognostic potential given the correlation between high tumor uptake and poor treatment outcome (Mortensen et al. 2012). Other promising PET tracers of the nitroimidazole family with optimized pharmacokinetic and clearance prop18 erties include F-2-nitroimidazolpentafluoropropyl acetamide (18F-EF5) and 18 F-flortanidazole (18F-­ HX4), both previously tested in head and neck cancer and showing reproducible and spatially stable results (Silvoniemi et  al. 2018; Zegers et  al. 2015). However, more studies on the direct comparison of these tracers and on their use during the course of radiotherapy are needed.

5.2.5  Imaging of Molecular Targets An important imaging target in patients with squamous cell head and neck carcinoma is the epidermal growth factor receptor (EGFR), which is involved in treatment resistance mechanisms. Radiolabeled EGFR inhibitors such as cetuximab have great potential to assess EGFR expression across the entire tumor burden before and during treatment (van Dijk et al. 2015). However, most studies were carried out in animal models and more research on the possible clinical role of these tracers is needed.

5.3  T  echnical Aspects of FDG-PET and Integrated FDG-PET/CT in Head and Neck Cancer FDG-PET relies on the detection of metabolic alterations in cancer cells, independent of structural features as provided by morphologic imaging techniques like contrast-enhanced CT and MRI.  Especially in the posttreatment setting, morphologic imaging modalities have distinct limitations in accurate identification of viable

59

tumor within residual masses, in the identification of small tumor deposits (e.g., in normal-size lymph nodes), and in the characterization of secondary enlarged inflammatory lymph nodes (McCabe and Rubinstein 2005). Postradiation changes and anatomic distortions limit the diagnostic accuracy of these anatomic-based imaging studies in the head and neck region. Since biochemical and cellular changes precede size reduction, there is an increasing interest in reponse imaging in vivo using molecular imaging with FDG-PET.  Moreover, FDG-PET has an advantage over conventional imaging because of its whole body coverage and its sensitivity to distant lesions that may be missed by conventional imaging. A major limitation of FDG-PET in the neck is the inadequate anatomic information inherent in metabolic imaging. The interpretation of FDG-­ PET images of the head and neck is particularly difficult because of the presence of multiple sites showing a variable degree of physiological FDG accumulation, such as the salivary glands, lymphoid tissues, muscles, and brown adipose tissue (“USA-Fat”). The introduction of combined PET/CT scanners has overcome this problem by providing coregistered metabolic and anatomic information (Fig.  15). Combined PET/CT has proven to be more accurate than stand-alone FDG-PET by improving the differentiation between physiologic and pathologic FDG uptake and by providing a more accurate tumor localization, hence reducing the number of equivocal findings (Goshen et al. 2006). In addition, FDG-­ PET image quality can be improved by instructing patients not to speak during the FDG uptake phase and by pharmacological interventions with diazepam, reducing muscular uptake, or with propranolol, reducing FDG uptake in brown adipose tissue. Until today, independent of the tumor type, clinically combined PET/CT imaging is usually performed as a whole body imaging tool, which has intrinsic limitations in terms of resolution (affecting sensitivity) and false positive findings (affecting specificity). Moreover, the CT-part of this examination is most often performed as a low-dose nondiagnostic non-contrast CT, which

R. Hermans et al.

60

CT

PET

Fig. 15  A schematic design of a modern PET/CT camera, combining a PET camera and a diagnostic CT scanner within one gantry. A standard whole body FDG-PET/CT protocol is started 60 min after the intravenous injection

of FDG and includes a low-dose spiral CT acquisition without contrast, followed by a PET acquisition of the same region and followed by a complementary contrast-­ enhanced CT study

is sufficient for attenuation correction and anatomic correlation of the PET images, but does not offer the image quality of a diagnostic CT scan. Hence, this whole body approach using general reconstruction algorithms is less suitable for imaging of a complex and dense anatomic region like the neck. More recent data advocate the use of high-resolution head and neck PET/CT protocols for the detection of small lymph node metastases in the neck (Vogel et  al. 2005; Yamamoto et al. 2007; Rodrigues et al. 2009). In such a setting, whole body imaging is being performed for the detection of metastatic disease, while dedicated high-resolution head and neck PET/CT imaging is added for the detection of the primary tumor and/or lymph node metastases in the neck. Ideally, dedicated head and neck PET/CT imaging is being performed with the arms down (along the side of the patient) and should include a diagnostic CT scan with the use of intravenous contrast. The PET acquisition and reconstruction parameters should be optimized for the head and neck region and smaller pixels should be used, reducing the partial volume effect which causes

an underestimation of FDG activity in all lesions smaller than twice the spatial resolution. These improvements enable to optimize integrated FDG-PET/CT imaging for the detection of small primary tumors and small nodal metastases in the head and neck region (Fig. 16). Because of further technical improvements like the introduction of digital PET/CT and PET scanners with a very large axial field of view, the role of PET/CT imaging in the management of head and neck tumors is likely to become even more important in the future.

5.4  PET/MRI Despite the proven value of hybrid PET/CT in oncology, it suffers from some limitations like the lack of true simultaneous imaging (with potential artifacts), the lack of soft tissue contrast of CT images, and the radiation dose of CT. In contrast to CT, MRI gives no additional radiation dose, offers high differentiation of soft tissues, and has functional imaging capabilities. These benefits

Imaging Techniques

61

a

b

Fig. 16  FDG-PET and fused FDG-PET/CT slices illustrating the improvement in image quality between standard whole body images and dedicated high-resolution head and neck images in a patient with a lymph node positive for squamous cell carcinoma on the right side of the neck, and an unknown primary tumor. (a) Whole body images demonstrate avid FDG uptake at the base of the tongue (red arrow) and at both sides of the neck (blue and green arrows). Interpretation of these FDG spots is diffi-

cult because of CT-related artifacts, insufficient image resolution of both PET and CT, and the presence of FDG uptake in brown fat and muscle. (b) Dedicated head and neck images clearly show the primary tumor at the base of the tongue (red arrow) and a positive lymph node at both sides of the neck (blue and green arrows). These dedicated fused head and neck images allow easy differentiation between tumor and physiological FDG uptake in brown fat or muscle

62

have led to the development of hybrid PET/MRI systems, first for small animal imaging, followed by a PET/MRI scanner for human brain imaging and more recently the introduction of PET/MRI scanners for human whole body imaging. To allow simultaneous acquisition of MRI and PET data, analog photomultiplier tubes have been replaced by solid-state photodetectors, which are insensitive to magnetic fields. The PET detector system is inserted into the bore of the MRI scanner, allowing true simultaneous imaging (Slomka et  al. 2016). In comparison with PET/CT, PET/ MRI can reduce the patient’s radiation dose significantly but has a main disadvantage of a higher cost due to expensive cameras and a slower patient throughput. Moreover, several technical challenges, like difficulties with attenuation correction, have delayed the introduction of this new combined imaging modality into routine clinical practice (Boellaard and Quick 2015). Concerning head and neck cancer, additive value of integrated whole body FDG-­PET/MRI may be expected in the detection of perineural spread of tumors, the detection of infiltration of the prevertebral fascia or great vessel walls, the assessment of skull base tumors or tumors with intracranial extension, and in depicting the extent of disease in the presence of dentures or inlays (Queiroz and Huellner 2015). The first results of studies on the role of combined whole body FDG-PET/MRI in staging and follow-­up of head and neck cancer are promising, but more studies are needed to define its role in clinical practice (Chan et al. 2018; Becker et al. 2018).

References Becker M, Varoquaux A, Combescure C et al (2018) Local recurrence of squamous cell carcinoma of the head and neck after radio(chemo)therapy: diagnostic performance of FDG-PET/MRI with diffusion-weighted sequences. Eur Radiol 28:651–663 Been L, Hoekstra H, Suurmeijer A et al (2009) 18F-FLT-­ PET and 18F-FDG-PET in the evaluation of radiotherapy for laryngeal cancer. Oral Oncol 45:211–215 Boellaard R, Quick H (2015) Current image acquisition options in PET/MR. Semin Nucl Med 45:192–200 Carmeliet P, Jain RK (2000) Angiogenesis in cancer and other diseases. Nature 407:249–257

R. Hermans et al. Chan S, Yeh C, Yen T et al (2018) Clinical utility of simultaneous whole-body 18F-FDG PET/MRI as a single-step imaging modality in the staging of primary nasopharyngeal carcinoma. Eur J Nucl Med Mol Imaging 45:1297–1308. Connolly M, Srinivasan A (2018) Diffusion-weighted imaging in head and neck cancer: technique, limitations, and applications. Magn Reson Imaging Clin N Am 26:121–133 Curtin HD, Ishwaran H, Mancuso AA et  al (1998) Comparison of CT and MR imaging in staging of neck metastases. Radiology 207:123–130 De Foer B, Vercruysse JP, Bernaerts A et  al (2008) Detection of postoperative residual cholesteatoma with non-echo-planar diffusion-weighted magnetic resonance imaging. Otol Neurotol 29:513–517 Driessen JP, van Kempen PM, van der Heijden GJ et  al (2015) Diffusion-weighted imaging in head and neck squamous cell carcinomas: a systematic review. Head Neck 37:440–448 Forghani R, Kelly HR, Curtin HD (2017) Applications of dual-energy computed tomography for the evaluation of head and neck squamous cell carcinoma. Neuroimaging Clin N Am 27:445–459 Gaddikeri S, Mossa-Basha M, Andre JB, Hippe DS, Anzai Y (2018) Optimal fat suppression in head and neck MRI: comparison of multipoint Dixon with 2 different fat-suppression techniques, spectral presaturation and inversion recovery, and STIR.  AJNR Am J Neuroradiol 39:362–368 Goshen E, Davidson T, Yahalom R et al (2006) PET/CT in the evaluation of patients with squamous cell cancer of the head and neck. Int J Oral Maxillofac Surg 35:332–336 Henrot P, Blum A, Toussaint B, Troufleau P, Stines J, Roland J (2003) Dynamic maneuvers in local staging of head and neck malignancies with current imaging techniques: principles and clinical applications. Radiographics 23:1201–1213 Higashi K, Clavo AC, Wahl RL (1993) In vitro assessment of 2-fluoro-2-deoxy-D-glucose, L-Methionine and Thymidine as agents to monitor the early response of a human adenocarcinoma cell line to radiotherapy. J Nucl Med 34:773–779 Jones T (1996) The imaging science of positron emission tomography. Eur J Nucl Med 23:807–813 Kabadi SJ, Fatterpekar GM, Anzai Y, Mogen J, Hagiwara M, Patel SH (2018) Dynamic contrast-enhanced MR imaging in head and neck cancer. Magn Reson Imaging Clin N Am 26:135–149 Keberle M, Tschammler A, Hahn D (2002) Single-bolus technique for spiral CT of laryngopharyngeal squamous cell carcinoma: comparison of different contrast material volumes, flow rates, and start delays. Radiology 224:171–176 King AD, Thoeny HC (2016) Functional MRI for the prediction of treatment response in head and neck squamous cell carcinoma: potential and limitations. Cancer Imaging 16(1):23

Imaging Techniques Kubota R, Yamada S, Kubota K et al (1992) Intratumoural distribution of fluorine-18-fluorodeoxyglucose in vivo: high accumulation in macrophages and granulation tissues studied by microautoradiography. J Nucl Med 33:1972–1980 Le Bihan D, Breton E, Lallemand D, Aubin ML, Vignaud J, Laval-Jeantet M (1988) Separation of diffusion and perfusion in intravoxel incoherent motion MR imaging. Radiology 168:497–505 Lell MM, Greess H, Hothorn T, Janka R, Bautz WA, Baum U (2004) Multiplanar functional imaging of the larynx and hypopharynx with multislice spiral CT. Eur Radiol 14:2198–2205 López-Mora D, Flotats A, Fuentes-Ocampo F et al (2019) Comparison of image quality and lesion detection between digital and analog PET/CT. Eur J Nucl Med Mol Imaging 46:1383–1390 Magnano M, Bongioannini G, Cirillo S et  al (2005) Virtual endoscopy of laryngeal carcinoma: is it useful? Otolaryngol Head Neck Surg 13:776–782 McCabe KJ, Rubinstein D (2005) Advances in head and neck imaging. Otolaryngol Clin North Am 38:307– 319, vii Mortensen L, Johansen J, Kallehauge J et al (2012) FAZA PET/CT hypoxia imaging in patients with squamous cell carcinoma of the head and neck treated with radiotherapy: results from the DAHANCA 24 trial. Radiother Oncol 105:14–20 Noij DP, de Jong MC, Mulders LG et al (2015) Contrast-­ enhanced perfusion magnetic resonance imaging for head and neck squamous cell carcinoma: a systematic review. Oral Oncol 51:124–138 Noij DP, Martens RM, Marcus JT et al (2017) Intravoxel incoherent motion magnetic resonance imaging in head and neck cancer: a systematic review of the diagnostic and prognostic value. Oral Oncol 68: 81–91 Padhani AR (2003) MRI for assessing antivascular cancer treatments. Br J Radiol 76:S60–S80 Pauleit D, Zimmermann A, Stoffels G et al (2006) 18F-­ FET PET compared with 18F-FDG PET and CT in patients with head and neck cancer. J Nucl Med 47:256–261 Pauwels E, McCready VR, Stoot JH et  al (1998) The mechanism of accumulation of tumour-localising radiopharmaceuticals. Eur J Nucl Med 25:277–305 Port RE, Knopp MV, Brix G (2001) Dynamic contrast-­ enhanced MRI using Gd-DTPA: interindividual variability of the arterial input function and consequences for the assessment of kinetics in tumors. Magn Reson Med 45:1030–1038 Queiroz M, Huellner M (2015) PET/MR in cancers of the head and neck. Semin Nucl Med 45:248–265 Rajendran J, Schwartz D, O’Sullivan J et al (2006) Tumor hypoxia imaging with F-18-fluoromisonidazole positron emission tomography in head and neck cancer. Clin Cancer Res 12:5435–5441 Robert YH, Chevalier D, Rocourt NL et  al (1993) Dynamic maneuver acquired with spiral CT in laryngeal disease. Radiology 189:298–299

63 Rodrigues RS, Bozza FA, Christian PE et  al (2009) Comparison of whole-body PET/CT, dedicated high-­ resolution head and neck PET/CT, and contrast-­ enhanced CT in preoperative staging of clinically M0 squamous cell carcinoma of the head and neck. J Nucl Med 50:1205–1213 Roele ED, Timmer VCML, Vaassen LAA, van Kroonenburgh AMJL, Postma AA (2017) Dual-energy CT in head and neck imaging. Curr Radiol Rep 5:19. https://doi.org/10.1007/s40134-017-0213-0 Schöder H, Fury M, Lee N, Kraus D (2009) PET monitoring of therapy response in head and neck squamous cell carcinoma. J Nucl Med 50(Suppl 1):74S–88S Silverman PM, Zeiberg AS, Sessions RB, Troost TR, Zeman RK (1995) Three-dimensional imaging of the hypopharynx and larynx by means of helical (spiral) computed tomography. Comparison of radiological and otolaryngological evaluation. Ann Otol Rhinol Laryngol 104:425–431 Silvoniemi A, Suilamo S, Laitinen T et  al (2018) Repeatability of tumour hypoxia imaging using 18F-­ EF5 PET/CT in head and neck cancer. Eur J Nucl Med Mol Imaging 45:161–169 Slomka P, Pan T, Germano G et al (2016) Recent advances and future progress in PET instrumentation. Semin Nucl Med 46:5–19 Takahara T, Imai Y, Yamashita T, Yasuda S, Nasu S, Van Cauteren M (2004) Diffusion weighted whole body imaging with background body signal suppression (DWIBS): technical improvement using free breathing, STIR and high resolution 3D display. Radiat Med 22:275–282 Takes RP, Knegt P, Manni JJ et al (1996) Regional metastasis in head and neck squamous cell carcinoma: revised value of US with US-guided FNAB. Radiology 198:819–823 Takes RP, Righi P, Meeuwis CA et  al (1998) The value of ultrasound with ultrasound-guided fine-needle aspiration biopsy compared to computed tomography in the detection of regional metastases in the clinically negative neck. Int J Radiat Oncol Biol Phys 40: 1027–1032 Thoeny HC, De Keyzer F, Vandecaveye V et  al (2005) Effect of vascular targeting agent in rat tumor model: dynamic contrast-enhanced versus diffusion-weighted MR imaging. Radiology 237:492–499 Tofts PS, Kermode AG (1991) Measurement of the blood-­ brain barrier permeability and leakage space using dynamic MR imaging. 1. Fundamental concepts. Magn Reson Med 17:357–367 Townsend D (2008) Dual-modality imaging: combining anatomy and function. J Nucl Med 49:938–955 Troost E, Bussink J, Hoffmann A et  al (2010) 18F-FLT PET/CT for early response monitoring and dose escalation in oropharyngeal tumors. J Nucl Med 51:866–874 Troost E, Vogel W, Merkx M et al (2007) 18F-FLT PET does not discriminate between reactive and metastatic lymph nodes in primary head and neck cancer patients. J Nucl Med 48:726–735

64 van den Brekel MW, Castelijns JA, Stel HV et al (1991) Occult metastatic neck disease: detection with US and US-guided fine-needle aspiration cytology. Radiology 180:457–461 van der Vos C, Koopman D, Rijnsdorp S et  al (2017) Quantification, improvement, and harmonization of small lesion detection with state-of-the art PET.  Eur J Nucl Med Mol Imaging 44(Suppl 1): S4–S16 van Dijk L, Boerman O, Kaanders J et  al (2015) PET imaging in head and neck cancer patients to monitor treatment response: a future role for EGFR-targeted imaging. Clin Cancer Res 21:3602–3609 Vandecaveye V, De Keyzer F, Dirix P et  al (2010) Applications of diffusion-weighted magnetic resonance imaging in head and neck squamous cell carcinoma. Neuroradiology 52:773–784 Vogel WV, Wensing BM, van Dalen JA et  al (2005) Optimised PET reconstruction of the head and neck area: improved diagnostic accuracy. Eur J Nucl Med Mol Imaging 32:1276–1282

R. Hermans et al. Warburg O (1956) On the origin of cancer cells. Science 123:309–314 Widmann G, Henninger B, Kremser C, Jaschke W (2017) MRI sequences in head & neck radiology – state of the art. Fortschr Röntgenstr 189:413–422 Wirth S, Meindl T, Treitl M, Pfeifer KJ, Reiser M (2006) Comparison of different patient positioning strategies to minimize shoulder girdle artifacts in head and neck CT. Eur Radiol 16:1757–1762 Yabuuchi H, Kamitani T, Sagiyama K et al (2018) Clinical application of radiation dose reduction for head and neck CT. Eur J Radiol 107:209–215 Yamamoto Y, Wong T, Turkington T et  al (2007) Head and neck cancer: dedicated FDG PET/CT protocol for detection  – phantom and initial clinical studies. Radiology 244:263–272 Zegers C, van Elmpt W, Szardenings K et  al (2015) Repeatability of hypoxia PET imaging using 18F-­ HX4  in lung and head and neck cancer patients: a prospective multicenter trial. Eur J Nucl Med Mol Imaging 42:1840–1849

Laryngeal Neoplasms Robert Hermans

Contents 1

Introduction 

2 N  ormal Laryngeal Anatomy  2.1  L  aryngeal Skeleton  2.2  Mucosal Layer and Deeper Laryngeal Spaces  2.3  Normal Radiological Anatomy 

Abstract    65    66    66    67    68

3 S  quamous Cell Carcinoma     68 3.1  General Imaging Findings     71 3.2  Neoplastic Extension Patterns of Laryngeal Cancer     74  rognostic Factors for Local Outcome of P Laryngeal Cancer  4.1  Treatment Options  4.2  Impact of Imaging on Treatment Choice and Prognostic Accuracy  4.3  Use of Imaging Parameters as Prognostic Factors for Local Outcome Independently from the TN-Classification  4

   82    82    85    85

 osttreatment Imaging in Laryngeal P Cancer  5.1  Expected Findings After Treatment  5.2  Persistent or Recurrent Cancer  5.3  Treatment Complications 

   91    91    97  101

 on-squamous Cell Laryngeal Neoplasms N  6.1  Minor Salivary Gland Neoplasms  6.2  Mesenchymal Malignancies  6.3  Hematopoietic Malignancies 

 104  105  106  107

References 

 109

5

Most laryngeal neoplasms are squamous cell cancers, and the larynx is a relative frequent site of head and neck malignancy. Although most laryngeal cancers are detected clinically, and their superficial extent can be well evaluated by endoscopic examination, imaging is required to evaluate the frequent submucosal spread of these tumors. Accurate staging of laryngeal cancer requires imaging, and the radiological findings affect tumor staging and treatment choice. This chapter reviews the normal anatomy of the larynx, and focuses on the imaging findings in laryngeal squamous cell cancer, both before and after treatment. The prognostic value of imaging-derived parameters is explained. The radiological findings in less common laryngeal tumor types are also reviewed.

6

R. Hermans (*) Department of Radiology, University Hospitals, KU Leuven, Leuven, Belgium e-mail: [email protected]

1  Introduction The larynx is one of the most frequent head and neck cancer sites. Nearly all laryngeal malignancies are squamous cell carcinomas. Cigarette smoking and excessive alcohol consumption are well-known risk factors. An important factor in the treatment planning of laryngeal neoplasms is the accuracy of pretherapeutic staging. As most

Med Radiol Diagn Imaging (2020) https://doi.org/10.1007/174_2020_224, © Springer Nature Switzerland AG Published Online: 23 April 2020

65

66

R. Hermans

laryngeal tumors are mucosal lesions, they often can be seen directly or indirectly, but the limitations of clinical and endoscopic tumor evaluation are well recognized. The clinical and radiological evaluation of laryngeal tumors are complementary; the combination of the obtained information will lead to the most accurate determination of tumor extent. Imaging may be used to monitor tumor response and to detect recurrent or persistent disease as early as possible.

2  Normal Laryngeal Anatomy Essentially the larynx consists of a supporting skeleton, a mucosal surface, and in between a soft tissue layer containing fat, some ligaments and muscular structures (Figs. 1, 2, 3, and 4).

2.1  Laryngeal Skeleton The laryngeal skeleton is made up of cartilage and fibrous bands. The foundation of the larynx is the cricoid cartilage. The cricoid cartilage is the only complete cartilaginous ring in the airway. Its horizontal ring-shaped part is known as the arch (arcus), while the higher posterior part is called the lamina. Two paired facets are found at the upper margin of the lamina, allowing articulation with the arytenoid cartilages. The largest supporting cartilage is the thyroid cartilage, essentially consisting of two wings or laminae. The teardrop-shaped epiglottis extends downward and attaches to the inner side of the thyroid cartilage. Only a small part of the epiglottis extends above the hyoid bone, the suprahyoid or free margin of the epiglottis. The vocal ligament stretches from the vocal process of the arytenoid to the inner side of the thyroid cartilage; it forms the medial support of the true vocal cord. The ventricular ligament stretches from the upper arytenoid to the thyroid cartilage, forming the medial margin of the false cord. The epiglottis is held in place by the hyo-­ epiglottic ligament, running through the fatty preepiglottic space.

Fig. 1  Lateral diagram of the larynx showing the cartilaginous skeleton (mucosa, intrinsic laryngeal muscles, and paraglottic fat removed). The vocal ligament (single arrowhead) stretches from the vocal process of the arytenoid (A) to the anterior thyroid cartilage. The ventricular ligament (double arrowhead) runs from the upper arytenoid to the anterior thyroid cartilage. T thyroid lamina, SC superior cornu (of thyroid). The superior cornua are attached to the hyoid by the thyrohyoid ligament (unlabeled thick arrow), which forms the posterior margin of the thyrohyoid membrane. C cricoid cartilage, E epiglottis, H hyoid bone. Note: The small structure at the upper tip of the arytenoid is the corniculate cartilage. It has no clinical significance, but is occasionally seen on CT. The small hole (unlabeled thin arrow) in the thyrohyoid membrane transmits the internal branch of the superior laryngeal nerve that provides sensation to the laryngeal mucosa

Laryngeal Neoplasms

67

Fig. 3  Coronal diagram of the larynx showing the laryngeal subsites. (1) True vocal cord (TVC) consist mainly of the bellies of the thyroarytenoid muscle. (2) False vocal cord (FVC) consists mainly of fatty tissue. TVC and FVC are separated by the slitlike laryngeal ventricle (sinus of Morgagni), extending superolaterally as the sacculus laryngis or appendix. E epiglottis, H hyoid bone, T thyroid cartilage, C cricoid cartilage

Fig. 2  Lateral diagram of the larynx sectioned sagittally in the midline. The slitlike ventricle separates true vocal cord (unlabeled arrow) and false vocal cord (large arrowhead). Small arrowheads aryepiglottic fold, T thyroid cartilage, C cricoid cartilage (lamina), dashed line projection of the arch of the cricoid cartilage, E epiglottis, H hyoid bone, V vallecula

2.2  M  ucosal Layer and Deeper Laryngeal Spaces All these structures are covered by mucosa; the inner larynx is dominated by two prominent parallel bands, the true and false cord, separated by a slitlike opening towards the laryngeal ventricle. The true cord largely consists of a muscle, running parallel and lateral to the vocal ligament, between the arytenoid and thyroid cartilage, hence known as the thyro-arytenoid

Fig. 4  Inner view of the larynx, seen from above, after removal of most soft tissues. A arytenoid cartilage, C cricoid lamina, T thyroid cartilage. The bulk of the TVC is made up of the thyroarytenoid muscle (in dark) running from the inner aspect of the thyroid lamina to the arytenoid cartilage, paralleling the vocal ligament (large arrowhead). The thyroarytenoid muscle can be separated in two bellies. Only the medial portion (vocalis muscle) is seen on the right. The vocal ligament extends from the vocal process (small arrowhead) to the anterior commissure (unlabeled arrow)

R. Hermans

68 Table 1  Subsites within the larynx (UICC 2017) Supraglottis  Suprahyoid epiglottis (including tip, lingual, and laryngeal surfaces)  Infrahyoid epiglottis  Aryepiglottic fold, laryngeal aspect  Arytenoid  Ventricular bands (false vocal cords) Glottis  True vocal cords  Anterior commissure  Posterior commissure Subglottis

muscle. The false cord largely consists of fat. In between the cords, the ventricle is rising into the laryngeal tissue space between the mucosa and supporting skeleton. The relationship of pathological conditions to these three parallel structures is significant in the evaluation of laryngeal cancer. Above the false cords, from the arytenoid cartilages, the mucosa reflects upwards towards the epiglottis, forming the aryepiglottic folds. The part of the larynx at the level of the true vocal cords is called the glottis. The region beneath the undersurface of the true vocal cords until the undersurface of the cricoid cartilage is the subglottis. Above the level of the true vocal cords is the supraglottis. Within these different levels, further subsites are distinguished, important for staging purposes (Table 1). The bare area between the anterior attachment of the true vocal cords, where no or only minimal soft tissue is present against the cartilage, is known as the anterior commissure. The area between the arytenoids is known as the posterior commissure. The fat-containing space between the mucosa and the supporting skeleton is variable in size. The part of this deep space, anterior to the epiglottis, is known as the preepiglottic space. This preepiglottic space is continuous with the more lateral submucosal spaces, extending into the aryepiglottic folds and false vocal cords. These lateral spaces are known as the paraglottic spaces.

At the level of the glottis, the paraglottic spaces are reduced to a very thin stripe of fat just lateral to the thyro-arytenoidal muscles. The paraglottic fat tissue is continuous with a thin infraglottic fat plane, bordered by the conus elasticus. The preepiglottic and paraglottic spaces together are sometimes called the paralaryngeal space.

2.3  Normal Radiological Anatomy The normal radiological anatomy of the larynx is shown in Fig. 5. The appearance of the laryngeal cartilages can vary considerably, depending on the degree of ossification and the amount of fatty marrow in the ossified medullar space. In children, the CT density of the laryngeal cartilages is similar to soft tissue. The (endochondral) ossification of hyaline cartilage starts early in the third decade of life. A high degree of variation exists between individuals. The thyroid cartilage shows the greatest variability in ossification; its ossification may also occur in an asymmetrical fashion. The cricoid and arytenoids show less pronounced variability in ossification. The epiglottis and vocal process of the arytenoids are composed of yellow fibrocartilage; this type of cartilage usually does not ossify.

3  Squamous Cell Carcinoma Squamous cell carcinoma, originating from the mucosal lining, is the most common malignant tumor in the larynx. Mucosal abnormalities can be far better evaluated by the clinician than with even sophisticated imaging methods such as CT or MRI.  However, these tumors have the tendency to spread submucosally, and this extension into the deeply lying tissue planes may be difficult to evaluate by clinical examination alone. The clinical criteria used for giving a tumor a particular T-classification are site-dependent; in

Laryngeal Neoplasms

69

a

b

c

d

Fig. 5  Axial CT images through normal appearing larynges (different patients), from cranial to caudal, illustrating normal radiological anatomy. (a) Level of the free epiglottic margin (arrowhead), at the superior edge of the hyoid bone. (b) Level of the hyoid bone (H). The glosso-epiglottic ligament (curved arrow) separates both valleculae (asterisks). The epiglottis separates the oropharyngeal valleculae from the laryngeal vestibule (dots). The pharyngoepiglottic folds (arrows) correspond to the anterocranial margin of the piriform sinuses. Submandibular salivary gland (SM). (c) Level of superior margin of thyroid cartilage (black arrowhead). Epiglottis (white arrowhead), aryepiglottic fold (arrow), piriform sinuses (asterisks). The fatty space just in front to the epiglottis is the preepiglottic space (PES). The more lateral fatty spaces are called the paraglottic spaces; in the left paraglottic space, the aircontaining tip of the laryngeal ventricle is seen (curved arrow). (d) Level of thyroid cartilage (black arrowhead). Thyroid notch (curved arrow). Superior thyroid cornu (arrow). (e) Level of false vocal cords. Within the fatty

paraglottic space, some tissue with higher density can be seen, corresponding to intrinsic laryngeal muscles and the collapsed laryngeal ventricles (white arrow). The thyroid cartilage shows areas of calcification (black arrowheads), ossification (black arrows), and non-calcified cartilage (white arrowheads). (f) Level of true vocal cords. Arytenoid cartilage (A, partially ossified); lamina of cricoid cartilage (C). The fatty paraglottic spaces are reduced to a thin fatty line (white arrows) between the thyroid cartilage and vocal muscles. Posteriorly, the paraglottic spaces are continuous with the anterior submucosal fat plane in the retrocricoidal part of the hypopharynx (black arrowhead). Hypopharyngeal mucosa (black arrows), posterior submucosal fat plane in retrocricoidal hypopharynx (white arrowheads), pharyngeal constrictor muscle (curved arrow). (g) Level of subglottis. Arch of cricoid cartilage (C). The denser areas correspond to islands of non-ossified cartilage within the otherwise ossified cricoid. Inferior thyroid cornu (black arrowhead). Posterior cricoarytenoid muscle (white arrowhead)

R. Hermans

70

e

f

g

Fig. 5 (continued) Table 2  T-staging of glottic cancer (UICC 2017)

the larynx involvement of different laryngeal subsites and reduced vocal cord mobility are important criteria. The local staging criteria for glottic, supraglottic, and subglottic cancer, as well as the stage grouping, are summarized in Tables 2, 3, 4, and 5. About 65–70% of laryngeal cancers originate at the glottic level, and about 30% at the supraglottic level; laryngeal cancer originating from the subglottic region is rare. The regional (neck) staging criteria for laryngeal cancer are similar to those for hypopharyngeal cancer and sinonasal cancer. The validity of any classification is dependent on the diagnostic methods employed. It is recognized that clinical classification of laryngeal can-

T1

T2 T3 T4

Tumor limited to vocal cord(s) with normal mobility (may involve anterior or posterior commissure) T1a: limited to one vocal cord T2b: involving both vocal cords Extension into supra- and/or subglottis, and/or with impaired vocal cord mobility Vocal cord fixation and/or invasion of paraglottic space, and/or inner cortex of the thyroid cartilage Extralaryngeal tumor spread T4a: tumor invading through the outer cortex of the thyroid cartilage, or tissues beyond the larynx (e.g., trachea, soft tissues of the neck, deep/ extrinsic muscles of the tongue,a strap muscles, thyroid gland, esophagus) T4b: tumor invading prevertebral space, mediastinum, or encasing carotid artery

Genioglossus, hypoglossus, palatoglossus, and styloglossus muscle

a

Laryngeal Neoplasms

71

Table 3  T-staging of supraglottic cancer (UICC 2017) T1 T2

T3

T4

Tumor limited to one subsite of supraglottis with normal vocal cord mobility Tumor invades mucosa of more than one adjacent subsite of supraglottis, glottis or region outside of supraglottis, without fixation of the larynx Vocal cord fixation or invasion of postcricoid area, preepiglottic and/or paraglottic space, and/ or inner cortex of thyroid cartilage Extralaryngeal tumor spread T4a: tumor invading through thyroid cartilage, or tissues beyond the larynx (e.g., trachea, soft tissues of the neck, deep/extrinsic muscles of the tongue,a strap muscles, thyroid gland, esophagus) T4b: tumor invading prevertebral space, mediastinum, or encasing carotid artery

Genioglossus, hypoglossus, palatoglossus, and styloglossus muscle

a

Table 4  T-staging of subglottic cancer (UICC 2017) T1 T2 T3 T4

Tumor limited to subglottis Tumor extends to vocal cord(s) with normal or impaired mobility Vocal cord fixation Extralaryngeal tumor spread T4a: tumor invades cricoid or thyroid cartilage, and/or invades tissues beyond the larynx (e.g., trachea, soft tissues of the neck, deep/extrinsic muscles of the tongue,a strap muscles, thyroid gland, esophagus) T4b: tumor invading prevertebral space, mediastinum, or encasing carotid artery

Genioglossus, hypoglossus, palatoglossus, and styloglossus muscle

a

Table 5  Stage grouping of laryngeal cancer (UICC 2017) Stage 0 Stage I Stage II Stage III Stage IVa Stage IVb Stage IVc

Tis T1 T2 T1,T2 T3 T4a T1, T2, T3 T4b Any T Any T

N0 N0 N0 N1 N0, N1 N0, N1, N2 N2 Any N N3 Any N

M0 M0 M0 M0 M0 M0 M0 M0 M0 M1

cer is insufficient when compared with pathologic classification (Pillsbury and Kirchner 1979). As some criteria (such as vocal cord fixation or

impaired mobility) are prone to subjective interpretation, difficulties occur to clinically determine the extension of a laryngeal tumor, or to reproduce this assessment (Takes et al. 2010). A marked improvement in accuracy is obtained when the results of CT or MRI are added to the clinical findings (Zbären et al. 1996). Imaging is mainly of benefit in detecting deep soft tissue extension, such as in the preepiglottic space, the laryngeal cartilages, and base of tongue. Findings from imaging studies frequently result in an upclassification of the disease.

3.1  General Imaging Findings Criteria used for tumor involvement are abnormal contrast enhancement, soft tissue thickening, presence of a bulky mass, infiltration of fatty tissue (even without distortion of surrounding soft tissues), or a combination of these. Any tissue thickening between the airway and the cricoid arch is considered to represent subglottic tumor. Several studies have compared the CT/MRI findings with the results of whole organ sectioning after total or partial laryngectomy, showing that both techniques are accurate methods to visualize laryngeal pathology (Zbären et  al. 1996). These studies correlating whole organ sectioning and imaging have also revealed some pitfalls. Small foci of mucosal tumor may be difficult to detect or may be invisible, and associated inflammatory and edematous changes may cause overestimation of the tumor extent. Distortion of adjacent normal structures may mimic tumoral involvement. Gross cartilage invasion can be detected with CT.  Due to the large variability in the ossification pattern of the laryngeal cartilages, CT often fails to detect early cartilage invasion. Nonossified hyaline cartilage shows more or less the same density values as tumor on CT images. Demonstration of tumor on the extralaryngeal side of the cartilage is a reliable, but late, sign of cartilage invasion. Asymmetrical sclerosis, defined as thickening of the cortical margin and/ or increased medullary density, comparing one arytenoid to the other, or one side of the cricoid

R. Hermans

72

a

b

c

Fig. 6  Contrast-enhanced axial CT images in a patient with a clinically T3 glottic cancer on the left side. (a) Level of true vocal cords. The left true vocal cord appears thickened and slightly enhancing. The tumor reaches the anterior commissure (black arrowhead). The left paraglottic space is infiltrated (compared to normal opposite side (arrows)). Marked sclerosis of the left arytenoid (white arrowhead). There appears to be some sclerosis of the left thyroid lamina. (b) Level of subglottis. Enhancing soft tissue thickening on left side (arrowheads). Note slight sclerosis cricoid arch on the left (curved arrow). Slight enhancement is seen anteromedially to the subglottis, cor-

responding to subtle extralaryngeal tumor spread or peritumoral inflammation (arrow). (c) Level of false vocal cords. Soft tissue infiltration of the paraglottic space along the thyroid cartilage (arrows). Area of non-ossified thyroid cartilage (arrowhead); as the surrounding ossified thyroid cartilage shows no abnormalities, most likely normal variant. The patient was treated by extended hemilaryngectomy. Pathologic examination confirmed glottic squamous cell carcinoma extending in the subglottic and supraglottic region without evidence of extralaryngeal tumor extension. The arytenoid showed focal neoplastic invasion; in the other cartilages only inflammatory changes were noted

or thyroid cartilage to the other side, is a sensitive but nonspecific finding on CT (Fig.  6) (Becker et  al. 1995). Erosion or lysis has been found to be a specific criterion for neoplastic invasion in all cartilages (Fig.  7). Other signs,

such as cartilaginous blowout or bowing, a serpiginous contour, or obliteration of the medullary space, are not very reliable for cartilage invasion. The combination of several diagnostic CT criteria for neoplastic invasion of the

Laryngeal Neoplasms

a

73

b

Fig. 7  Contrast-enhanced CT images in a patient with a large left-sided glottic squamous cell carcinoma. (a) Axial image. The tumor mass massively invades and destructs the left wing of the thyroid cartilage, growing into the extralaryngeal soft tissues (arrows). (b) Coronal reformatting.

Involvement of the glottic and supraglottic laryngeal level (arrowheads) is seen, as well as massive destruction of left thyroid cartilage wing and extralaryngeal tumor spread (white arrows). Extralaryngeal extension also occurs through the lateral cricothyroid membrane (black arrow)

laryngeal cartilages seems to constitute a reasonable compromise: when extralaryngeal tumor and erosion or lysis in the thyroid, cricoid, and arytenoid cartilages were combined with sclerosis in the cricoid and arytenoid (but not the thyroid) cartilages, an overall sensitivity of 82%, an overall specificity of 79%, and an overall negative predictive value of 91% were obtained (Becker et al. 1995). The controversy on which modality should be preferred to image the larynx dealt for a great part with the accuracy to detect cartilage invasion. MRI was recommended to be the best method to determine the status of the cartilages in the presence of a laryngeal tumor (Becker et al. 1997a). MRI is a more sensitive technique than CT to detect cartilage abnormalities. Areas of cartilage abnormality will result in an increase in signal intensity T2-weighted images and contrast-­ enhanced T1-weighted MRI images. However, due to its high sensitivity for intracartilaginous alterations, MRI causes in a considerable number of cases a false positive result, as distinction between true cartilage invasion and

reactive inflammation, edema, fibrosis, or ectopic red bone marrow is not possible (Becker et  al. 1995). Peritumoral inflammatory changes without tumoral invasion are common coincidental findings in laryngeal cartilages, especially in the thyroid cartilage. The positive diagnosis of neoplastic invasion of the thyroid cartilage should be made with caution on MRI; it has been suggested that one should rather talk about “abnormal signal intensity in the cartilage” instead of “invasion of cartilage” (Castelijns et al. 1996b). A later study suggests that reactive inflammatory changes and true neoplastic involvement of the laryngeal ­cartilages can be better distinguished by comparing the T2-weighted and postcontrast T1-weighted cartilage signal intensity with that of the adjacent tumor tissue. If the cartilage signal intensity on these sequences is higher than that of the tumor, this more likely indicates inflammation; the reported specificity of this sign is 82% (Fig. 8) (Becker et al. 2008). A recent study suggests that dual-energy CT may have additional value compared to conventional

R. Hermans

74

a

b

c

Fig. 8  Axial MR images in a patient suffering transglottic cancer. A large soft tissue mass (asterisk) is seen, centered on the left glottis/subglottis, extending over the midline. On the T1-weighted spin echo image (a), signal loss is seen in the adjacent part of the thyroid cartilage (arrowheads). On the T2-weighted image (b) the tumor shows a lower signal

intensity than the thyroid cartilage. On the gadolinium-­ enhanced T1-weighted image (c) both the tumor and the cartilage show similar enhancement. This cartilage signal behavior suggests cartilaginous inflammation, rather than tumoral invasion. Total laryngectomy was performed; no neoplastic cartilage involvement was present

CT, allowing to distinguish iodine-­ enhancing cartilaginous tumor invasion from non-ossified cartilage, increasing study specificity (Kuno et al. 2018). The prognostic importance of minimal cartilage involvement in laryngeal cancer, as seen on imaging studies, remains debated (Ginsberg 2018). In the absence of extralaryngeal tumor spread, cartilage infiltration does not exclude the possibility of organ preservation therapy (see also below).

3.2

 eoplastic Extension Patterns N of Laryngeal Cancer

3.2.1

Glottic Cancer

3.2.1.1  Local Tumor Spread The most common site of involvement is the anterior portion of the vocal cord, usually at the free margin or upper surface. Involvement of the anterior commissure is commonly present and such lesions may extend over the midline in the

Laryngeal Neoplasms

contralateral vocal cord. As the amount of normal soft tissue visible at the level of the anterior commissure is somewhat variable (Kallmes and Phillips 1997), radiological detection of subtle tumor spread into this structure by imaging can be challenging; however, usually the anterior commissure can be well evaluated during endoscopic examination. Lesions limited to the anterior commissure are rarely seen (3.5 ml had only a local control rate of 22%). The results of the studies by Hermans et  al. (1999a, b) corroborate well these previous findings. Both for glottic and supraglottic cancer, tumor volume was found to be a significant prognostic indicator of local control. In glottic cancer, failure probability analysis showed a clear relation between larger tumor volume and increasing risk for local failure (Fig. 19); a tumor volume of 3.5 ml correlated with a risk for local failure of approximately 50%. From the graph published by Pameijer et al. (1997), an approximately 40% chance of local failure in T3 glottic cancer with a similar tumor volume can be inferred. Also for supraglottic cancer, Hermans et al. (1999a) found a significant relation between tumor volume and risk for local failure (Fig. 20). Compared to glottic cancer, larger supraglottic tumor volumes were found for similar local control rates; similar results can be inferred from other publications (Mancuso et al. 1999). The reason for this different critical tumor volume between glottic and supraglottic cancer is not clear; it might be related to a different local environment in the glottic and supraglottic region, but also (and maybe predominantly) to the more exophytic growth pattern exhibited by supraglottic tumors. However, tumor volume was not found to be an independent predictor of local outcome when a multivariate analysis was performed. In glottic carcinoma, involvement of the paraglottic space at the level of the true vocal cord and involvement of the preepiglottic space were found to be independent predictors of local outcome (Hermans et al. 1999b). Deep involvement of the paraglottic space at the glottic level, as seen on imaging

Laryngeal Neoplasms 25 tumour volume lower 95% CI upper 95% CI

20

Tumour volume (ml)

Fig. 19  Glottic cancer: probability of local failure after definitive radiation therapy versus CT-determined primary tumor volume. Local failure rate is significantly higher with larger primary tumor volume. The 95% confidence intervals for tumor volume are indicated. (From Hermans et al. (1999b) with permission)

87

15

10

5

p=0,01 0 0.15

0.25

0.35

0.45

0.55

0.65

0.75

0.85

0.95

Probability of failure

80 75

tumour volume

70

lower 95% CI

65

upper 95% CI

60 55 Tumour volume (ml)

Fig. 20 Supraglottic cancer: probability of local failure after definitive radiation therapy versus CT-determined primary tumor volume. As for glottic cancer, local failure rate is significantly higher with larger primary tumor volume. The 95% confidence intervals for tumor volume are indicated. (From Hermans et al. (1999a) with permission)

50 45 40 35 30 25 20 15 10

p=0,005

5 0 0.10

0.20

0.30

0.40

0.50

0.60

0.70

Probability of failure

0.80

0.90

0.99

88

studies, is also called the “adjacent sign.” This sign was found to be the only independent predictor of local outcome and survival in a series of 130 patients suffering T1–T2 glottic cancer (Murakami et  al. 2005). In a study where MRI was used as imaging method, intermediate signal in the thyroid cartilage on a T2-weighted sequence and hypopharyngeal extension were found as independent predictors of local control (Ljumanovic et al. 2007). However, these authors did not include involvement of the paraglottic space in their analysis; as mentioned below, cartilage signal alterations can be regarded as an indirect parameter reflecting tumor spread in the deep tissues. An association between tumor volume and thyroid cartilage penetration was demonstrated in a study on laryngectomy specimens (Kats et al. 2013). In supraglottic carcinoma, involvement of the preepiglottic space and subglottic extension were the strongest independent predictors of local control (Hermans et al. 1999a). Also in a study using MRI as imaging tool, preepiglottic space involvement as well as abnormal signal intensities in the thyroid cartilage adjacent to the anterior commissure and /or cricoid cartilage were the independent predictors of local control (Ljumanovic et al. 2004). Again, these cartilage abnormalities can be regarded as reflecting extensive invasion of the deep laryngeal tissues (see below). Tumor volume and degree of involvement of the laryngeal deep tissues are correlated to some extent. However, these descriptive CT parameters may also reflect a more aggressive tumoral behavior, which could explain their stronger association with local recurrence. Fletcher and Hamberger (1974) stated that the preepiglottic space is poorly vascularized; they suggested that the anoxic compartment of tumors invading this space must be significant, and thus relatively radioresistant. Imaging-determined tumor volume was also found to predict the local outcome in patients suffering advanced head and neck cancer, including laryngeal cancer, when treated by chemoradiotherapy (Hoebers et al. 2008; Shiao et al. 2017). In another study, including patients with T3-laryngeal tumors, not tumor volume but deep

R. Hermans

tumor extension was found to independently predict local outcome (Kamal et al. 2018). 4.3.1.2  Cartilage Involvement Laryngeal cartilage invasion is often considered to predict a low probability of radiation therapy alone to control the primary tumor site, and to indicate an increased risk of late complications, such as severe edema or necrosis (Lloyd et  al. 1981; Castelijns et al. 1990). Before the era of computer-assisted cross-­ sectional imaging only gross cartilage destruction, usually occurring in large volume laryngeal tumors, could be detected clinically or by conventional radiography. More limited laryngeal cartilage invasion can be detected with modern cross-sectional imaging methods (Becker et  al. 1995). Earlier studies described an association between CT-depicted cartilage involvement in laryngeal carcinoma and poor outcome after radiation therapy (Silverman 1985; Isaacs et  al. 1988). However, according to others involvement of laryngeal cartilage is not necessarily associated with a reduced success rate of radiation therapy (Million 1989). More recent studies correlating laryngeal cartilage abnormalities, detected on CT, with local outcome after RT seem to corroborate this last point of view. The cartilage most often showing abnormalities is the arytenoid cartilage; usually this cartilage appears sclerotic. An abnormal appearance of this cartilage was not found to be associated with poorer local control, and may be unimportant in terms of prognosis (Tart et  al. 1994; Hermans et al. 1999b). The majority of sclerotic arytenoid cartilages do not contain tumor within ossified bone marrow, which can help to explain why radiation therapy is efficient in a large percentage of patients with isolated arytenoid sclerosis on CT (Becker et al. 1997a, b). Pameijer et al. (1997) found a lower probability of local control in patients with T3 glottic carcinoma, when both arytenoid and cricoid showed sclerosis. These authors assume that if both the arytenoid and cricoid cartilage are sclerotic, the probability of microscopic cartilage invasion will increase. Hermans et  al. (1999a) did also find that cricoid cartilage abnormalities in glottic

Laryngeal Neoplasms

carcinoma yielded a statistically significant lower control rate. Ten out of the 13 patients with sclerosis of the cricoid in this study had also sclerosis of the arytenoid cartilage, corresponding with the “double sclerosis” situation described by Pameijer et  al. (1997). However, the multivariate analysis performed in the study by Hermans et al. (1999b) showed that an abnormal appearing cricoid cartilage is not an independent predictor of poor local control in glottic carcinoma: it lost significance when paraglottic and preepiglottic space involvement were entered in the statistical model. Even relatively subtle cartilage abnormalities, as detected in this study population (sclerosis of the cartilage being the most frequent alteration seen), seem to be correlated with deep tumor extension. More destructive cartilage changes are associated with very bulky tumors, which are not selected for radiation therapy. There are only few data available on the correlation between thyroid cartilage abnormalities as seen on CT and local outcome of glottic cancer after definitive radiation therapy. Some studies explicitly excluded patients showing evidence of thyroid cartilage involvement (Mukherji et  al. 1995; Pameijer et  al. 1997). In the study by Hermans et al. (1999b), where tumor visible on both sides of the cartilage and lysis of ossified cartilage were used as signs of thyroid cartilage invasion, only a limited number of patients with glottic carcinoma had an abnormal appearance of this cartilage. No evidence was found that thyroidal cartilage involvement on itself as seen on CT is associated with a poorer local outcome after definitive radiation therapy but, as said, the number of patients in this study with signs of neoplastic involvement of this cartilage was small. In a study including patients suffering laryngeal cancer, regardless of the site of origin and tumor stage, all treated with primary radiation therapy (with or without chemotherapy), cartilage sclerosis as seen on CT was not associated with a different outcome compared to patients without such cartilage sclerosis (Moubayed et al. 2012). On MRI, cartilage involvement in patients with small sized tumors (under 5 cc) is not correlated with tumor recurrence; abnormal MR sig-

89

nal pattern in cartilage combined with large tumor volume (above 5 cc) worsens the prognosis significantly (Castelijns et  al. 1996a). Consequently abnormal MR signal pattern in laryngeal cartilage should not automatically imply laryngectomy, especially in lesions with smaller volumes. It is incorrect to postulate that radiotherapy cannot cure a substantial number of lesions with cartilage involvement on MRI. Castelijns et al. agree with Million (1989) that minimal cartilage involvement in patients with low staged tumors does not imply a bad prognosis (Castelijns et al. 1995, 1996b). Similar to CT, the presence of cartilage abnormalities on MRI studies may be just reflecting a large tumor volume and deep tumor spread, and as such being only indirectly correlated with local outcome after radiotherapy (Ljumanovic et al. 2004, 2007; Kats et al. 2013). Recent experience shows that organ preservation after chemoradiotherapy is possible in advanced laryngeal cancer invading the cartilage, or even spreading through the cartilage, as visible on imaging studies (Knab et  al. 2008; Worden et al. 2009; Wagner et al. 2012). However, the use of organ preservation as end point in such studies may be questioned; in patients with pretreatment gross cartilage destruction, a poor functional outcome may be expected, because of breakdown of a significant part of the larynx during tumor regression (Wolf 2010). As discussed above, quantification of tumor bulk may be a more reliable way to predict success of therapy. 4.3.1.3  Imaging of the Tumoral Micro-Environment Multiple factors determine the resistance of tumors against radiation treatment and chemotherapy. Tumors may show an intrinsic, genetically determined inherent resistance. However, extrinsic physiological (environmental) factors are also important. Most critical is the presence of less or inadequate and heterogeneous vascular networks leading to chronic “diffusion-limited” tumor hypoxia. There is strong evidence that for some human tumors treatment may fail due to the presence of hypoxia (Overgaard and Horsman 1996).

90

R. Hermans

Identification and quantification of tumor hypoxia is useful as predictor of outcome, but also to select patients for concomitant radiosensitizing therapy to overcome the hypoxia effect. Treatments such as hyperbaric oxygen or carbogen (95–98% O2 with 2–5% CO2) breathing during RT have been extensively investigated and initiated in clinics (Kaanders et al. 2002). The adequate appreciation of tumor hypoxia may also lead to the efficient use of hypoxia-directed treatments such as bioreductive drugs or gene therapy. Until now one has to rely on invasive methods, e.g., biopsy-based immunohistochemistry techniques, or the use of Eppendorf oxygen-sensitive electrodes to screen tumors for hypoxia. However, oxygen-sensitive needle electrodes can only to a certain extent be used, as some primary tumors (such as laryngeal cancers) are deeply seated and difficult to reach. There is a clear need for noninvasive methods to investigate the tumoral micro-environment. Nuclear imaging methods (such as PET imaging) may provide important information on tumor physiology. There is evidence that CT and MR studies, classically used to demonstrate the anatomic position and extent of tumors, are able to provide additional, biological information (Rijpkema et al. 2001, 2002; Hermans et al. 2003; Bisdas et al. 2010).

contained within the resected specimen in a successfully performed laryngectomy. Large tumors are more likely invading the laryngeal framework and grow extralaryngeally (Mukherji et al. 2000). It is often suggested that cartilage involvement precludes voice-sparing partial laryngectomy (Tart et  al. 1994; Becker et  al. 1995; Castelijns et  al. 1996b). However, one study indicated that cartilage alterations, as seen on preoperative CT, are not correlated with the local outcome of patients treated by a speechpreserving surgical technique: no increased local failure rate was observed in the patients with cartilage alterations (1 of 11) over those without cartilage abnormalities (1 of 5) (Thoeny et al. 2005). The used surgical technique in this study (extended hemilaryngectomy with tracheal autotransplantation) allows resection of the hemilarynx, including half of the cricoid cartilage. Therefore, areas of possible neoplastic cartilage involvement are very likely to be included in the resection specimen. The inability of other speech-preserving surgical techniques to adequately resect areas of laryngeal framework invasion may falsely lead to the belief that cartilage involvement, in itself, is a contraindication for partial laryngectomy (Thoeny et al. 2005).

4.3.2 P  redicting Local Outcome After Surgery One study addressed the correlation between volume of supraglottic cancer, as assessed on imaging studies, and outcome after surgical therapy. This study examined a small population with few local recurrences; patients with a tumor volume over 16 ml were found to have a significantly worse local outcome than those with smaller volumes (Mukherji et  al. 2000). The threshold tumor volume in this surgical series is greater than threshold tumor volumes reported for supraglottic cancer treated by radiotherapy (see above). This can be expected as during laryngectomy the tumor is resected en bloc. The endolaryngeal soft tissues of the larynx are contained within a cartilaginous framework; the primary tumor should therefore be completely

4.3.3 T  owards Risk Profiles Incorporating Imaging Findings As staging procedure, CT and MRI have an important function in corroborating clinical findings and ruling out more extensive disease. Accurate staging is critical in decision-making in oncology (Barbera et al. 2001). However, to what extent CT or MRI influence treatment decisions in laryngeal cancer is currently not very clear, and likely varies from institute to institute. This influence depends on the conducted treatment policy, more precisely on the relative role of radiotherapy and surgery as primary treatment modality in more advanced laryngeal cancer. The parameters defined in the T-classification are mainly based on clinical examination; the addition of modern imaging methods in staging

Laryngeal Neoplasms

91

laryngeal cancer may change the prognostic information of the T-classification itself, by causing stage migration (Piccirillo and Lacy 2000; Champion and Piccirillo 2004). Furthermore, imaging-derived parameters such as tumor volume and depth of invasion in the deep tissues are stronger related to local outcome than the T-categories. Pure morphologic criteria cannot explain entirely the biologic behavior of a tumor and its response to treatment. Ongoing research focuses on the evaluation with radiological methods of tumor microvascularization, perfusion and oxygenation, factors known to be of important prognostic value. New classification systems should be conceived, incorporating not only morphologic tumor extent as within the present TNM system, but also including other variables with independent prognostic significance (Takes et al. 2010).

5.1.1 E  xpected Tissue Changes After Radiotherapy Within the first 2 weeks after radiotherapy, there is an acute inflammatory reaction within the deep tissues. Increased permeability, due to detachment of the lining endothelial cells within small blood and lymphatic vessels, results in interstitial edema. After this initial period of a few weeks, there is progressive thickening of the connective tissue. Endothelial proliferation is also seen, eventually resulting in complete obstruction of the vessels. The reduction in venous and lymphatic drainage results in further accumulation of interstitial fluid. Then the fibrosis becomes progressively more advanced, but the interstitial edema may be reduced by formation of collateral capillary and lymphatic channels.The changes visible on posttreatment CT and MR images depend on the radiation dose and rate, the irradiated tissue volume, and the time elapsed since the end of radiation therapy (Mukherji et al. 1994a; Nömayr et al. 2001). Changes which may be seen include (Fig. 21):

5  Posttreatment Imaging in Laryngeal Cancer

–– Thickening of the skin and platysma muscle. –– Reticulation of the subcutaneous fat and the deep tissue fat layers. –– Edema in the retropharyngeal space. –– Increased enhancement of the major salivary glands, followed by size reduction of these glands: postirradiation sialadenitis. –– Atrophy of lymphatic tissue, in both the lymph nodes and Waldeyer’s ring. –– Thickening and increased enhancement of the pharyngeal walls. –– Thickening of the laryngeal structures, with increased density of the fat in the preepiglottic and paraglottic spaces.

5.1  E  xpected Findings After Treatment After treatment of a head and neck cancer, a number of tissue changes become visible on CT and MR images of the neck. These expected ­alterations should be known, so that they are not misinterpreted as evidence of persistent or recurrent tumor. Imaging may be used to monitor tumor response and to try to detect recurrent or persistent disease before it becomes clinically evident, possibly with a better chance for successful salvage. Treatment complications, such as soft tissue or cartilage/bone necrosis, are less frequent than tumor recurrences, but these conditions may be clinically sometimes difficult to distinguish. Although definitive distinction between necrosis and recurrent tumor may also radiologically be difficult, imaging findings may be helpful in guiding further patient management (Hermans 2004).

These tissue changes are most pronounced during the first few months after the end of radiation therapy, and diminish or even resolve with time. It is important to note that the expected tissue changes after radiation therapy appear symmetrical, unless the neck was irradiated using asymmetric radiation portals. The laryngeal cartilages do not show changes after irradiation. Reduction in the degree of cartilage sclerosis in the neighborhood of the tumor

R. Hermans

92

a

b

c

d

Fig. 21  Patient with supraglottic squamous cell carcinoma, staged T3N0, treated by definitive radiotherapy. Axial contrast-enhanced CT images are shown, obtained just before and 3  months after completion of radiation treatment. (a, b). Level of lingual tonsil. After radiotherapy (b), apart from diffuse increased attenuation of the neck fatty tissue, thickening of the free edge of the epiglottis (white arrowhead), platysma muscles (curved arrows), and oropharyngeal walls is seen. Slight amount of retropharyngeal edema is present (black arrowhead). Note also increased enhancement of the submandibular salivary glands (asterisks), corresponding to radiation

sialadenitis, and volume reduction of lingual tonsil (arrows). (c, d) Level of supraglottis. Before radiotherapy (c), a large supraglottic tumor mass (asterisk) is seen, infiltrating the preepiglottic and right paraglottic space; normal left ventricle, containing air bubble, in left paraglottic space (arrow). After radiotherapy (d), the tumor mass disappeared; increased attenuation of the paraglottic fat spaces, somewhat more pronounced in former tumor bed; no mass lesion can be recognized. Laryngeal ventricle is now visible on both sides (arrows). Thickening and increased enhancement of the hypopharyngeal walls (arrowhead)

has been described, and this appears to correlate with local control (Pameijer et al. 1999).

the introduction of various reconstructive materials, such as pedicled or free soft tissue flaps, grafts, and protheses.

5.1.2 E  xpected Findings After Laryngeal Surgery The limits of surgical therapy are determined by the balance between obtaining cure by radical resection of the tumor, and leaving the patient in a functionally and esthetically acceptable situation. More extensive resections are possible by

5.1.2.1  Laser Resection The expected findings after transoral laser excision of a laryngeal cancer depend on the amount of tissue resected. The laryngeal soft tissues may appear normal, or show a focal tissue defect (Fig.  22). After a more extensive resection, the

Laryngeal Neoplasms

a

93

b

Fig. 22 (a) Recurrent glottic squamous cell cancer, presenting as soft tissue thickening (arrows), 2 years after radiotherapy for a right-sided glottic cancer (T2N0). (b) Situation

7 months after partial cordectomy by transoral laser resection: a soft tissue defect is seen in the anterior half of the right true vocal cord (arrows); no evidence for recurrent cancer

laryngeal soft tissue may be replaced by scar, appearing as homogenous but relatively dense tissue with a straighter inner border (Maroldi et  al. 2001); in such cases, differentiation with recurrent tumor may be difficult and correlation with endoscopic findings is necessary. In case of doubt, biopsy is warranted.

(Maroldi et al. 2001). The differentiation between redundant or hypertrophic mucosa, as well as scar tissue, from recurrent cancer, may be difficult. Horizontal supraglottic laryngectomy can be performed in supraglottic cancer staying above the level of the ventricles; this procedure is not performed when the tumor infiltrates both arytenoids (one arytenoid can be resected), the posterior commissure, the postcricoid area, the apex of the sinus piriformis, the glottis, or the thyroid cartilage. Minimal tongue base invasion is not a contraindication. Almost all of the larynx above the level of the ventricles is removed. The residual thyroid cartilage is pulled upwards and sutured to the hyoid bone. Limited glottic cancer can be treated by vertical hemilaryngectomy. The most limited variant of this procedure is a cordectomy, where the entire vocal cord is removed from the anterior commissure to the vocal process of the arytenoid. In a frontolateral laryngectomy, the true and false vocal cord is removed, as well as the greatest part of the ipsilateral thyroid cartilage, including the angle to encompass the anterior commissure; the vocal process of the arytenoid can also be included. In a frontal laryngectomy, the anterior portion of both vocal cords is removed, together with the anterior commissure; a modified frontal laryngectomy is Tucker’s‚ near-total technique, using the epiglottis as reconstructive tissue (Fig. 23).

5.1.2.2  Partial Laryngectomy The aim of partial laryngectomy is to combine radical tumor resection with preservation of laryngeal function. This requires continuity and patency of the airway, separation of the airway and digestive tract, and sparing or reconstruction of the glottic phonation function. Traditional partial laryngectomies include horizontal supraglottic laryngectomy and vertical hemilaryngectomy, but more complex surgical techniques are also being employed (Maroldi et al. 1997; Delaere et al. 2007; Ferreiro-­Argüelles et al. 2008). The postoperative radiological findings depend on the technique employed. Changes in the laryngeal framework offer landmarks for interpreting postoperative findings. However, a somewhat different appearance for the same technique may be encountered among different patients, depending on technical adaptations needed for adequate tumor resection. The postoperative soft tissue changes are less predictable, depending on individual differences in healing, and variations in amount of edema and scarring

R. Hermans

94

a

b

c

Fig. 23  Contrast-enhanced CT images in a patient who was treated by a frontal laryngectomy (according to Tucker) for a carcinoma in the anterior commissure. Three years later, the patient presents with increasing dysphonia. Clinically, swelling of the right false vocal cord is noted with an intact mucosa. (a) Axial section at the level of the arytenoid cartilages (arrowheads). Defect in the anterior part of the thyroid cartilage (arrows); the anterior part of the left true vocal cord (curved arrow) has been resected. On the right side, a centrally necrotic soft tissue mass is seen (asterisk), indicating tumor recurrence. (b) Coronal

reformatting. Level of true vocal cord is indicated on left side by arrow. The recurrent tumor on the right (arrowheads) grows from the false vocal cord region into the true vocal cord; early subglottic extension may be present (lower arrowhead). (c) Sagittal reformatting. The upper part of the epiglottis (arrows) has a more anterior course as normally expected, as this structure was used to close the thyroid cartilage defect. The recurrent tumor (asterisk) abuts the upper margin of the arch of the cricoid cartilage, appearing sclerotic (arrowhead). No neoplastic cartilage invasion was present histologically

Laryngeal Neoplasms

95

If more extensive involvement of the arytenoid is present (possibly with involvement of the cricoarytenoid joint) and/or subglottic extension is present, these procedures are not performed. Extended hemilaryngectomy may than be an alternative. During this procedure, half of the larynx, including half of the cricoid cartilage, is removed. The large defect in the larynx is a

reconstructed with a tracheal patch, revascularized by a freely transplanted radial forearm soft tissue flap. Full height cricoid defects can be closed using this patch in a position comparable to unilateral laryngeal paralysis. This is a functional ­reconstruction, allowing the patient to breathe and speak through his larynx, and swallow ­without aspiration (Delaere et al. 2007) (Fig. 24). b

c

Fig. 24  Contrast-enhanced CT images in a patient treated by extended hemilaryngectomy for a right-sided true vocal cord carcinoma. (a) Axial section at the level of the left true vocal cord. Left arytenoid (arrowhead). The right hemilarynx was resected, and the defect closed by a tracheal patch (arrows). The fatty structure along the tracheal patch (aster-

isk) corresponds to the radial forearm fascial flap. (b) Axial section at the level of the subglottis. The subglottic airway is reconstructed by the tracheal transplant. (c) Coronal reformatting shows restoration of the laryngeal airway by the tracheal transplant (arrows). Left true vocal cord (arrowhead); cricoid cartilage (c); thyroid cartilage (t)

96

Some advanced glottic and supraglottic cancer can be treated by supracricoid partial laryngectomy (SPL), entailing en bloc resection of all tissues between the upper margin of the cricoid cartilage and the inferior margin of the hyoid bone, including the true and false vocal cords. Only the arytenoid on the less involved site is left in place. For glottic cancers without involvement of the supraglottis, the upper two thirds of the epiglottis can be preserved; this variant is known as SPL with cricohyoidoepiglottopexy (CHEP) (Gavilan 2000). 5.1.2.3  Total Laryngectomy Complete removal of the larynx may be required as primary treatment of extensive laryngeal cancer or for salvage of tumor recurrence after radiation treatment or failed partial laryngectomy. When the larynx is removed, the airway and upper digestive tract become completely separated. The airway will then end at a tracheostomy in the base of the neck. If, following the laryngectomy, not sufficient hypopharyngeal tissue is left for creating a neopharyngeal lumen of acceptable diameter, a soft tissue flap is used to create a wider lumen. A pedicled pectoralis major musculocutaneous flap is commonly used for this purpose (Fig. 25). The pectoralis major flap

Fig. 25  Axial contrast-enhanced CT image. Situation after total laryngectomy. The neopharynx is reconstructed by residual pharyngeal tissue (arrows) and a musculocutaneous soft tissue flap (pectoral major flap), containing skin (arrowheads), subcutaneous fat (black asterisk), and muscle (white asterisk)

R. Hermans

has an excellent blood supply. The skin of the flap borders the lumen, while the bulk of the flap fills the soft tissue neck defect, creating a more acceptable aesthetic appearance. On imaging studies, the pectoralis major flap appears initially as a bulky soft tissue structure, showing the characteristics of muscle; gradually, denervation atrophy appears, causing volume loss and fatty replacement of the muscle. At the time of imaging, the muscle denervation may be incomplete; fiber-like structures with muscle density within the flap should not be confused with tumor recurrence. Sometimes a radial forearm flap is used to create a neopharynx (Fig.  26), or an intestinal structure is transplanted to function as neopharynx. Between the proximal trachea and esophagus, a small one-way valve (such as a Provox voice prothesis) is placed, allowing escape of air from the proximal trachea to the esophagus if the tracheostome is closed by the patient. In this way the patient has a lot of air available for producing pharyngeal speech, allowing more rapid speech rehabilitation. Such a valve is visible on imaging studies as a small tube, situated in the wall between the proximal trachea and upper esophagus (Fig. 27). Commonly during laryngectomy, tissue of the thyroid gland is removed. Unilateral thyroidectomy may be performed, to facilitate surgical access to the larynx and to remove at the same time a site of potential direct spread of the cancer. Another option is to remove the isthmus of the thyroid gland, leaving the two thyroid lobes. This remnant thyroid tissue is usually easy to recognize because it shows a high density, related to the high iodine concentration in the gland and its strong vascularization. However, as the normal shape of the thyroid gland is lost, these remnants usually show a rounded or oval appearance. Thyroid tissue may appear inhomogeneous due to the presence of nodular hyperplasia, adenomas, or cysts. It is important that these thyroid remnants are not confused with recurrent cancer; unlike recurrent cancer, these have well-defined borders (Fig. 28).

Laryngeal Neoplasms

Fig. 26  Axial contrast-enhanced CT image, in a patient treated by total laryngectomy. The neopharynx is reconstructed by a free radial forearm flap (arrowheads; inner enhancing rim is skin); the soft tissues are anteriorly covered by a pedicled pectoralis major flap (arrows)

Fig. 27  Axial CT image at the level of the tracheostomy, in a patient who underwent total laryngectomy. Normal appearance of a voice prothesis (arrow), placed through the tracheo-esophageal septum

5.2

Persistent or Recurrent Cancer

5.2.1 Imaging Strategies and Findings Posttreatment imaging is useful to confirm the presence of clinically suspected tumor recurrence.

97

Fig. 28  Axial contrast-enhanced CT image. The neopharynx is seen lying between both thyroid lobes (black asterisks). The thyroid isthmus was resected during the laryngectomy. The inhomogeneous appearance of the thyroid lobes is caused by nodular hyperplasia. Absence of left internal jugular vein along the common carotid artery (arrow), resected during radical neck dissection. Soft tissue flap (white asterisk)

On CT or MRI, tumor recurrence appears after radiation therapy as a soft tissue mass at the primary site and/or as an enlarged (and/or centrally liquefied) neck adenopathy. After surgical treatment, the most reliable imaging finding in recurrent tumor is an enhancing soft tissue mass (Figs.  23 and 29); after partial laryngectomy, destruction of residual laryngeal cartilage may be seen. Early tumor recurrence may be difficult to distinguish from tissue changes induced by therapy. Therefore, it is recommended to obtain a follow­up CT or MR study after surgical, radiation, or combined treatment for a laryngeal neoplasm with high-risk profile (Hermans et  al. 2000; Schwartz et al. 2003). Probably the best time to obtain such a baseline study is about 3–6 months after the end of treatment. Such a baseline study allows treatment-caused changes in the head and neck tissues to be documented. By comparing subsequent studies with the baseline study, it becomes possible to detect with more confidence tumor recurrences or treatment complications, and this at an earlier stage than is possible with clinical follow-up alone (Fig.  30). In patients with laryngeal cancer, CT is an adequate imaging modality for pre- and posttreatment imaging, but similar results can be obtained using MRI (Ljumanovic et al. 2008).

98

R. Hermans

these patients will develop a local failure (Pameijer et  al. 1999). Further exploration in such post-radiotherapy CT-score 3 patients is warranted. PET-CT imaging may prove to be a useful intermediate step in cases where biopsy is considered too risky, or if a biopsy result is returned as negative (Fig. 31). Indeed, the predictive value of a negative biopsy for local control is reported to be only 70% (Keane et al. 1993); this is likely due to sampling error, as tumor recurrences initially develop submucosally and can therefore not be accurately targeted. In cases of contradiction between the clinical findings, CT findings, results of radionuclide studies and/or Fig. 29  Axial contrast-enhanced CT image, after total biopsy, close clinical follow-up and repeat imaglaryngectomy for squamous cell carcinoma. Enhancing soft tissue mass (asterisk) at the anterolateral side of the ing studies are needed. neopharyngeal lumen (arrows): recurrent cancer The local outcome of patients initially classified as post-radiotherapy CT-score 2 is indetermiThere is evidence that the baseline study after nate. Unless clinical examination is already radiotherapy carries important predictive infor- suspect for local failure, further follow-up CT mation regarding the eventual local outcome: studies are needed in these patients; a time interseveral studies show that CT may be useful in the val of 3 to 4 months is recommended, to be conearly differentiation of treatment responders from tinued up to 2 years after completion of radiation nonresponders in irradiated laryngeal and hypo- treatment. pharyngeal cancer (Hermans et  al. 2000; Another strategy is to use PET-CT as the iniMukherji et al. 1994b). tial baseline study, in patients treated with Based on the appearance of the larynx/hypo- advanced disease with low clinical suspicion of pharynx on an early post-radiotherapy CT study, recurrence, and in patients with nonspecific a prediction of long-term local outcome can be symptoms that could indicate recurrence but made according to the following scores: without a clinically obvious mass. PET-CT has a 1  =  expected post-radiotherapy changes, i.e., high negative predictive value; however, false complete resolution of the tumor at the primary positive results are not uncommon (Purohit et al. site and symmetrically appearing laryngeal and 2014); cross-sectional imaging should then be hypopharyngeal tissues, as described above; performed for an equivocal or positive PET study 2 = focal mass with a maximal diameter of 1 cm, or 2 cm but ≤4 cm in greatest diameter without fixation of hemilarynx T3 Tumor >4 cm in greatest dimension or resulted in clinical fixation of the hemilarynx or involves esophageal mucosa T4  T4a Moderate advanced local disease: Tumor invades thyroid/cricoid cartilage, hyoid bone, thyroid gland, esophageal musculature, or surrounding muscles and/or subcutaneous fat planes  T4b Very advanced local disease: Tumor invades prevertebral structures, encases carotid artery, or involves mediastinal structures Regional lymph nodes in HPV negative cancers NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1 Metastasis to a single ipsilateral lymph node ≤3 cm in greatest dimension without extracapsular tumor spread N2 Metastasis to a single ipsilateral lymph node >3 cm but ≤6 cm in greatest dimension, or to multiple ipsilateral lymph nodes of ≤6 cm in greatest dimension, or to bilateral or contralateral lymph nodes ≤6 cm in greatest dimension without extracapsular tumor spread N3 Metastasis to at least one lymph node >6 cm in greatest dimension and or presence of extracapsular tumor spread Regional lymph nodes in HPV positive cancers NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1 Metastasis to a single ipsilateral lymph node ≤6 cm in greatest dimension without or with extracapsular tumor spread N2 Metastasis to a single ipsilateral lymph node >6 cm, multiple ipsilateral or bilateral lymph nodes Note: Metastases at the level VII are considered regional lymph node metastasis. Source: Adapted from American Joint Committee on Cancer (2018)

Table 2 Staging of primary tumors of the cervical esophagus Primary tumor TX Tumor cannot be staged due to missing information Tis High-grade dysplasia confined by the basement membrane T1 Tumor involves the lamina propria, muscularis mucosae, or submucosa T2 T1a: Tumor involves lamina propria or muscularis mucosae T1b: Tumor involves submucosa Tumor invades muscularis propria T3 Tumor involves adventitia T4 Tumor infiltrates adjacent structures T4a: Resectable tumor invading pleura, pericardium, peritoneum, diaphragm, or azygos vein T4b: Unresectable tumor invading other adjacent structures, such as aorta, vertebral body, or trachea Regional lymph nodes NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1 Metastasis in 1–2 regional lymph nodes N2 Metastasis in 3–6 regional lymph nodes N3 Metastasis in ≥7 regional lymph nodes Source: Adapted from reference American Committee on Cancer (2018)

3.3

Joint

 econdary Involvement by S Other Tumors

Infiltration of the hypopharynx and cervical esophagus by surrounding tumors is rare. It may occur with advanced head and neck tumors as well as thyroid, tracheal, and bronchogenic carcinomas (Roychowdhury et al. 2000). MR is more sensitive than CT for detection of hypopharyngeal and or esophageal invasion by an adjacent malignancy due to its higher soft tissue definition. Focal areas of increased T2 signal intensity raise the suspicion for invasion (Fig.  14). Focal enhancement following contrast administration might also be a sign of infiltration; however, it is not as specific as the T2 changes (Fig.  14b, c). Circumferential mass at the level of the cervical esophagus has been reported to be the most sensitive and specific sign of invasion (accuracy of 100%). In contrast, intact adjacent fat planes,

Neoplasms of the Hypopharynx and Proximal Esophagus

a

129

b

c

Fig. 14 (a–c) Axial contrast-enhanced T1-weighted image (a) reveals are partially necrotic, aggressive right thyroid mass (TM) that is extending into the right tracheoesophageal groove and severely displacing the esophagus to the left. Notice the normal appearance of the different layers of the posterior and left lateral cervical esophagus (black arrowheads in (a)) with full thickness involvement right laterally and anteriorly (small arrows in (a)). In addition, the thyroid mass is growing through the common wall of the trachea and esophagus (large arrows in (a)) to involve the posterior tracheal lumen (white arrowheads in

(a)). In addition, the axial, non-contrasted T1-weighted image (b) at the post-cricoid level reveals subtle foreshortening of the right lateral submucosal fat plane (arrowheads in (b)) when compared to the left (arrow in (b)) suggesting early invasion of the hypopharynx. This is confirmed on the axial T2-weighted image (c) at the same level as (b) clearly showing abnormal increased signal intensity of the lateral aspect of the right posterior pharyngeal wall (arrowheads in (c)) at the post-cricoid level when compared to the normal dark signal intensity of the left posterior pharyngeal wall musculature (arrows in (c))

absence of wall thickening, and normal T2 wall signal intensity indicate no invasion with a very high degree of confidence (Fig. 1) (Roychowdhury et al. 2000).

that the clinical tumor stage increases in up to 90% of patients with cross-sectional imaging. Changes in T-stage account for two-thirds of the tumor upstaging due to detection of lateral soft tissue involvement in 88%, and bone or cartilage invasion in 23% of the patients. In one-third of the patients, the N-stage was responsible for the tumor upstaging. Comparison of the accuracy of tumor staging with pathological findings revealed that the clinical examination is less accurate with 58% than CT and MR imaging with 80% and 85% accuracy, respectively. These facts emphasize the essential role of cross-sectional imaging in staging of hypopharyngeal and esophageal

4  Cross-Sectional Imaging Cross-sectional imaging with CT and MR is critical in the evaluation of patients with hypopharyngeal and or cervical esophagus malignancies (Wenig et al. 1995; Becker et al. 2008; Aspestrand et al. 1990; Nowak et al. 1999; Prehn et al. 1998; Thabet et  al. 1996). Overall, it has been shown

I. M. Schmalfuss

130

cancer. Interestingly, none of the other cancers of the head and neck region show such a significant impact of cross-sectional imaging upon staging. CT and MR imaging is, however, only helpful when performed with an appropriate protocol covering the potential sites of tumor spread, u­ tilizing the appropriate imaging study, sequence and window display. Although each radiologist has personal preferences and there are vendor-­related variations in hardware and software, the subsequent generic imaging guidelines should be followed. 1. Image the patient in supine position with the neck slightly hyperextended to elongate the airway. Use dedicated neck coil for MR imaging to improve the spatial resolution and signal-to-noise ratio. 2. Perform images in axial plane from the body of the mandible to the thoracic inlet. Obtain MRI images parallel to the true vocal cords and review the volumetrically acquired CT images in the same plane. Increase the upper extent to the skull base if posterior pharyngeal wall cancer is suspected to capture the entire possible extent of the tumor and to include all retropharyngeal lymph nodes. 3. Utilize a slice thickness of ≤1 mm for CT and contiguous 3 mm images for MR imaging for adequate display of the pertinent anatomical structures. Utilize the multi-planar reformations capabilities of the volumetric CT acquisition to review images in different planes. 4. Use a field-of-view of ≤16  cm through the neck. Additional reconstructions of the CT images through the hypopharynx with a field-of-view of 10 cm are recommended to increase spatial resolution. Magnification of the images performed with the larger field-­ of-­view is not sufficient as the special resolution will stay the same. 5. Use image matrix of at least 256 × 256 for MR and 512 × 512 for CT for optimal spatial resolution. 6. Inject intravenous contrast for better tumor border delineation and detection of nodal metastatic foci. Scan in the capillary phase as tumor might not significantly enhance when scanned too early, e.g., during the arterial phase.

Fig. 15  Axial contrast-enhanced CT image performed during Valsalva maneuver reveals a tiny mass along the posterior pharyngeal wall (arrows) on the right side that would have been difficult to detect with collapsed pyriform sinus as typically seen during quite breathing

7. MRI: A minimum of non-contrasted and contrast T1-weighted images and fast spin-­ echo T2-weighted images should be performed to emphasize soft tissue detail. Other sequences may be included to better evaluate certain structures such as the intramural fat planes (see Sect. 2.2). 8. CT: Reconstruction of the images through the laryngeal cartilages in bone algorithm is helpful for detection of cartilage destruction and or sclerosis. 9. Utilize the multi-planar capabilities of both imaging modalities to facilitate the assessment of craniocaudal tumor extension. 10. Consider scanning of patient during Valsalva maneuver when small tumor is suspected to open the piriform sinuses (Fig. 15).

5  Radiologist’s Role 5.1

Pretreatment

The radiologist’s pretreatment role is multifold: detection of the subsite of origin of the primary tumor, delineation of the extent of the primary tumor, assessment of the nodal status, and

Neoplasms of the Hypopharynx and Proximal Esophagus

d­ etection of a second primary cancer. When the full extent of the primary tumor is assessed and reported, the following pertinent issues have to be emphasized as they may influence the T staging of the tumor.

5.1.1 Submucosal Spread As mentioned before, hypopharyngeal and or esophageal cancers like to grow in a submucosal fashion and, therefore, remain undetectable in a significant number of patients on clinical and endoscopic examination (Figs. 10 and 11) (Saleh et  al. 1993; Hermans 2006). Occasionally, the entire tumor is submucosal in location and difficult or impossible to biopsy endoscopically. In such cases, the radiologist may offer percutaneous biopsy under CT guidance. 5.1.2 Cartilage Involvement Cartilage involvement may be studied with CT and MR. Both modalities struggle with the problem of lack of ossification of the different cartilages in the younger patient and heterogeneous as well as variable ossification with advanced age (Yeager et al. 1982). This is in particular true for the thyroid cartilage. Since the different cartilages tend to ossify in a symmetric fashion, asymmetric attenuation might be a helpful hint for the presence of cartilage invasion and needs to be included in the pretreatment assessment of CT and or MR studies (Fatterpekar et al. 2004). The recent introduction of dual-energy CT imaging for the evaluation of head and neck cancers and cartilage invasion is showing promising results when iodine overlay technique is used in iodine enhancing tumors (Kuno et  al. 2012; Forghani et  al. 2015). The evaluation of iodine overlays has shown substantially higher specificity in detection of cartilage invasion than CT images with 96% versus 70%, respectively, at same sensitivity (Kuno et al. 2012). Dual energy CT is in particular helpful in distinguishing areas of non-ossification as anatomical variation versus tumor invasion. On the contrary, dual energy CT is of very limited use assessing ossified parts of a cartilage, as this technique cannot differentiate iodine from calcium (Kuno et al. 2012; Forghani et al. 2015). In comparison to MRI, dual energy

131

CT has been shown to have higher specificity than MRI for detection of laryngeal cartilage invasion with 98% versus 84%, respectively, with a trend of MRI to higher sensitivity than dual energy CT (Kuno et al. 2018). The role of CT and MR in detection of cartilage involvement has been studied extensively in the past decades as it has been shown that non-­ removal of an involved cartilage carries a risk of 50–60% of leaving tumor behind. The consensus is that MR is the most sensitive imaging modality in detection of cartilage involvement; however, it suffers from low specificity as inflammation, edema, and sclerosis can show similar MR findings as tumor invasion (Wenig et al. 1995; Becker et al. 1995, 2008; Castelijns et al. 1988). For CT, the reported sensitivity is lower but the overall specificity is higher than for MR (Bukachevsky 1992). In addition, the specificity of both imaging modalities depends upon the cartilage type. The thyroid cartilage has the lowest specificity due to its variable ossification and the arytenoid cartilage has the highest. Therefore, the diagnosis of thyroid cartilage involvement by tumor should only be made with caution with CT and MRI.  Fortunately, dual energy CT with iodine overlay technique has shown superior specificity in the evaluation of the non-ossified areas of thyroid cartilage for tumor invasion (Kuno et  al. 2012; Forghani et  al. 2015). Therefore, dual energy CT and MRI in combination might be the best approach for most reliable detection of cartilage invasion with dual energy CT being superior in areas of lysis and MRI in areas of sclerosis. Cartilage involvement can manifest as cartilage sclerosis (Fig. 16), cortical erosions or lysis (Fig. 17) and bone marrow replacement (Fig. 18). Cartilage sclerosis is discernible as increased density on CT and decreased attenuation on all MR sequences with lack of enhancement following contrast administration (Fig. 16). It has been shown to be the most sensitive criterion for cartilage invasion; however, it often corresponds to “benign” reactive inflammation secondary to the adjacent tumor or occasionally to a normal anatomical variation as reported for the arytenoid cartilage (Schmalfuss et  al. 2000; Becker et  al. 1997; Munoz et  al. 1993). Therefore, it is not

I. M. Schmalfuss

132

a

Fig. 16 (a, b) Axial contrast-enhanced CT image displayed in soft tissue window (a) shows diffuse fullness of the esophageal verge (ev) with subtle left lateral extension (black arrow in (a)). There are focal areas of cortical dehiscence (arrowheads in (a)) in the posterior cricoid cartilage with associated soft tissue attenuation (asterisk in (a)) when compared to the fatty replacement of the anterior cricoid

a

b

cartilage bilaterally (white arrows in (a)). The constellation of findings is concerning for cricoid cartilage invasion by tumor. The axial contrast-enhanced T1-weighted image (b) confirms invasion of the left cricoid cartilage (arrowheads in (b)) with marked decrease of the anterior posterior thickness with preservation of the normal appearance of the right cricoid cartilage (arrows in (b))

b

Fig. 17 (a, b) Axial contrast-enhanced CT images displayed in soft tissue (a) and bone (b) windows show severe foreshortening of the posterior thyroid cartilage

caused by extensive pyriform sinus cancer (T) involvement on the right, when compared to its normal length on the left (arrow)

surprising that the positive predictive value of cartilage sclerosis for tumor invasion has been reported to be only about 50%. The use of dual energy CT with iodine overlays is not helpful in this situation, as it cannot distinguish iodine from calcium (Kuno et al. 2012).

ment plan based on outcome measures. Beside cartilage invasion, tumor volume and amount of disease at the pyriform sinus apex have been identified to play critical prognostic factors in patients with T1 and T2 stage pyriform sinus cancer (Pameijer et al. 1998). A tumor volume of 6.5 ml demarcated the threshold between favorable and non-favorable to reach control at the primary site with radiation therapy alone. Similar threshold criteria were also detected in regard to pyriform sinus apex involvement. A tumor

5.1.3 Tumor Volume In the past decade, focus has been placed on stratifying patients in high- or low-risk groups to help with the selection of the most optimal treat-

Neoplasms of the Hypopharynx and Proximal Esophagus

Fig. 18  Axial contrast-enhanced CT image through the arytenoid cartilage level demonstrates replacement of the normally seen fatty bone marrow (white arrow) by soft tissue density (black arrow) caused by bone marrow replacement of this portion of the thyroid cartilage by a pyriform sinus cancer on the right (T). Note the abrupt cutoff of the submucosal fat planes caused by the tumor on the right (white arrowhead) when compared to their normal appearance on the left (black arrowheads). The tumor also wraps around the posterior border of the thyroid cartilage to involve the strap muscles (asterisks)

d­ iameter of 3 mm with homogeneous contrast enhancement and/or an adenoid without contrast-enhancing septa. This pattern was present in 66/144 cases (45.8%) in a study by King et  al. (2018). In BH2 the mucosa measures >3  mm in thickness, with a deeper band of greater contrast enhancement (described as a deep mucosal white line) compared to the superficial layer of the wall; and/or an adenoidal mass with contrast-enhancing septa separated by linear bands of minimally

Nasopharyngeal Neoplasms

197

a

b

c

d

e

f

g Fig. 4  Variations in the normal appearance of the nasopharynx. (a, b) Axial T1W post-contrast images. Pencil thin nasopharyngeal mucosa (broad arrow) and the pharyngeal recesses (fossae of Rosenmuller; thin arrows). Remnant adenoidal tags (b) may be seen. (c, d) Enlarged nasopharyngeal adenoid in a normal patient. Axial T1W unenhanced image (c) shows thickened nasopharyngeal mucosa. On the post-contrast images (d), preserved deep

mucosal line and symmetric striations in the nasopharyngeal adenoid. (e) Axial T1W post-contrast image showing thickened nasopharyngeal mucosa and enlarged adenoid. Underlying deep mucosal line preserved. Biopsy showed benign hyperplasia. BH II. (f, g) Axial T2W fat-saturated images showing cysts in the midline nasopharyngeal adenoid (f) and right fossa of Rosenmuller (arrow, g)

J. Goh and A. Karandikar

198 Table 1  Features of benign hyperplasia of nasopharyngeal mucosa Benign hyperplasia (BH) of nasopharyngeal mucosa BH1 (46%) BH2 (54%) Mucosal >3 mm >3 mm thickness Enhancement Homogeneous Deep layer enhances more than superficial layer Adenoid No enhancing Enhancing septa if septa if present present

a

contrast-­ enhancing tissue (striped appearance). This can be seen in up to 54% of cases (Table 1).

3.3  L  ocal Extension and Patterns of Spread On CT, NPC is isodense to muscle in unenhanced images, with mild to moderate enhancement after contrast administration (Fig. 5a). On MRI, NPC

b

c Fig. 5  CT vs. MRI. (a) CT image of patient with NPC (asterisk) showing nasopharyngeal mass. Extent of involvement of adjacent soft tissues is difficult to determine. (b, c) Axial T1W contrast-enhanced and T2W

images. Tumour shows intermediate enhancement, less than normal mucosa. Tumour extent and prevertebral muscle involvement are also much better seen on MRI

Nasopharyngeal Neoplasms

is usually iso- to mildly hypointense to muscle in unenhanced T1W images. After contrast administration, the tumour shows less enhancement compared to normal mucosa (Fig. 5b). It is mildly hyperintense to muscle and less bright than mucosa on T2W images (Fig. 5c).

Fig. 6  T1 NPC.  T1W post-contrast axial image. Tumour fills right FOR and is confined to the nasopharyngeal cavity. Tumour enhancement is less than normal mucosa (thin arrow). Note the disruption of the deep mucosal line (arrow)

a

199

NPC usually begins in the FOR (Figs. 2b and 6) and extends into the submucosal tissues with early infiltration of the deep cervical spaces. It may also arise on the posterior nasopharyngeal wall (Fig.  7). Subsequent involvement of the Eustachian tube orifice and/or levator veli palatini muscle lead to serous otitis media, a common finding in patients with NPC.  Beyond the nasopharynx, NPC spreads in various directions and patterns as described by several authors (Sham et  al. 1991a) although the tumour has a propensity for lateral, posterior and superior spread. An important mimic of NPC is spread of infection from necrotising external otitis (NEO) and skull base osteomyelitis (SBOM). These entities cause soft tissue swelling of the ipsilateral nasopharynx that can mimic a mass (Figs. 8 and 9). SBOM arises secondary to spread of infection from NEO or from the sphenoid sinus. Patients are often immunocompromised (e.g. diabetes, HIV, post-transplant patients). It can also be seen in post-radiotherapy patients. The most common bacterial organisms are Pseudomonas Aeruginosa, S. Aureus, and Streptococcus. Fungal organisms such as Aspergillus can also cause SBOM.

b

Fig. 7  NPC arising from posterior nasopharyngeal wall and replacing the adenoid (a). No FOR involvement or prevertebral muscle involvement. T1 tumour (b) accompanying fused PET-CT image

J. Goh and A. Karandikar

200

a

c

b

d

Fig. 8  Sixty-seven-year-old male with DM, right ear pain and discharge. (a) Unenhanced CT showing right FOR fullness. (b) T2 MR brain also shows right FOR swelling. However, T2W signal is similar to nasal mucosa, unlike

NPC. (c, d) T1W and T2W images show architectural preservation. IV contrast not given as patient had renal impairment. Inflammatory markers were raised. Biopsy— no malignancy. NEO. This can mimic NPC

On MRI, features of NEO or sphenoid sinusitis may be found. The bone of the skull base shows increased T2-weighted signal, coupled with enhancement post-contrast administration and low ADC values. Extensive soft tissue involvement can occur. Notably, there is architectural preservation. This is unlike NPC and other malignancies, where architectural distortion is the norm. Abscess formation may also occur. Neural involvement is often seen, presenting with cranial nerve palsy (commonly VII and V nerves). Intracranial extension can occur and cause abscess formation, meningitis and cavernous sinus thrombosis. Thrombosis of the internal carotid artery and internal jugular vein can also occur (Goh et al. 2017).

sphenopalatine foramen. This is recognised by loss of the normal fatty signal in the PPF on unenhanced T1W images (Fig. 10). PPF infiltration is seen in up to 15% of patients at presentation (Luo et al. 1998). The PPF is an important crossroads in the deep face as the tumour may show retrograde PNTS along the maxillary and vidian nerves, extend laterally into the pterygomaxillary fissure and masticator space, or invade orbital apex via the inferior orbital fissure, and subsequently reach the cranial cavity via the superior orbital fissure (Fig. 11).

3.3.1  Anterior Spread NPC can spread anteriorly into the nasal cavity, and from here, it can invade the PPF via the

3.3.2  Lateral Spread This is the most common route of spread, due to tumour extension through the sinus of Morgagni to the parapharyngeal space. Parapharyngeal spread is associated with increased risk of spread to the first echelon nodes (retropharyngeal, level II) (Ai et  al. 2018b).

Nasopharyngeal Neoplasms

201

a

b

c

d

e Fig. 9  Patient with NEO. (a) T1W image shows left nasopharyngeal swelling with intermediate T2W signal (b). (c) Post-contrast, the swelling showed enhancement

greater than nasal mucosa, unlike NPC. (d) Post-contrast image at a different level showed no architectural distortion. (e) Abscess in a different patient

J. Goh and A. Karandikar

202

a

b

Fig. 10  NPC.  Axial T1W pre- (a) and coronal post-­ clivus (double) and sphenoid sinus (starred). Right paracontrast (b) images show tumour infiltrating right ptery- pharyngeal space invasion also seen. T3 tumour gopalatine fossa (arrow), pterygoid process (chevron),

Parapharyngeal extension may be recognised by direct lateral tumour extension into the parapharyngeal fat by the following signs: (a) through the pharyngobasilar fascial plane, and the intrapharyngeal portions of the levator palatini muscle and Eustachian tube in the upper nasopharynx and (b) through the pharyngeal constrictors in the lower nasopharynx (King et  al. 2000b). The normal fat signal on T1W unenhanced images will be lost and the tumour enhances after contrast administration (Fig. 10). However, care must be taken to distinguish the normal enhancing pterygoid venous plexus from tumour. Subsequent lateral extension from the parapharyngeal space to the masticator space can then occur, with tumour invading the pterygoid muscles (Fig.  12). In advanced cases, tumour can extend beyond the masticator space (e.g. into the parotid space). Masticator space involvement can be recognised by architectural distortion of the muscle fibres and replacement of the normal muscle signal with tumour signal (Fig. 12a). Perineural infiltration of the mandibular division of the trigeminal nerve (V3, the nerve of the masticator space) can also occur (Fig.  12b); denervation

atrophy of the muscles of mastication may be present.

3.3.3  Posterior Spread NPC can infiltrate the retropharyngeal and prevertebral spaces, distorting the fibres of the longus capitis muscles. Prevertebral muscle invasion has been identified as a negative prognostic factor and is associated with a higher risk of distant metastases (Ai et al. 2018a; Lee et al. 2008; Feng et al. 2006). In more advanced cases, tumour infiltrates the clivus (Figs. 10a and 12a) and upper cervical vertebra, and may extend into the posterior cranial fossa. Posterolateral spread into the carotid space can affect the ipsilateral cranial nerves IX–XII (CN IX–XII); however, CN XII involvement is more often due to extranodal extension from the retropharyngeal nodes rather than direct infiltration by the primary tumour. CN X dysfunction may produce fatty infiltration and wasting of the superior constrictor muscle, soft palate muscles and vocal cord paralysis. CN XI dysfunction produces trapezius muscle wasting and atrophy, while CN XII dysfunction causes fatty infiltration and wasting of the hemitongue, with

Nasopharyngeal Neoplasms

a

203

b

c Fig. 11  Orbital invasion. Forty-three-year-old man presenting with left optic neuropathy. Axial pre- and post-­ contrast (a, b) and coronal post-contrast images (c) show an enhancing mass invading the left inferior orbital fissure and orbital apex (arrows), compressing the optic nerve.

Primary was in the nasopharynx (not shown). Histology showed non-keratinizing squamous cell carcinoma. T4 disease. (Images courtesy of Dr. Yu Wai Yung, National Neuroscience Institute, Singapore)

posterior displacement (Harnsberger and Dillon 1985; Chong and Fan 1998).

oncologist, for treatment planning purposes. NPC may also spread into the intracranial cavity by various routes. The foramen lacerum is one route, whereupon tumour may gain access to the carotid canal and cavernous sinus (Silver et  al. 1983) (Fig. 13). A second route of spread is by direct skull base erosion (Sham et  al. 1991b). Alternatively, retrograde perineural spread via

3.3.4  Superior Spread Superior extension to involve the sphenoid sinus can occur; the proximity of the tumour to the optic chiasm and pituitary gland (Fig. 13d) is an important finding to convey to the radiation

J. Goh and A. Karandikar

204

a

b

Fig. 12  Perineural spread along V3. (a) Invasion of left medial pterygoid muscle (asterisk) with architectural distortion (from muscle fibre infiltration) compared to normal right pterygoids (starred). Note also left V3 perineural

infiltration (arrow). Normal right V3, compound (arrow); clival (chevron) and left petrous apex invasion (thin arrow). (b) Right V3 perineural spread (arrow) in another patient

b

a

d c Fig. 13 (a–c) T4 NPC with right cavernous sinus invasion. The patient had right CN III and VI palsy. (a, b) T1W axial post-contrast and (c) unenhanced coronal T1W images show a nasopharyngeal mass (asterisk) invading the right foramen lacerum (thin arrow) and cavernous

sinus (block arrow). The cavernous segment of the right ICA is encased. (d) T4 NPC in a different patient. Tumour invades the sphenoid sinus and invades the right side of the pituitary fossa (double arrow)

Nasopharyngeal Neoplasms

the mandibular division of the trigeminal nerve after masticator space invasion can occur (Fig. 12b).

3.3.5  O  rbital and Paranasal Sinus Involvement Paranasal sinus involvement occurs as a result of direct extension. The ethmoids can be affected after nasal invasion, while sphenoid sinus invasion occurs due to superior spread from the nasopharyngeal cavity; up to 25% of patients show direct sphenoid sinus invasion (Chong and Fan 1993). Maxillary sinus invasion may occur after nasal or infratemporal maxillary wall erosion (seen in 6% of cases) (Huang et al. 2005). Here, one can see loss of contiguity of the bony sinus walls, with or without intrasinus extension; sclerosis and cortical irregularity may be seen on CT. On CT, reactive mucosal thickening is distinguished from tumour by its hypodense ­appearance and the lack of enhancement relative to tumour. On MRI, mucosal thickening shows uniform hyperintense signal on T2W images, distinguishing it from the intermediate signal of tumour. Post-contrast administration, mucosal thickening shows little to no enhancement, compared to enhancing tumour. Orbital invasion and ophthalmic involvement are uncommon findings in NPC, with reported incidences ranging from 3.2% (Hsu and Wang 2004) to 7.6% (Wong et  al. 2017). Invasion occurs most commonly from direct extension from the PPF through the inferior orbital fissure (Fig.  12) (Wong et  al. 2001). Direct extension from the ethmoid and/or sphenoid sinus through the lamina papyracea is the second most common pathway for orbital invasion (Luo et  al. 1998). Other possible, albeit less common, pathways include the superior orbital fissure via the cavernous sinus; or direct invasion through the orbital floor from the maxillary sinus (Hsu and Wang 2004). In Wong et al.’s study (Wong et al. 2001), orbital invasion was seen in almost half of the patients with ophthalmic involvement from NPC. The most common presenting symptom in newly diagnosed patients may vary; in Wong’s study blurred vision was the most common symptom (75%), followed by facial numbness, while in Hsu’s study, the most common presenting

205

complaints were proptosis (92.3%) and diplopia (69.2%). This difference may have been due to the different methodologies used in these studies.

3.3.6  Perineural Tumour Spread and Intracranial Extension Intracranial extension denotes T4 disease. Features of intracranial extension include meningeal involvement, cavernous sinus invasion, masses in the middle and/or posterior cranial fossa, and perineural tumour spread (PNTS). Meningeal involvement can be recognised by the presence of nodular thickening of the meninges. Diffuse smooth meningeal thickening may also be present but can be reactive; neoplastic invasion should be considered if the meningeal thickening is >5 mm. PNTS may produce symptoms that may present before the primary tumour is diagnosed; hypoaesthesia was found in 50% of patients after PPF invasion (Chong et al. 1995a). However, up to 40% of patients with PNTS may be asymptomatic. Failure to detect PNTS will result in early recurrence as this will not be addressed during radiotherapy. Sites distant to the primary tumour, e.g. the cavernous sinus (Figs.  12b and 13a, c), may become involved. Best shown on post-contrast T1W images, PNTS is seen as nodular thickening and abnormal enhancement of the central nerve, with bony remodelling and expansion of the canals these nerves travel in. Cavernous sinus expansion and replacement of CSF in Meckel’s cave may also be seen (Maroldi et  al. 2008). Apart from the mandibular division of the trigeminal nerve, other possible routes of PNTS into the cranial cavity may include spread along the maxillary division of the trigeminal nerve and its branches; the vidian nerve and the hypoglossal nerve. Maxillary and mandibular nerve involvement is best demonstrated on coronal T1W post-­ contrast MRI. 3.3.7  Carotid Artery Encasement Carotid artery encasement is defined as tumour tissue surrounding more than 270° of the vessel circumference. The presence of carotid artery encasement renders the patient inoperable if salvage surgery is contemplated. In addition,

J. Goh and A. Karandikar

206

vascular invasion and potential carotid blow-out post-radiotherapy are possible complications.

3.4  T-Staging Accurate demonstration of the primary tumour and its anatomical extent is essential for staging and treatment planning. MRI is the imaging technique of choice for the reasons described above. CT has poorer soft tissue contrast resolution and may be unable to separate the primary tumour from retropharyngeal nodes (RPNs). MRI is also superior to CT in demonstrating marrow infiltration of the clivus. Whilst PET/CT is able to identify the primary tumour, there are limitations. PET/CT may underestimate the tumour extent in 27–50% of cases, and may even miss small volume tumours (King et  al. 2008). Ng et al. also found that while PET/CT was able to alter the T-staging in NPC, discrepancies in mapping the tumour boundaries were seen between PET/CT and MRI.  These occurred in bony structures such as the skull base (16% of cases), intracranial region (14%) and parapharyngeal soft tissue spaces (19%) (Ng et al. 2009). A meta-analysis by Vellayappan et al. also found that PET/CT did not have good accuracy for T-staging (Vellayappan et al. 2014). According to the 8th edition of the AJCC Cancer Staging Manual (Table 2), T1 lesions are now defined as being confined to the nasopharynx, or may extend to the oropharynx and/or nasal cavity, without parapharyngeal extension. If the tumour is confined to the nasopharynx, it

can be recognised by the mucosal thickening and asymmetry produced (Figs.  6 and 7). However, 11.9% of T1 NPC cases may show symmetry (King et al. 2018). BH can be mistaken for early NPC although the features described earlier help differentiate the two entities. Disruption of the deep mucosal line is also a helpful sign in distinguishing early NPC from BH (King et al. 2018). T2 lesions are now classified as lesions extending to the parapharyngeal space and/or adjacent soft tissue involvement (medial pterygoid, lateral pterygoid, prevertebral muscles) (Fig. 14b). Masticator space involvement, previously staged as T4, is now staged as T2 disease. T3 tumours are characterised by paranasal sinus and bony involvement (the skull base, pterygoid structures and cervical vertebra) (Fig. 10a). Bony involvement may take the form of bone marrow infiltration and/or cortical erosion. Bone marrow infiltration is best shown on MRI, as seen by the loss of the normal hyperintense T1W fatty marrow signal on unenhanced images, with enhancement post-contrast administration. While clival marrow can have a heterogeneous appearance, the normal clivus should not appear less intense than the pons on T1-weighted images (Chong and Fan 1993; Kimura et al. 1990). T4 lesions are generally very extensive lesions, as seen when NPC invades beyond the lateral margin of the lateral pterygoid muscle or shows direct invasion of the orbit and parotid gland (Fig.  11). The presence of intracranial extension also upstages the tumour to T4, as does cranial nerve involvement (Figs.  12b and 13a, c). These have been described above.

Table 2 T-staging T category TX T0 T1 T2 T3 T4

T criteria Primary tumour cannot be assessed No tumour identified, but EBV-positive cervical node(s) involvement Tumour confined to nasopharynx, or extension to oropharynx and/or nasal cavity without parapharyngeal involvement Tumour with extension to parapharyngeal space, and/or adjacent soft tissue involvement (medial pterygoid, lateral pterygoid, prevertebral muscles) Tumour with infiltration of bony structures at skull base, cervical vertebra, pterygoid structures, and/ or paranasal sinuses Tumour with intracranial extension, involvement of cranial nerves, hypopharynx, orbit, parotid gland, and/or extensive soft tissue infiltration beyond the lateral surface of the lateral pterygoid muscle

Nasopharyngeal Neoplasms

a

207

b

Fig. 14 (a) Blue line—T2; yellow line—T4 (TNM 7th edition). White line—T2; red line—T4 (TNM 8th edition). (b) T2 NPC. Mass in left FOR invading left longus capitis muscle

Hypopharyngeal invasion, which indicates very extensive disease and upstages the tumour to T4, is rarely seen in current clinical practice. On CT, features suggesting bony involvement include cortical erosion and sclerosis. However, sclerosis is non-specific and may be due to either reactive change from irritation by tumour or direct tumour spread. Also, bony sclerosis and thickening of the pterygoids and paranasal sinus walls from chronic sinusitis are commonly seen. However, if sclerosis is seen in a bony structure adjacent to the primary tumour, in the absence of adjacent inflammatory sinus disease, the radiologist should be alert to the possibility of tumour involvement. Despite the changes in the T-staging, it is felt that further refinement to the T-staging is required as there appears to be a lack of separation in local control for T2 and T3 tumours, especially in the IMRT era. Tang et al. have suggested either merging the T2 and T3 categories, or further identifying distinguishing features specific to these two categories (Tang et al. 2017).

4  Metastatic Disease Metastatic disease may occur regionally to the draining lymph nodes (nodal metastasis) or to distant structures (distant or systemic metastasis). These are discussed separately.

4.1  Nodal Metastases Nodal metastases are commonly seen in NPC, with up to 85% of patients having regional nodal metastasis (Ho et  al. 2012), and occurs in an orderly fashion from the upper to lower cervical nodes. Bilateral lymphadenopathy is seen in up to 80% of cases (Chong and Ong 2008). Nodal metastases are most commonly seen in the retropharyngeal nodes (RPNs) and level II nodes, which are thought to be the first echelon nodes (Liu et  al. 2006) (Fig.  15a, b). The presence of abnormal RPNs is thought to be a negative prognostic factor for disease-free survival (DFS) and distant metastasis-free survival (DMFS) in NPC (Tang et al. 2014). Although both medial and lateral RPNs exist anatomically, only the lateral RPNs are seen consistently on imaging (Chong et  al. 1995b). The lateral RPNs can be found from the level of the occipital bone down to the C3 body. Metastases can bypass the lateral RPNs, with the reported incidence ranging from 6% (King et al. 2000a) to 35% of cases (Chong et al. 1995b). The medial group of nodes may be due to intercalated nodules rather than a discrete nodal chain and thus are not seen on imaging (Smoker and Gentry 1986). Hence, any medial RPNs seen on imaging are considered abnormal. The level III, IV (Fig.  15d) and V nodes are thought to represent the second echelon of nodes,

J. Goh and A. Karandikar

208

a

b

c

d

Fig. 15  Lymphadenopathy from NPC. (a) Bilateral retropharyngeal nodes 6  cm in greatest dimension; and/or (b) extension below the caudal border of the cricoid (Fig. 15d). The previous designations of N3a and N3b have been removed (AJCC 8th edition).

4.1.1  Imaging Evaluation When evaluating the lymph nodes for nodal disease, several criteria are used. The most commonly used is nodal size. While the shortest transaxial diameter is measured on imaging studies, it must be remembered that in clinical evaluation, using the AJCC classification, the maximum dimension is assessed. The upper normal limits (shortest transaxial diameter) are 11 mm for level II nodes and 10  mm for the level III–VII nodal groups, with a specificity and sensitivity of 82% (Van den Brekel et al. 1990). A minimum transaxial diameter of 5  mm has been proposed for RPNs (King et al. 2000a) although other authors have suggested that 6 mm may be more accurate (Zhang et al. 2010). However, no one size criterion can actually correctly predict lymph node metastasis. Moreover, in NPC, in imaging studies, metastatic lymphadenopathy often presents as clustered and/or confluent nodes (Figs.  15c and 16), a feature that is not included in the current AJCC staging system. The size criterion for these clustered and/or confluent nodes has not

been established. In a paper by Ai et al. (2018a, b), it was suggested that using the maximum unidimensional measurement of not only single, but also confluent and contiguous nodes was the best method of identifying the maximal dimension for nodes. In addition, the authors found that this method best correlated to nodal volume and was an independent predictor of survival. A threshold of >5.6 cm was also thought to be a better predictor of survival in Stage IVb disease. Nodal volume is thought to be a better predictor of survival (Ai et  al. 2017; Yuan et  al. 2017) but requires painstaking and laborious measurements to obtain. Other criteria that may suggest metastatic lymphadenopathy from NPC include the following: –– shape (rounded nodes with loss of their fatty hila) –– a cluster of ≥3 nodes with minimum axial diameters of 9–10 mm (level II) and 8–9 mm (levels III–VII) if these are in the drainage area of the primary tumour –– necrosis Necrosis is a reliable feature of metastatic disease and can be found in 20–44% of NPC patients (Lan et  al. 2015) (Fig.  16). Necrosis is also an important prognostic factor in overall survival as shown in Lan et al.’s study (Lan et al. 2015), where reduced overall survival, locoregional DFS and DMFS were found in patients with N1–N3 disease who had necrotic nodes. Conversely, Ting et  al. (2017) found the presence of necrosis only had a significant impact on survival in N2 disease.

J. Goh and A. Karandikar

210

a

b

Fig. 16  Necrotic nodes in NPC patient. Post-contrast T1-weighted (a) and T2-weighted (b) images show matted nodes, extranodal extension and necrosis (starred)

Extranodal extension (ENE; previously termed extracapsular spread) carries a grave prognostic significance and should be highlighted to the clinician. Features suggesting ENE include loss or irregularity of the node margin, stranding in the perinodal fat, and invasion of adjacent structures, e.g. internal or common carotid artery, prevertebral muscles, sternocleidomastoid muscle (Figs. 15c and 16). If a clear plane separates the node from an adjacent soft tissue structure then, it is unlikely that ENE and invasion of said structure has occurred. Ai et al. (2019) have suggested that ENE involving skin, muscle and the salivary glands was an independent predictor or poor outcome and may be a new criterion for determining N3 disease. Unfortunately, these morphologic criteria do not always reflect the presence of metastases. Using the most common criterion (short axis diameter, 10 mm), the sensitivity and specificity of CT and MRI vary from 65–87% to 47–94%, respectively (Sigg et al. 2003; Curtin et al. 1998; Al-Ibraheem et al. 2009; Comoretto et al. 2008). In this regard, 18F-FDG PET/CT imaging better demonstrates nodal metastases (Adams et  al. 1998; Chang et al. 2005) and has been shown to improve staging in 15–20% of patients with a sensitivity and specificity of 74–94 and 82–100%

(Al-Ibraheem et  al. 2009). FDG PET/CT has been found to have a sensitivity of 97–100% and a specificity of 73–97% for assessing cervical lymph nodes, and it is more accurate in the assessment of lower cervical metastases (Mohandas et al. 2014). However, PET/CT is less accurate in identifying metastatic RPNs; Ng et al. found MRI was superior to PET/CT in identifying RPNs, with a discrepancy of 13% between the two modalities (Ng et al. 2009). This can be due to blooming artefact making it difficult to distinguish RPNs from the primary tumour, and the inferior contrast resolution of PET/CT.  The use of contrast-enhanced PET/CT may improve detection of RPNs (Mohandas et al. 2014). PET/ MRI is currently still being investigated as a clinical staging tool although Chan et  al. (2018) found that PET/MRI had a sensitivity of 99.5% in N staging, compared to head and neck MRI alone (94.2%) and PET/CT (90.9%). PET/MRI was also found to be particularly useful for distinguishing retropharyngeal nodal metastasis from adjacent nasopharyngeal tumours although this is likely due to the MRI component of the examination. A caveat related to RPNs is the superior cervical ganglion, which can be mistaken for metastatic lymphadenopathy in the RPNs, particularly

Nasopharyngeal Neoplasms

in the post-treatment setting. The ganglion can be recognised from its position; ovoid shape; greater enhancement and higher ADC values compared to metastatic RPNs and the presence of a central hypointensity on T2-weighted and T1-weighted post-contrast images (described as a black dot and line on axial and coronal images, respectively) (Loke et al. 2016; Lee et al. 2016; Yokota et al. 2018) (Fig. 17).

a

c Fig. 17  Prominent bilateral superior cervical ganglia (SCG). T1W unenhanced (a) and T1W post-contrast (b) images show intensely enhancing ganglion with a central black dot. (c) T2W image without fat saturation shows the

211

4.2  Distant Metastases Five to fifteen percent of untreated NPC patients will have distant metastases at diagnosis (Chang et  al. 2005; Teo et  al. 1996; Sham and Choy 1990). Teo et  al.’s study found that the 1-year mortality in patients with metastatic disease found at diagnosis may be up to 90% and that the likelihood of metastasis increases with higher T

b

d hyperintense signal of the ganglion, again with a central black dot. (d) Coronal T1W post-contrast image shows ovoid appearance with tapered ends (arrowed). The central black dot is now seen as a central black line

J. Goh and A. Karandikar

212

and N stages; approximately 26% (247 out of 945 patients) developed metastases within 3 years of radiotherapy. The most common sites are the

bones (Fig. 18a, g), thorax (pulmonary deposits and mediastinal nodes (Fig. 18b), liver (Fig. 18h), distant lymph nodes (non-mediastinal) and other

b

a

c

e Fig. 18 (a) Bone scan showing multiple bony metastases from NPC. (b) Mediastinal lymphadenopathy from NPC in a different patient. Lymph node involvement below the clavicle is deemed distant metastasis. (c–h) Middle-aged

d

f male with NPC. MRI (c, d) shows extensive locoregional and intracranial extension. PET images show uptake in primary and neck nodes (e, f) and metastases in the bone and liver (g, h)

Nasopharyngeal Neoplasms

213

h

g Fig. 18 (continued)

a

b

Fig. 19  Unusual thyroid gland metastasis from NPC. US (a) shows an irregular markedly hypoechoic nodule right lower pole thyroid nodule invading the capsule. Coronal

post-contrast image (b) shows the nodule (arrowed). (b) Histology showed undifferentiated carcinoma

Table 4 M-staging

However, this requires the use of multiple imaging modalities to identify metastatic disease and has a low sensitivity (32.8%) (Liu et  al. 2007). In the same study, the authors found that 18FFDG PET had a sensitivity of 82%. Yen et  al. (2005) also found that conventional workup missed ­metastatic disease in 12.8% of patients. Chang et al.’s study also found that the use of 18FFDG PET was superior to conventional workup, identifying all 14/95 patients with metastases, whereas conventional workup only identified four patients (Chang et al. 2005). A meta-analysis of three PET and five PET/ CT studies by Chang et al. (2013) also showed that the pooled sensitivity, specificity, positive

M category M0 M1

M criteria No distant metastasis Distant metastasis

distant sites (e.g. thyroid; Fig. 19a, b). M-staging remains unchanged in the current 8th AJCC edition; M0 means there is no metastasis, while M1 means metastatic disease is present (Table 4).

4.2.1  Metastatic Workup Traditionally, metastatic workup has consisted of CXRs, liver ultrasound and bone scintigraphy.

J. Goh and A. Karandikar

214

likelihood ratio, and negative likelihood ratio of FDG-PET or PET/CT were 0.83, 0.97, 23.38, and 0.19, respectively. Whole-­body 18FFDG PET/CT is now the preferred modality for M-staging; the NCCN guidelines also state that contrast-enhanced CT thorax or PET/CT should be performed for imaging metastases (NCCN guidelines).

5  Staging and Treatment Based on the various TNM features, NPC patients are staged from Stage 0 to Stage IV (Table  5). Several features to note are as follows: (a) Stage I and Stage II NPC are considered early disease, while Stage III and Stage IV NPC are considered as advanced disease. (b) In Stage II disease, there is no distant metastasis. The tumour may be in the nasopharynx only (T1) but has nodal metastasis (N1); or spread beyond the nasopharynx with or without nodal spread (N0, N1). (c) T3 disease indicates a patient has at least Stage III disease. (d) T4 disease indicates a patient has at least Stage IV disease. (e) N2 disease indicates a patient has at least Stage III disease. (f) M1 disease places the patient at stage IVB.

Table 5  NPC staging Stage 0 I II

III

IV A

IV B Recurrent

T Tis T1 T1/T0 T2 T2 T1/T0 N2 M0 T2 N2 M0 T3 N0 M0 T3 N1 M0 T3 N2 M0 T4 T4 T4 Any T Any T

N N0 N0 N0 N0 N1 N2 N2 N0 N1 N2 N0 N1 N2 N3 Any N

M M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M0 M1

Correct staging is important as it allows the clinician to determine the best which treatment modality/ies for the patient. A detailed discussion of the treatment options is beyond the scope of this chapter. However, the following points should be noted. Radiotherapy is the cornerstone of treatment in NPC. The current standard of care is intensity-­ modulated radiotherapy (IMRT). IMRT is able to cover the irregular anatomy of the nasopharynx and adjacent structures, achieving good results while reducing morbidity from radiation damage to adjacent normal structures, which is of paramount importance. However, the tumour size in very large tumours located close to vital structures imposes a limitation on the full dose delivered. For early stage NPC (Stages I, II), definitive radiotherapy (RT) is used. The RT dose administered generally is 66–70  Gy over 6–7  weeks (National Comprehensive Cancer Network (NCCN) 2019). Advanced NPC (Stages III, IV) is treated with concurrent chemoradiotherapy (CRT). Cisplatin and 5-FU are the agents commonly used although newer agents are now available; cisplatin is often used as a single agent. In recent years, studies have suggested improved control in Stage 2 patients by combining chemotherapy with RT. However, this needs further validation (Sze et  al. 2014, Lee et  al. 2012). Also, induction chemotherapy followed by concurrent CRT has also been considered as a treatment option, with some reports showing improved disease control in advanced stage NPC (Tan et al. 2018).

6  Post-treatment Changes and Follow-Up 6.1  Post-treatment Changes The goal of post-treatment imaging is to (a) demonstrate that treatment has been successful and (b) identify residual or recurrent disease post CRT with a view to salvage surgery where amenable. The timing of post-treatment imaging can be difficult. There is a need to balance resolution of reactive changes from treatment, and identifying

Nasopharyngeal Neoplasms

residual disease early so that timely intervention may be employed. While the NCCN guidelines have suggested that post-treatment imaging be performed 6–8  weeks after therapy, performing post-treatment imaging at this time often results in diagnostic difficulty as the inflammatory oedema and soft tissue swelling caused by treatment is still resolving. For NPC, most centres will perform the first post-treatment evaluation at 8–16 weeks (Lee et al. 2019), although in our practice the first baseline post-treatment imaging is usually performed at 12  weeks (3 months) post-therapy to (a) confirm resolution of the primary tumour; (b) identify residual disease and direct biopsy; and (c) determine if the patient will benefit from salvage surgery.

215

Performing the baseline post-treatment imaging study at 16  weeks (4  months) may more accurately differentiate post-treatment images from residual tumour. However, this needs to be studied in detail as the additional 4 weeks wait may negatively impact salvage treatment. MRI is the imaging modality of choice. If cost is not an issue and facilities are available, PET-CT is also performed as the results are complementary with MRI.  Post-treatment MRI evaluation can be challenging, due to the superimposed granulation tissue (in the acute or subacute phase) and scarring (in the chronic phase). Granulation tissue typically shows intense T2W signal with intense enhancement and may be diffuse or polypoidal in morphology (Fig. 20). It can be par-

a

b

c

d

Fig. 20 (a, b) Axial T1W fat suppressed post Gd images. Left FOR mass seen in (a). (b) Post-treatment scan reveals diffuse intensely enhancing soft tissue (arrows). (c, d) Axial T2W in another patient, (c) right FOR mass seen

pre-treatment. (d) Post-treatment scan shows a midline polypoidal intense T2W hyperintense soft tissue bulge (arrow). Biopsy proven granulation tissue seen in both cases

J. Goh and A. Karandikar

216

ticularly difficult to distinguish scar tissue from recurrent tumour. Useful pointers that suggest residual disease include persistent soft tissue at the primary site demonstrating intermediate T2W signal and enhancement akin to the pre-treatment primary, and restricted diffusion. In recurrent disease, tumours are typically expansile with intermediate T2W hyperintensity and intermediate enhancement (Razek and King 2012). A mature scar typically shows retraction, T2W hypointensity, and no contrast enhancement (Fig. 21). DWI also helps improve diagnostic value and increases sensitivity and specificity of detecting locally recurrent NPC (Wang et al. 2018).

6.1.1  Resolution If the primary tumour has partially resolved, post-treatment imaging shows interval reduction in the bulk and T2W hyperintensity of the nasopharyngeal tumour (Fig.  22). If complete treatment response has occurred, there is complete resolution of the primary tumour. Here, intensely enhancing soft tissue thickening may be seen in the nasopharynx blunting the FOR, which is likely due to granulation tissue (Fig.  23). T2W hypointensity and retraction of the submucosal tissues, e.g. in the prevertebral muscles, can also be seen. In our experience, enhancement at certain sites, such as the skull base and PPF, can

a

b

c

d

Fig. 21 (a) Axial T2W reveals a right FOR mass. (b) 3 years post treatment scan shows soft tissue thickening in nasopharynx blunting the bilateral FOR with a midline T2W hypointense area (arrow) due to mature scar. (c) Axial T2W and (d) T1W post-contrast fat-saturated post-­

treatment scan in another patient reveals a midline T2W hypointense scar (arrows in c, d) with intermediate or reduced enhancement. It is important to evaluate both post-contrast and T2W sequences for accurate soft tissue characterisation

Nasopharyngeal Neoplasms

217

a

b

c

d

e

f

Fig. 22  Axial T2W images in a patient with left nasopharyngeal cancer (FOR, lateral wall) and an enlarged left RPN. (a, d) Pre-treatment scan. (b, e) Post induction chemotherapy. (c, f) Post completion of concurrent CRT. Note

the progressive reduction in size, volume and T2W signal of the primary tumour (arrows) and the RPN (arrows) from (a) to (c) and (d) to (f), respectively

persist for varying lengths of time post-treatment (Fig.  24). As it may be difficult to differentiate residual disease from post-treatment inflammatory changes, sequential close follow-up MRI studies or PET/CT may have to be performed. Here, DWI may be of benefit. In Hirshoren et al.’s study, where ADC values before and after treatment were calculated, they found that an ADC

threshold of 0.965  ×  10−3  mm2/s yielded 100% positive and negative predicted values and was able to differentiate pre-treatment NPC from post-treatment changes (Hirshoren et al. 2019). Similarly cervical nodes also show interval reduction in size and T2W hyperintensity. The nodes may completely regress, while in some patients they may remain enlarged, showing

J. Goh and A. Karandikar

218

a

b

c

d

Fig. 23 (a) Axial T2W and (b) post Gd axial T1W left T1 NPC localised to the nasopharynx pre-treatment. Similar sequences (in c, d) in the post-treatment scan reveals

smooth soft tissue thickening blunting the left FOR. Note is also made of partial opacification and fluid levels in both maxillary sinuses

internal heterogeneity (Fig.  25). In such cases, either follow-up MRI or correlation with PET/ CT and histology should be considered to exclude residual disease.

ation. In our experience, for very bulky tumours, evolving post-treatment changes may still be seen in this first baseline scan. A targeted biopsy will also be performed for confirmation. An exception to targeted biopsy would be suspected residual disease in inaccessible regions (e.g. foramen lacerum), whereupon sequential scans will be performed for close monitoring. Disease recurrence is defined as tumour appearing after an established disease-free interval, and may be local (in the nasopharynx, Fig. 27), regional (in the neck nodes) or distant metastases. Recurrence which is amenable to surgery is treated as such, e.g. nasopharyngectomy (either endoscopic or via a maxillary swing) or neck dissection.

6.1.2  Residual and Recurrent Tumour Residual or recurrent tumours are usually detected by endoscopy. Imaging helps with soft tissue characterisation, evaluating disease extent and differentiating post-treatment changes from tumour. Histology is needed to confirm residual or recurrent disease. Residual disease is defined as incomplete tumour resolution in the first post CRT assessment scan. Residual disease can be in the nasopharynx or in the neck nodes (Fig. 26). Tumour bulk however has to be taken into consider-

Nasopharyngeal Neoplasms

219

a

b

c

d

e

f

Fig. 24 (a) Axial T2W; (b, c, e, f) Axial post Gd T1W FS images. (d) Axial T1W unenhanced image. (a, b) Infiltrative left FOR mass extending to the left PPS and the left PPF (arrow in a). There is early invasion into the left clivus (arrow in b). (c, d) Persistent enhancement in

the left PPF (arrow, c) and abnormal marrow signal in the left clivus (arrow, d) 1 year post treatment. (e and f) Stable enhancement in left PPF (arrow, f) and left clivus (arrow, e) 4 years post-treatment

J. Goh and A. Karandikar

220

a

b

c

d

e

b

Fig. 25  Axial T2W images. (a, b) Right FOR mass (arrow in a) with a bulky intermediate T2W hyperintense right level 2 nodal mass. (c, d) Post CRT scan shows resolution of the primary nasopharyngeal mass. The right level 2 nodal mass has reduced in size and shows reduced

T2W signal, but there are intermediate T2W hyperintense areas within the nodal mass (arrow in d). (e, f) Repeat scan 3 months later shows the right level 2 nodal mass had increased in size (arrow in f) suspicious for residual disease (confirmed at histology)

Nasopharyngeal Neoplasms

221

a

b

c

d

e

f

Fig. 26 (a–d) Axial post Gd T1W FS, (e–f) PET/CT images. (a, b) reveal a left nasopharyngeal mass extending to the left nasal cavity with a metastatic left level 2 node (arrow in b). (c, d) Post CRT scan shows resolution of nasopharyngeal mass. Residual disease in left level 2

node seen (arrow in d). PET/CT shows FDG avidity in the left level 2 node (arrow in f, SUV-7), confirms the MRI findings. The FDG uptake in the tongue is due to physiological activity. No nasopharyngeal uptake seen.

It is important for radiologists to be familiar with the altered anatomy in post-treatment patients and to recognise if recurrence has occurred (Fig.  28). This is where good quality imaging, interpreted by dedicated readers, is required. Preferentially, sequential imaging studies should be performed in the same institution.

crosis was 1.04%, with a latency period of 45.57 months. Risk factors included the T-stage, total radiation dose, and anaemia. Patients can present with facial pain, swelling or deformity. The common sites affected are the skull base (clivus, floor of sphenoid sinus and sometimes petrous apex), temporal bone, mandible, maxilla and hyoid bone. Imaging reveals cortical destruction, abnormal marrow signal (hypointense on T1W, intense T2W hyperintensity with enhancement). Associated soft tissue thickening, necrotic non-enhancing areas and fistula formation may also be seen (Figs. 29 and 30). Abscesses or gas pockets may be seen if there is secondary infection and osteomyelitis (Saito et  al. 2012). The appearance may mimic tumour recurrence but

6.2  Post-treatment Changes 6.2.1  Skull Base Osteoradionecrosis Osteoradionecrosis may occur 1–3  years post-­radiotherapy (Razek and King 2012; Alhilali et  al. 2014). In Han et  al.’s study (Han et  al. 2018), the incidence of skull base osteoradione-

J. Goh and A. Karandikar

222

a

b

c

d

Fig. 27 (a–c) Axial post Gd T1W FS. (d) FDG PET/CT. (a) NPC involving bilateral FOR. (b) Scan 1  year post CRT shows no tumour. Granulation tissue is seen blunting

the bilateral FOR. Figure (c) shows a recurrent tumour in the right FOR 3 years post-treatment which shows FDG avidity (SUV7.3, d)

cortical defects, the absence of a soft tissue mass and necrotic non-enhancing areas may prompt the diagnosis of radionecrosis (Figs. 30 and 31a). A change in the imaging appearance, e.g. new tissue mass occupying a site of previously known radionecrosis, should alert the radiologist to the possibility of recurrent tumour, which should be confirmed with histology (Fig. 31b, c, e, f).

6.2.2  Radiation-Induced Brain Necrosis Temporal lobe necrosis occurs with a latent period of 1.5–13 years post-radiotherapy (Razek and King 2012). This usually affects the medial and inferior aspects of the temporal lobes as they lie within the radiation field. Temporal lobe necrosis usually presents as vasogenic oedema

Nasopharyngeal Neoplasms

a Fig. 28 (a) Axial T2W and (b) post Gd-T1W FS reveals altered anatomy in the left nasopharynx due to prior left nasopharyngectomy with flap reconstruction for a

223

b previous recurrent tumour. These images reveal a second deep recurrence (arrows in a, b) on the left side

a

b

c

d

Fig. 29 (a) Axial T1W, (b) axial T2W, (c) axial post Gd T1W FS, (d) coronal post Gd T1W FS.  Post CRT for NPC.  Diffuse intensely enhancing soft tissue thickening in the nasopharynx. Necrotic area (arrows in c, d) in the

left nasopharynx and PPS extending to inferior clivus and foramen ovale suggesting osteo- and soft tissue radionecrosis. Inflammation and necrosis seen on biopsy. Follow-up over 3 years showed no recurrent tumour

J. Goh and A. Karandikar

224

a

b

c

Fig. 30  Post Gd-T1W FS (a—axial), (b, c—coronal). Erosion of petrous apex and occipital condyle seen in (b) with necrotic area and a soft tissue defect in the left naso-

a

b

d

e

pharynx suggesting skull base and soft tissue radionecrosis. The nasopharyngeal air column extends posterolaterally beyond the margins of the nasopharynx

c

f

Fig. 31 (a–c) Post Gd T1W FS, (d–f) PET/CT images. (a) Scan 2 years post CRT revealed a defect (arrow in a) in the right skull base with soft tissue thickening showing intense enhancement. Biopsy revealed inflammation, no malignancy was found. Diagnosis: Osteoradionecrosis (b,

c) New soft tissue (arrows in b, c) filling up the defect in follow-­up scan 1 year later. This was avid on FDG PET/ CT with an SUV of 9.8 suggesting tumour recurrence on a background of radionecrosis. Biopsy revealed malignancy. Recurrent tumour

with heterogeneous rim enhancing intra-axial mass lesions. The intra-axial location helps differentiate temporal lobe necrosis from intracranial invasion by recurrent NPC, which is extra-axial in  location (Ong and Chong 2011). Intracranial abscess, metastasis or an incidental primary brain tumour such as glioma are differentials. Combined imaging modalities like MRI

with DWI, spectroscopy and PET-CT may aid differentiation (Fig. 32) (Chao et al. 2001).

6.2.3  Radiation-Induced Tumours These are rare and usually occur 4–27 years post-­ radiotherapy. Different histologic types have been reported such as meningioma, sarcoma and squamous cell cancers (Offiah and Hall 2011).

Nasopharyngeal Neoplasms

225

a

b

c

d

Fig. 32 (a) Axial T2W, (b) axial post Gd T1W FS, (c) coronal STIR, (d) FDG/PT CT. Large intra-axial heterogeneous T2W mass in the left temporal lobe with significant vasogenic oedema. Differentiation from primary

intracranial tumours may be difficult. (d) Tracer uptake by the mass is less than grey matter (SUV of 2), which suggests temporal lobe necrosis

Radiation-induced tumours present as bulky heterogeneous masses in the irradiated field with no specific imaging features (Figs.  33 and 34). Diagnosis requires histologic correlation (Saito et al. 2012; Makimoto et al. 2007). Sarcomas and squamous cell cancers arise in high-dose radia-

tion zones and usually arise around the maxillary region (palate, maxillary sinus, nasal cavity and alveolar process). Squamous cell cancers also arise in the low-dose radiation field, and may also involve peripheral sites such as the temporal bones (Razek and King 2012).

J. Goh and A. Karandikar

226

a

b

c

d

Fig. 33 Radiation-induced tumour. (a) CECT scan 5  years post-treatment reveals blunted left FOR, bony defects of the walls of the left maxillary sinus from radionecrosis, with associated left maxillary sinusitis. (b) Axial T2W, (c) axial post Gd T1W FS and (d) coronal T2W

images 10 years post-treatment. New large heterogeneous enhancing mass in the nasopharynx and nasal cavities with extension into the left maxillary sinus. Biopsy showed sarcoma

6.2.4  B  rain Stem and Spinal Cord Encephalomyelopathy The incidence of this complication was about 3% until the late twentieth century. However, with the use of more sophisticated radiotherapy techniques, it is now much less frequently seen. The anteriorly located corticospinal tracts in the brainstem and cervical cord were at highest risk for radiation-induced injury, which in turn

resulted in spastic paraparesis or quadriparesis (Ong and Chong 2011).

6.2.5  Radiation-Induced Cranial Neuropathy This is a rare complication as the cranial nerves are generally radioresistant. Recurrent tumour with perineural invasion must first be excluded. The hypoglossal nerve is most commonly

Nasopharyngeal Neoplasms

227

a

b

c

d

Fig. 34 (a, c) Post Gd T1W FS, (b) T1W and (d) PET/ CT. (a) Focal enhancing lesion (arrow, a) in the left mid/ lower neck involving skin and subcutaneous tissue in a patient 15  years post-RT for NPC.  Biopsy showed radiation-­induced dermatofibrosarcoma. (b) Post excision

of the left neck lump and flap reconstruction. (c) A new focal enhancing lesion (arrow in c) in the left supraclavicular fossa with an SUV of 9 (d) 25  years postRT.  Biopsy revealed a recurrent radiation-induced ­ myxofibrosarcoma

affected followed by the optic and abducens nerves. Multiple cranial nerve palsies can be sometimes seen in brainstem encephalopathy (Ong and Chong 2011).

evaluation can be difficult. CT angiography may be necessary for diagnosis (Chin et  al. 2005). Pseudoaneurysms are fragile leading to rupture and carotid blow out with a need for endovascular stenting or embolization (Karthik et al. 2017).

6.2.6  Vascular Complications Accelerated atherosclerosis or thrombosis of the internal or common carotid artery and internal jugular vein are known complications from radiotherapy. The imaging findings are similar to age-­ related atherosclerotic narrowing and is limited to radiation fields. A serious and potentially life threatening vascular complication is pseudoaneurysm of the internal carotid artery due to radiotherapy and background radiation-induced soft tissue necrosis (Fig.  35). These patients usually present with epistaxis post-treatment and the clinical

6.2.7  Xerostomia and Trismus Xerostomia can be seen in post-treatment patients due to parotid and submandibular gland exposure to radiation. This can lead to oral infections and dental caries. As xerostomia is a clinical finding, salivary gland injury can be recognised on imaging by glandular enhancement, followed by atrophy in late cases (Ong and Chong 2011). Trismus is caused due to radiation-induced fibrosis involving the muscles of the masticator space or due to osteoradionecrosis involving the

228

J. Goh and A. Karandikar

Fig. 35  Left ICA pseudoaneurysm post-RT (arrows). The patient subsequently developed a carotid blowout

mandibular ramus and the temporomandibular joint (Razek and King 2012).

6.2.8  Radiation-Induced Lung Disease They usually involve the apical segments of upper lobes. Acute stages manifest as consolidations or ground glass densities due to pneumonitis with eventual volume loss, scarring, traction bronchiectasis and pleural thickening (Choi et al. 2004).

7  Future Directions? With the implementation of newer techniques, investigators are looking into using these techniques to try and better analyse malignancies in various areas of the body, including the head and neck. These include DWI and tumour perfusion. Various investigators have analysed the use of DWI in NPC, in various areas including improved

tumour identification and prognostication. However, to date, the use of DWI in standard clinical practice is still not clearly defined. Similarly, tumour perfusion analysis for both the identification and prognostication of NPC is still in the investigational phase, with varying results. Successful treatment of NPC is dependent on many factors, including the tumour microcirculatory environment and level of oxygenation within tumour tissue. Reduced oxygenation is an adverse factor in the treatment of malignancies. With this goal in mind, tumour oxygenation is also being studied, using techniques such as MRI and PET/ CT, with the aim of trying to determine if the primary tumour and nodal metastases have reduced oxygenation. This would help clinicians identify tumours which require either a boost in treatment or modification of treatment regimens during therapy to maximise outcome. However, this is still an area of research and requires further study to determine its clinical applicability.

Nasopharyngeal Neoplasms

8  Other Nasopharyngeal Neoplasms and Infections 8.1  Pleomorphic Adenoma Pleomorphic adenomas are rare tumours in the nasopharynx, arising from the minor salivary glands in the pharyngeal mucosal space. The palate is the most common site of origin (Downer et  al. 2011; Waldron et  al. 1988). These benign tumours can cause an array of symptoms, including nasal obstruction, epistaxis, dysphagia and hearing loss from obstruction of the Eustachian tube. These usually have smooth surfaces (Lee et  al. 2006; Yeğin et  al. 2015). On MRI, they present as lobulated smooth lesions arising from the pharyngeal mucosal space, with intermediate T1W and hyperintense T2W signal (compared to the intermediate T2W signal of NPC). Enhancement is usually seen. Remodelling of adjacent bony structures can occur although aggressive bony changes have been described (Downer et  al. 2011). Surgical excision is the treatment of choice.

8.2  Inflammatory Pseudotumour This condition can be seen in many different locations in the body but rarely affects the nasopharynx. The aetiology is unknown. Unlike NPC, inflammatory pseudotumours are submucosal in location on MRI, with little overlying mucosal thickening, with hypointense T2W signal (compared to the brainstem) and show weak to moderate enhancement. ICA encasement occurs more frequently in inflammatory pseudotumours. Pachymeningeal enhancement has also been described. Lymphadenopathy is uncommon in this condition (Lu et al. 2010).

8.3  Lymphoma The head and neck is the second most common site for extranodal lymphoma, after the gastrointestinal tract. Waldeyer’s ring comprises the lymphoid tissue in the nasopharynx and oropharynx. In the nasopharynx itself, the

229

lymphoid tissue is found in the midline pharyngeal adenoid (the ‘pharyngeal tonsil’) and around the tubal openings of the Eustachian tubes (the tubal or Gerlach’s tonsils). The nasopharynx represents a common location for lymphoma. Diffuse large B-cell lymphoma is the most common type. Unlike NPC, it generally arises in the midline (Fig.  36b, c) although it may originate within the fossa of Rosenmuller itself (Fig.  36a). Also, the bone is not commonly involved by lymphoma, unlike NPC. While nodal enlargement can be seen in lymphoma, involvement of the parotid and submandibular nodes can be seen, nodal sites not commonly affected by NPC.  On DWI, lymphoma shows a lower ADC compared to NPC (Tan 2004; Razek and King 2012).

8.4  Adenoid Cystic Carcinoma (ACC) ACC rarely occurs in the nasopharynx, with reported incidences ranging from 0.5 to 4% (Soprani et  al. 2007). It occurs in the fourth to sixth decades of life and is a locally aggressive tumour with a propensity for perineural spread (26.9–55%) (Liu et  al. 2008; Wang et  al. 1996) and skull base invasion (67%) (Dong et al. 2015). Symptoms include epistaxis, Eustachian tube obstruction and symptoms related to skull base and cranial nerve involvement. There is also a propensity for parapharyngeal space and pterygoid muscle invasion (Fig. 37) (Dong et al. 2015). Unlike NPC, cervical lymphadenopathy is not a common feature; there is also no relation to EBV. In a case series of 10 patients with nasopharyngeal ACC imaged with MRI, tumours arising from the nasopharyngeal walls tended to have irregular outlines. A homogeneous appearance, mildly increased T2-weighted signal and strong enhancement, was seen in six patients—this pattern was seen in patients with the tubular subtype of ACC. Of the remaining four patients, three had tumours with hypointense T2-weighted signal and no enhancement, while one patient had a tumour with central necrosis in the tumour. Twenty percent of patients had cervical lymphadenopathy. Forty percent of patients showed perineural infiltration (Liu et al. 2012).

J. Goh and A. Karandikar

230

a

b Fig. 36 (a) Left nasopharyngeal involvement in a 57-year-old male with diffuse large B-cell lymphoma. (Image courtesy of Wendy Smoker, Iowa.) (b, c) Nasopharyngeal lymphoma (different patient). Central

c nasopharyngeal mass with intermediate T2W signal and enhancement, displacing the septa of the nasopharyngeal adenoid in the T2W images

Nasopharyngeal Neoplasms

231

a

b

c

d

Fig. 37  Male patient with adenoid cystic carcinoma of the nasopharynx. (a) Primary at diagnosis in 2005. (b–d) Images in 2011. Progression of tumour mass with PPF

infiltration, right cavernous sinus and sphenoid sinus invasion, and perineural spread

Acknowledgements  The authors would like to acknowledge the following:

References

Drs. Lim Ming Yann (Department of Otolaryngology), Timothy Cheo (Department of Radiation Oncology, Tan Tock Seng Hospital), Dr. Mark Khoo (Amandela ENT Head and Neck Centre, Mt Elizabeth, Novena) and Dr. Tan Tiong Yong (Department of Radiology, Changi General Hospital) for their valuable input; and Drs. Amanda Liew and Khoo Hau Wei (Department of Diagnostic Radiology, Tan Tock Seng Hospital) for their assistance in proofreading the manuscript.

Adams S, Baum RP, Stuckensen T, Bitter K, Hör G (1998) Prospective comparison of 18 F-FDG PET with conventional imaging modalities (CT, MRI, US) in lymph node staging of head and neck cancer. Eur J Nucl Med 25(9):1255–1260 Ai QY, Hu CW, Bhatia KS, Poon DM, Hui EP, Mo FK, Law BKH, Tong M, Ma BB, Chan AT, King AD (2018a) Nasopharyngeal carcinoma: relationship between invasion of the prevertebral space and distant metastases. Eur Arch Otorhinolaryngol 275(2):497–505 Ai QY, King AD, Mo FK, Law BKH, Bhatia KS, Ma BB, Poon DM, Kam MK (2017) Prediction of distant metastases from nasopharyngeal carcinoma: improved diagnostic performance of MRI using nodal volume in N1 and N2 stage disease. Oral Oncol 69:74–79

232 Ai QY, King AD, Mo FKF, Law BKH, Bhatia KS, Poon DMC, Kam MKM, Ma BB (2018b) Staging nodal metastases in nasopharyngeal carcinoma: which method should be used to measure nodal dimension on MRI? Clin Radiol 73(7):640–646 Ai QY, King AD, Poon DM, Mo FK, Hui EP, Tong M, Ahuja AT, Ma BB, Chan AT (2019) Extranodal extension is a criterion for poor outcome in patients with metastatic nodes from cancer of the nasopharynx. Oral Oncol 88:124–130 Alhilali L, Reynolds AR, Fakhran S (2014) Osteoradionecrosis after radiation therapy for head and neck cancer: differentiation from recurrent ­disease with CT and PET/CT imaging. Am J Neuroradiol 35(7):1405–1411 Al-Ibraheem A, Buck A, Krause BJ, Scheidhauer K, Schwaiger M (2009) Clinical applications of FDG PET and PET/CT in head and neck cancer. J Oncol 2009:208725 Amin MB, Edge SB (eds) (2017) AJCC cancer staging manual. Springer Bei JX, Zuo XY, Liu WS, Guo YM, Zeng YX (2016) Genetic susceptibility to the endemic form of NPC. Chin Clin Oncol 5(2):15 Bhatia KS, King AD, Vlantis AC, Ahuja AT, Tse GM (2012) Nasopharyngeal mucosa and adenoids: appearance at MR imaging. Radiology 263(2):437–443 Buell P (1974) The effect of migration on the risk of nasopharyngeal cancer among Chinese. Cancer Res 34(5):1189–1191 Chan SC, Yeh CH, Yen TC, Ng SH, Chang JTC, Lin CY, Yen-Ming T, Fan KH, Huang BS, Hsu CL, Chang KP (2018) Clinical utility of simultaneous whole-body 18 F-FDG PET/MRI as a single-step imaging modality in the staging of primary nasopharyngeal carcinoma. Eur J Nucl Med Mol Imaging 45(8):1297–1308 Chang ET, Adami HO (2006) The enigmatic epidemiology of nasopharyngeal carcinoma. Cancer Epidemiol Prev Biomarkers 15(10):1765–1777 Chang JTC, Chan SC, Yen TC, Liao CT, Lin CY, Lin KJ, Chen IH, Wang HM, Chang YC, Chen TM, Kang CJ (2005) Nasopharyngeal carcinoma staging by (18) F-fluorodeoxyglucose positron emission tomography. Int J Radiat Oncol Biol Phys 62(2):501–507 Chang MC, Chen JH, Liang JA, Yang KT, Cheng KY, Kao CH (2013) Accuracy of whole-body FDG-PET and FDG-PET/CT in M staging of nasopharyngeal carcinoma: a systematic review and meta-analysis. Eur J Radiol 82(2):366–373 Chao ST, Suh JH, Raja S, Lee SY, Barnett G (2001) The sensitivity and specificity of FDG PET in distinguishing recurrent brain tumor from radionecrosis in patients treated with stereotactic radiosurgery. Int J Cancer 96(3):191–197 Chen C, Fei Z, Huang C, Ding J, Chen L (2018) Prognostic value of tumor burden in nasopharyngeal carcinoma. Cancer Manag Res 10:3169 Chen L, Liu LZ, Chen M, Li WF, Yin WJ, Lin AH, Sun Y, Li L, Ma J (2012) Prognostic value of subclassification using MRI in the T4 classification nasopharyngeal

J. Goh and A. Karandikar carcinoma intensity-modulated radiotherapy treatment. Int J Radiat Oncol Biol Phys 84(1):196–202 Chin SC, Jen YM, Chen CY, Som PM (2005) Necrotic nasopharyngeal mucosa: an ominous MR sign of a carotid artery pseudoaneurysm. Am J Neuroradiol 26(2):414–416 Choi YW, Munden RF, Erasmus JJ, Joo Park K, Chung WK, Jeon SC, Park CK (2004) Effects of radiation therapy on the lung: radiologic appearances and differential diagnosis. Radiographics 24(4):985–997 Chong V, Fan Y, Toh K, Khoo J, Lim T (1995a) Magnetic resonance imaging and computed tomography features of nasopharyngeal carcinoma with maxillary sinus involvement. Australas Radiol 39(1):2–9. https://doi. org/10.1111/j.1440-1673.1995.tb00222.x Chong VF, Fan YF (2000) Facial lymphadenopathy in nasopharyngeal carcinoma. Clin Radiol 55(5):363–367 Chong VFH, Fan YF (1993) MRI and CT assessment of paranasal sinus involvement in nasopharyngeal carcinoma (NPC). Clin Radiol 48(5):345 Chong VFH, Fan YF (1998) Hypoglossal nerve palsy in nasopharyngeal carcinoma. Eur Radiol 8(6):939–945 Chong VFH, Ong CK (2008) Nasopharyngeal carcinoma. Eur J Radiol 66(3):437–447 Chong VFH, Fan YF, Khoo JBK (1995b) Retropharyngeal lymphadenopathy in nasopharyngeal carcinoma. Eur J Radiol 21(2):100–105 Comoretto M, Balestreri L, Borsatti E, Cimitan M, Franchin G, Lise M (2008) Detection and restaging of residual and/or recurrent nasopharyngeal carcinoma after chemotherapy and radiation therapy: comparison of MR imaging and FDG PET/CT. Radiology 249(1):203–211 Curtin HD, Ishwaran H, Mancuso AA, Dalley RW, Caudry DJ, McNeil BJ (1998) Comparison of CT and MR imaging in staging of neck metastases. Radiology 207(1):123–130 Dong J, Zhang L, Mo Y, Tian L, Liu L, Wu P (2015) Discovery of invasion routes for nasopharyngeal adenoid cystic carcinoma. Journal of. Cancer 6(1):90 Downer J, Fryer E, Capper J, Woo EK (2011) Pleomorphic adenoma of the nasopharyngeal mucosal space with locally aggressive appearance (2010: 11b). Eur Radiol 21(2):443–446 Dubrulle F, Souillard R, Hermans R (2007) Extension patterns of nasopharyngeal carcinoma. Eur Radiol 17(10):2622–2630 Feng AC, Wu MC, Tsai SY, Chan KY, Cheng SH, Wang A, Chen SS, Jian JJ, Terng SD, Huang AT (2006) Prevertebral muscle involvement in nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 65(4):1026–1035 Goh JPN, Karandikar A, Loke SC, Tan TY (2017) Skull base osteomyelitis secondary to malignant otitis externa mimicking advanced nasopharyngeal cancer: MR imaging features at initial presentation. Am J Otolaryngol 38(4):466–471 Guo X, Johnson RC, Deng H, Liao J, Guan L, Nelson GW, Tang M, Zheng Y, de The G, O’Brien SJ, Winkler CA (2009) Evaluation of nonviral risk factors for

Nasopharyngeal Neoplasms nasopharyngeal carcinoma in a high-risk population of Southern China. Int J Cancer 124(12):2942–2947 Han P, Wang X, Liang F, Liu Y, Qiu X, Xu Y, Chen R, Yu S, Huang X (2018) Osteoradionecrosis of the Skull Base in nasopharyngeal carcinoma: incidence and risk factors. Int J Radiat Oncol Biol Phys 102(3):552–555 Harnsberger HR, Dillon WP (1985) Major motor atrophic patterns in the face and neck: CT evaluation. Radiology 155(3):665–670 Heng DM, Wee J, Fong KW, Lian LG, Sethi VK, Chua ET, Yang TL, Khoo Tan HS, Lee KS, Lee KM, Tan T (1999) Prognostic factors in 677 patients in Singapore with nondisseminated nasopharyngeal carcinoma. Cancer 86(10):1912–1920 Henle G, Henle W (1976) Epstein-Barr virus-specific IgA serum antibodies as an outstanding feature of nasopharyngeal carcinoma. Int J Cancer 17(1):1–7 Hirshoren N, Damti S, Weinberger J, Meirovitz A, Sosna J, Eliashar R, Eliahou R (2019) Diffusion weighted magnetic resonance imaging of pre and post treatment nasopharyngeal carcinoma. Surg Oncol 30:122–125 Ho FC, Tham IW, Earnest A, Lee KM, Lu JJ (2012) Patterns of regional lymph node metastasis of nasopharyngeal carcinoma: a meta-analysis of clinical evidence. BMC Cancer 12(1):98 Hsu WM, Wang AG (2004) Nasopharyngeal carcinoma with orbital invasion. Eye 18(8):833 Huang TT, Chen PR, Hsu YH, Sheen TS, Chang YL, Hsu LP (2005) Nasopharyngeal carcinoma invading the lacrimal apparatus. Tzu Chi Med J 17:349–352 Kataki AC, Simons MJ, Das AK, Sharma K, Mehra NK (2011) Nasopharyngeal carcinoma in the northeastern states of India. Chinese J Cancer 30(2):106 Kimura F, Kim KS, Friedman H, Russell EJ, Breit R (1990) MR imaging of the normal and abnormal clivus. AJR Am J Roentgenol 155(6):1285–1291 King AD, Ahuja AT, Leung SF, Lam WW, Teo P, Chan YL, Metreweli C (2000a) Neck node metastases from nasopharyngeal carcinoma: MR imaging of patterns of disease. Head Neck 22(3):275–281 King AD, Ma BB, Yau YY, Zee B, Leung SF, Wong JKT, Kam MKM, Ahuja AT, Chan ATC (2008) The impact of 18F-FDG PET/CT on assessment of nasopharyngeal carcinoma at diagnosis. Br J Radiol 81(964):291–298 King AD, Teo P, Lam WWM, Leung SF, Metreweli C (2000b) Paranasopharyngeal space involvement in nasopharyngeal cancer: dectection by CT and MRI. Clin Oncol 12(6):397–402 King AD, Vlantis AC, Bhatia KS, Zee BC, Woo JK, Tse GM, Chan AT, Ahuja AT (2011) Primary nasopharyngeal carcinoma: diagnostic accuracy of MR imaging versus that of endoscopy and endoscopic biopsy. Radiology 258(2):531–537 King AD, Vlantis AC, Tsang RKY, Gary TMK, Au AKY, Chan CY, Kok SY, Kwok WT, Lui HK, Ahuja AT (2006) Magnetic resonance imaging for the detection of nasopharyngeal carcinoma. Am J Neuroradiol 27(6):1288–1291 King AD, Wong LYS, Law BKH, Bhatia KS, Woo JKS, Ai QY, Tan TY, Goh J, Chuah KL, Mo FKF, Chan

233 KCA (2018) MR imaging criteria for the detection of nasopharyngeal carcinoma: discrimination of early-­ stage primary tumors from benign hyperplasia. Am J Neuroradiol 39(3):515–523 Kwong DLW, Lee V, Lam KO, Wong GK, Khong PL (2015) Biological tumor volume (BTV) as a predictor for local control in nasopharyngeal carcinoma (NPC) Lan M, Huang Y, Chen CY, Han F, Wu SX, Tian L, Zheng L, Lu TX (2015) Prognostic value of cervical nodal necrosis in nasopharyngeal carcinoma: analysis of 1800 patients with positive cervical nodal metastasis at MR imaging. Radiology 276(2):536–544 Lee AW, Foo W, Law SC, Poon YF, Sze WM, Tung SY, Lau WH (1997) Nasopharyngeal carcinoma: presenting symptoms and duration before diagnosis. Hong Kong Med J 3(4):355–361 Lee AW, Lin JC, Ng WT (2012) Current management of nasopharyngeal cancer. In Seminars in radiation oncology (Vol 22(3), pp 233–244). WB Saunders Lee AW, Lung ML, Ng WT (eds) (2019) Commentary on Chapter 10. In: Nasopharyngeal carcinoma: from etiology to clinical practice. Academic Press, pp 234–237 Lee CC, Chu ST, Chou P, Lee CC, Chen LF (2008) The prognostic influence of prevertebral space involvement in nasopharyngeal carcinoma. Clin Otolaryngol 33(5):442–449 Lee JY, Lee JH, Song JS, Song MJ, Hwang SJ, Yoon RG, Jang SW, Park JE, Heo YJ, Choi YJ, Baek JH (2016) Superior cervical sympathetic ganglion: normal imaging appearance on 3T-MRI. Korean J Radiol 17(5):657–663 Lee SL, Lee CY, Silver SM, Kuhar S (2006) Nasopharyngeal pleomorphic adenoma in the adult. Laryngoscope 116(7):1281–1283 Liang SB, Teng JJ, Hu XF, Yang XL, Luo M, Fang XN, Liu DS, Chen Y, Fu LW (2017) Prognostic value of total tumor volume in patients with nasopharyngeal carcinoma treated with intensity-modulated radiotherapy. BMC Cancer 17(1):506 Liu FY, Lin CY, Chang JT, Ng SH, Chin SC, Wang HM, Liao CT, Chan SC, Yen TC (2007) 18F-FDG PET can replace conventional work-up in primary M staging of nonkeratinizing nasopharyngeal carcinoma. J Nucl Med 48(10):1614–1619 Liu LZ, Zhang GY, Xie CM, Liu XW, Cui CY, Li L (2006) Magnetic resonance imaging of retropharyngeal lymph node metastasis in nasopharyngeal carcinoma: patterns of spread. Int J Radiat Oncol Biol Phys 66(3):721–730 Liu TR, Yang AK, Guo X, Li QL, Song M, He JH, Wang YH, Guo ZM, Zhang Q, Chen WQ, Chen FJ (2008) Adenoid cystic carcinoma of the nasopharynx: 27-year experience. Laryngoscope 118(11):1981–1988 Liu XW, Xie CM, Li H, Zhang R, Geng ZJ, Mo YX, Zhao J, Cai MY, Lv YC, Wu PH (2012) Nasopharyngeal adenoid cystic carcinoma: magnetic resonance imaging features in ten cases. Chinese J Cancer 31(1):19 Loke SC, Karandikar A, Ravanelli M, Farina D, Goh JPN, Ling EA, Maroldi R, Tan TY (2016) Superior cervical ganglion mimicking retropharyngeal adenopathy

234 in head and neck cancer patients: MRI features with anatomic, histologic, and surgical correlation. Neuroradiology 58(1):45–50 Lu CH, Yang CY, Wang CP, Yang CC, Liu HM, Chen YF (2010) Imaging of nasopharyngeal inflammatory pseudotumours: differential from nasopharyngeal carcinoma. Br J Radiol 83(985):8–16 Luo CB, Teng MM, Chen SS, Lirng JF, Guo WY, Chang TSUEN (1998) Orbital invasion in nasopharyngeal carcinoma: evaluation with computed tomography and magnetic resonance imaging. Chinese Med J 61(7):382–388 Luo DH, Yang J, Qiu HZ, Shen T, Chen QY, Huang PY, Sun R, Qian CN, Mai HQ, Guo X, Mo HY (2014) A new T classification based on masticator space involvement in nasopharyngeal carcinoma: a study of 742 cases with magnetic resonance imaging. BMC Cancer 14(1):653 Luo FF, Li ZY, Huang SN, Chen G, Xie TT, Li YQ, Xing WW, Li WY, Lu YK, Ding H (2017) Prevalence of human papillomavirus in patients with nasopharyngeal carcinoma: a meta-analysis. Int J Clin Exp Med 10(7):9837–9847 Makimoto Y, Yamamoto S, Takano H, Motoori K, Ueda T, Kazama T, Kaneoya K, Shimofusa R, Uno T, Ito H, Hanazawa T (2007) Imaging findings of radiation-­ induced sarcoma of the head and neck. Br J Radiol 80(958):790–797 Maroldi R, Farina D, Borghesi A, Marconi A, Gatti E (2008) Perineural tumor spread. Neuroimaging Clin N Am 18(2):413–429 Mimi CY, Ho JH, Lai SH, Henderson BE (1986) Cantonese-style salted fish as a cause of nasopharyngeal carcinoma: report of a case-control study in Hong Kong. Cancer Res 46(2):956–961 Mohandas A, Marcus C, Kang H, Truong MT, Subramaniam RM (2014) FDG PET/CT in the management of nasopharyngeal carcinoma. Am J Roentgenol 203(2):W146–W157 National Comprehensive Cancer Network (NCCN) (2019) Clinical Practice Guidelines in Oncology. Head and Neck Cancers. Version 1 Ng SH, Chan SC, Yen TC, Chang JTC, Liao CT, Ko SF, Liu FY, Chin SC, Fan KH, Hsu CL (2009) Staging of untreated nasopharyngeal carcinoma with PET/CT: comparison with conventional imaging work-up. Eur J Nucl Med Mol Imaging 36(1):12 Ng SH, Chang JTC, Chan SC, Ko SF, Wang HM, Liao CT, Chang YC, Yen TC (2004) Nodal metastases of nasopharyngeal carcinoma: patterns of disease on MRI and FDG PET.  Eur J Nucl Med Mol Imaging 31(8):1073–1080 Niedobitek G (2000) Epstein-Barr virus infection in the pathogenesis of nasopharyngeal carcinoma. Mol Pathol 53(5):248 Offiah C, Hall E (2011) Post-treatment imaging appearances in head and neck cancer patients. Clin Radiol 66(1):13–24 Old LJ, Boyse EA, Oettgen HF et al (1966) Precipitating antibody in human serum to an antigen present in cul-

J. Goh and A. Karandikar tured Burkittls lymphoma cells. Proc Natl Acad Sci U S A 56(6):1699–1704 Ong CK, Chong VF (2011) Neoplasms of the nasopharynx. In: Head and neck cancer imaging. Springer, Berlin, Heidelberg, pp 163–179 Pan JJ, Ng WT, Zong JF, Lee SW, Choi HC, Chan LL, Lin SJ, Guo QJ, Sze HC, Chen YB, Xiao YP (2016) Prognostic nomogram for refining the prognostication of the proposed 8th edition of the AJCC/UICC staging system for nasopharyngeal cancer in the era of intensity-modulated radiotherapy. Cancer 122(21):3307–3315 Pathmanathan R (1997) Pathology. In: Chong VFH, Tsao SY (eds) Nasopharyngeal carcinoma. Armour Publishing, Hong Kong, pp 6–13 Prasad KK, Sirsath NT, Naiknaware KV, Rani KS, Bhatia MS (2017) Carotid blowout syndrome: an oncological emergency less discussed. South Asian J Cancer 6(2):85 Raghupathy R, Hui EP, Chan AT (2014) Epstein-Barr virus as a paradigm in nasopharyngeal cancer: from lab to clinic. Am Soc Clin Oncol Educ Book 34: 149–153 Razek AAKA, King A (2012) MRI and CT of nasopharyngeal carcinoma. Am J Roentgenol 198(1):11–18 Saito N, Nadgir RN, Nakahira M, Takahashi M, Uchino A, Kimura F, Truong MT, Sakai O (2012) Posttreatment CT and MR imaging in head and neck cancer: what the radiologist needs to know. Radiographics 32(5):1261–1282 Sham JS, Choy D (1990) Prognostic factors of nasopharyngeal carcinoma: a review of 759 patients. Br J Radiol 63(745):51–58 Sham JS, Cheung YK, Choy D, Chan FL, Leong L (1991a) Nasopharyngeal carcinoma: CT evaluation of patterns of tumor spread. Am J Neuroradiol 12(2):265–270 Sham JS, Choy D, Cheung YK, Chan FL, Leong L (1991b) Cranial nerve involvement and base of the skull erosion in nasopharyngeal carcinoma. Cancer 68(2):422–426 Sigal R (1995) Tumeurs du nasopharynx. Imagerie Radiologique de la tête et du cou. P Bourjat, F Veillon. Vigot, pp 375–389 Sigg MB, Steinert H, Grätz K, Hugenin P, Stoeckli S, Eyrich GK (2003) Staging of head and neck tumors: [18F]fluorodeoxyglucose positron emission tomography compared with physical examination and conventional imaging modalities. J Oral Maxillofac Surg 61(9):1022–1029 Silver AJ, Mawad ME, Hilal SK, Sane P, Ganti SR (1983) Computed tomography of the nasopharynx and related spaces. Part II: pathology. Radiology 147(3):733–738 Smoker WRK, Gentry LR (1986) Computed tomography of the nasopharynx and related spaces. Seminars in Ultrasound 7(2):107–130 Soprani F, Armaroli V, Venturini A, Emiliani E, Casolino D (2007) A rare case of adenoid cystic carcinoma of the nasopharynx manifesting as Horner’s syndrome: discussion and review of the literature. Acta Otorhinolaryngol Ital 27(4):216

Nasopharyngeal Neoplasms Sze H, Chan LL, Ng WT, Hung AW, Lee MC, Chang AT, Chan OS, Cheung FY, Yeung RM, Lee AW (2014) Should all nasopharyngeal carcinoma with masticator space involvement be staged as T4? Oral Oncol 50(12):1188–1195 Tan LHC (2004) Lymphomas involving Waldeyer’s ring: placement, paradigms, peculiarities, pitfalls, patterns and postulates. Ann Acad Med Singapore 33:15–26 Tan TH, Soon YY, Cheo T, Ho F, Wong LC, Tey J, Tham IW (2018) Induction chemotherapy for locally advanced nasopharyngeal carcinoma treated with concurrent chemoradiation: a systematic review and meta-­ analysis. Radiother Oncol 129(1):10–17 Tang LL, Li WF, Chen L, Sun Y, Chen Y, Liu LZ, Mao YP, Lin AH, Li L, Ma J (2010) Prognostic value and staging categories of anatomic masticator space involvement in nasopharyngeal carcinoma: a study of 924 cases with MR imaging. Radiology 257(1):151–157 Tang LL, Chen YP, Mao YP, Wang ZX, Guo R, Chen L, Tian L, Lin AH, Li L, Sun Y, Ma J (2017) Validation of the 8th edition of the UICC/AJCC staging system for nasopharyngeal carcinoma from endemic areas in the intensity-modulated radiotherapy era. J Natl Compr Cancer Netw 15(7):913–919 Tang LL, Guo R, Zhou G, Sun Y, Liu LZ, Lin AH, Mai H, Shao J, Li L, Ma J (2014) Prognostic value and staging classification of retropharyngeal lymph node metastasis in nasopharyngeal carcinoma patients treated with intensity-modulated radiotherapy. PLoS One 9(10):e108375 Tao Q, Srivastava G, Chan AC, Chung LP, Loke SL, Ho FC (1995) Evidence for lytic infection by Epstein-­ Barr virus in mucosal lymphocytes instead of nasopharyngeal epithelial cells in normal individuals. J Med Virol 45(1):71–77 Teo PM, Kwan WH, Lee WY, Leung SF, Johnson PJ (1996) Prognosticators determining survival subsequent to distant metastasis from nasopharyngeal carcinoma. Cancer 77(12):2423–2431 Ting Y, Chee J, Charn TC, Loh KS, Choong CC, Ting E, Lim CM (2017) Prognostic significance of cystic lymph nodal metastasis in nasopharyngeal carcinoma. Head Neck 39(9):1832–1839 Toro C, Rinaldo AE, Silver CE, Politi M, Ferlito A (2009) Paraneoplastic syndromes in patients with nasopharyngeal cancer. Auris Nasus Larynx 36(5):513–520 Ung A, Chen CJ, Levine PH, Cheng YJ, Brinton LA, Chen IH, Goldstein AM, Hsu MM, Chhabra SK, Chen JY, Apple RJ (1999) Familial and sporadic cases of nasopharyngeal carcinoma in Taiwan. Anticancer Res 19(1B):661–665 Van den Brekel MWM, Stel HV, Castelijns JA, Nauta JJ, Van der Waal I, Valk J, Meyer CJ, Snow GB (1990) Cervical lymph node metastasis: assessment of radiologic criteria. Radiology 177(2):379–384 Vellayappan BA, Soon YY, Earnest A, Zhang Q, Koh WY, Tham IWK, Lee KM (2014) Accuracy of 18F-flurodeoxyglucosepositron emission tomography/

235 computed tomography in the staging of newly diagnosed nasopharyngeal carcinoma: a systematic review and meta-analysis. Radiol Oncol 48(4):331–338 Waldron CA, El-Mofty SK, Gnepp DR (1988) Tumors of the intraoral minor salivary glands: a demographic and histologic study of 426 cases. Oral Surg Oral Med Oral Pathol 66(3):323–333 Wang C, Liu L, Lai S, Su D, Liu Y, Jin G, Zhu X, Luo N (2018) Diagnostic value of diffusion-weighted magnetic resonance imaging for local and skull base recurrence of nasopharyngeal carcinoma after radiotherapy. Medicine 97(34):e11929 Wang CC, See LC, Hong JH, Tang SG (1996) Nasopharyngeal adenoid cystic carcinoma: five new cases and a literature review. J Otolaryngol 25(6):399–403 Wei KR, Zheng RS, Zhang SW, Liang ZH, Li ZM, Chen WQ (2017) Nasopharyngeal carcinoma incidence and mortality in China, 2013. Chinese J Cancer 36(1):90 Wong KT, Chan YL, Metreweli C (2001) Magnetic resonance imaging evaluation of the clivus and nasopharyngeal carcinoma. Hong Kong J Radiol 4:268–271 Wong WM, Young SM, Amrith S (2017) Ophthalmic involvement in nasopharyngeal carcinoma. Orbit 36(2):84–90 Yeğin Y, Çelik M, Kayhan FT, Erdur Ö (2015) Management of pleomorphic adenoma in the nasopharynx: a case report. Balkan Med J 32(1):118–120 Yen TC, Chang JTC, Ng SH, Chang YC, Chan SC, Lin KJ, Lin WJ, Fu YK, Lin CY (2005) The value of 18F-­ FDG PET in the detection of stage M0 carcinoma of the nasopharynx. J Nucl Med 46(3):405–410 Yokota H, Mukai H, Hattori S, Yamada K, Anzai Y, Uno T (2018) MR imaging of the superior cervical ganglion and inferior ganglion of the vagus nerve: structures that can mimic pathologic retropharyngeal lymph nodes. Am J Neuroradiol 39(1):170–176 Yom SS, Wee J (2014) Nasopharyngeal carcinoma: management strategies. Future Medicine Ltd. Yu MC, Garabrant DH, Huang TB, Henderson BE (1990) Occupational and other non-dietary risk factors for nasopharyngeal carcinoma in Guangzhou, China. Int J Cancer 45(6):1033–1039 Yuan H, Ai QY, Kwong DLW, Fong DYT, King AD, Vardhanabhuti V, Lee VHF, Khong PL (2017) Cervical nodal volume for prognostication and risk stratification of patients with nasopharyngeal carcinoma, and implications on the TNM-staging system. Sci Rep 7(1):10387 Zhang GY, Huang Y, Cai XY, Chen XP, Xu T, Wu J, Wei WH, Liu LZ, Huang ZL, Zhou MM, Xia ST (2014) Prognostic value of grading masticator space involvement in nasopharyngeal carcinoma according to MR imaging findings. Radiology 273(1):136–143 Zhang GY, Liu LZ, Wei WH, Deng YM, Li YZ, Liu XW (2010) Radiologic criteria of retropharyngeal lymph node metastasis in nasopharyngeal carcinoma treated with radiation therapy. Radiology 255(2):605–612

Parapharyngeal Space Neoplasms Robert Hermans

Contents 1

Introduction

2 A  natomy 2.1  F  ascial Layers and Compartments 2.2  Radiological Anatomy I maging Findings in Parapharyngeal Space Lesions 3.1  Primary Lesions of the Parapharyngeal Space 3.2  Secondary Lesions of the Parapharyngeal Space

Abstract

While primary neoplasms of the parapharyngeal space are rare, secondary displacement or infiltration of this space, by pathology originating in neighboring spaces, is more commonly seen. An important role of imaging in the evaluation of parapharyngeal space pathology is to identify the space of origin. In combination with the imaging characteristics of the mass, this allows to reduce the differential diagnosis to a limited number of possibilities. This chapter reviews the key anatomical features of the parapharyngeal space, and explains the imaging approach to mass lesions at this level.

 237  238  238  240

3

4

Conclusion

References

 241  241  247  250  251

1

R. Hermans (*) Department of Radiology, University Hospitals, KU Leuven, Leuven, Belgium e-mail: [email protected]

Introduction

The parapharyngeal space (PPS) is a deep space of the neck shaped as a tilted-up pyramid with its base attaching to the skull base and the apex reaching the level of the hyoid bone, and almost exclusively containing fat. Primary neoplasms arising in this space are rare, accounting for only 0.5% of all head and neck tumors (Olsen 1994; Miller et  al. 1996; Pang et al. 2002). About 70–80% of the tumors originating from the PPS itself are benign (Luna-­ Ortiz et al. 2005).

Med Radiol Diagn Imaging (2020) https://doi.org/10.1007/174_2020_228, © Springer Nature Switzerland AG Published Online: 23 April 2020

237

R. Hermans

238

The PPS is more commonly displaced or infiltrated by lesions arising in adjacent structures and spaces, including the pharynx, as well as the parotid, masticator, and retropharyngeal spaces. Small tumors of the PPS are often incidental findings. Larger tumors cause aspecific symptoms, including sore throat, ear fullness, dysphagia, and, less frequently, jaw pain combined with cranial nerve palsy. These last two symptoms are suggestive for a malignant lesion. Because of their deep location, the clinical assessment of tumors in this region is limited, commonly causing a delay between the onset of symptoms and diagnosis. Large lesions cause a bulging of the lateral oro- and/or nasopharyngeal wall. When a PPS lesion grows laterally, facial swelling may result at the level of the parotid or submandibular region. Rarely, the mandible will be displaced by a slowly growing tumor (Farina et al. 1999). An important role of imaging in the evaluation of PPS pathology is to identify the space of origin. Lesions arising from the adjacent spaces displace the PPS in a particular way; in combination with the imaging characteristics of the mass, this a

Fig. 1 (a) Axial T1-weighted spin-echo image at the level of the soft palate. The boundaries of the parapharyngeal space (PPS) (including prestyloid and retrostyloid compartment) are indicated by arrows and arrowheads on the right. On the left, the adjacent spaces are labeled: pharyngeal mucosal space (1), masticator space (2), parotid space (3), retropharyngeal/prevertebral space (4). (b)

information allows to reduce the differential diagnosis to a limited number of possibilities. Imaging studies should also identify hypervascular neoplasms, in order to avoid biopsy and its possible complications.

2

Anatomy

2.1

Fascial Layers and Compartments

The PPS is medially in close contact with the pharynx, bordered by the middle layer of the deep cervical fascia (also known as buccopharyngeal fascia) (Figs. 1 and 2). Superiorly, the pharyngeal constrictor muscle does not reach the skull base, and at that level, the lumen of the nasopharynx is held open by the thick pharyngobasilar fascia. This fascia lies within the middle layer of the deep cervical fascia. The pharyngobasilar fascia is interrupted at the level of the sinus of Morgagni, an opening through which the cartilaginous part of the Eustachian b

Coronal T1-weighted spin-echo images through prestyloid compartment of the PPS.  Inferiorly, this space is closed by the submandibular gland (5), while superiorly, it reaches the skull base (6). The foramen ovale (arrow), through which exits the mandibular nerve, communicates with the masticator space. The styloglossal muscles run through the PPS (arrowheads)

Parapharyngeal Space Neoplasms Pterygoid process

239 Masticator space

Mandible Superficial layer of deep cervical fascia

Tensor veli palatini muscle

Tensor-vascular-styloid fascia

Eustachian tube Levator veli palatini muscle

Stylopharyngeal fascia

Pharyngobasilar fascia Buccopharyngeal fascia

Parotid gland (deep lobe)

Alar fascia

Styloid process

Prevertebral fascia

Internal jugular vein

Retropharyngeal lymph node

Digastric muscle (posterior belly)

Charpy’s fascia

Sympathetic chain Internal carotid artery

Mastoid process

Vagal nerve

Deep layer of deep cervical fascia

Fig. 2  Topographical anatomy of the PPS in the axial plane, including the different fascial layers at this level

tube and the levator veli palatini muscle enter the nasopharynx. This area should be carefully inspected on imaging studies of the nasopharynx, as it is a common route of spread for nasopharyngeal carcinomas to the PPS and skull base (Fig. 3). The superficial layer of the deep cervical fascia is lateral to the PPS, separating this space from the masticator space. This fascia curves around the medial surface of the pterygoid muscles and extends from the mandible to the skull base, where it attaches just medial to foramen ovale. As a consequence, the mandibular nerve (V3), as it courses through this foramen, directly enters the masticator space. In its posterolateral portion, the PPS is in contact with the deep lobe of the parotid gland. The existence of a fascial layer at this level is controversial. The anterior border of the PPS is the pterygomandibular raphe. Inferiorly, the PPS gradually becomes narrower and ends at the level of the hyoid bone and the superior margin of the submandibular salivary gland.

The posterior border of the PPS is the most complex and controversial; different descriptions are found in literature. Some authors consider the PPS completely separated from the more posterior carotid space: the anterior surface of the carotid sheath (made up of the three layers of deep cervical fascia) draws the borderline between the two spaces (Som and Curtin 1995). Others consider the carotid space to be part of the PPS (Mukherji and Castillo 1998). Three more fascial structures are described, acting as anatomical landmarks subdividing the PPS.  The most important one is the tensor-­ vascular-­styloid fascia (TVS), a layer that extends from the inferior border of tensor veli palatini muscle, posterolaterally and inferiorly to the styloid process and muscles (Fig.  2). Anteriorly, it reaches the pterygomandibular raphe and therefore it closes the gap between the skull base, the tensor veli palatini muscle, and the styloid process (Som and Curtin 1995). The other two are the stylopharyngeal fascia and Charpy’s fascia, also known as “cloison sagittale” (Fig. 2).

R. Hermans

240

a

b

Fig. 3 Axial gadolinium-enhanced T1-weighted spin-­ echo images, showing the relationship between the superior constrictor muscle, the pharyngobasilar fascia, and a nasopharyngeal cancer. (a) Axial section at the level of soft palate. The superior constrictor muscle (arrowheads) is surrounding the nasopharynx; the tumor (asterisk) is confined to the nasopharyngeal lumen. (b) At a higher

level, the nasopharyngeal lumen is bordered by the pharyngobasilar fascia (arrowheads). Bilaterally, the tensor veli palatini muscle (black arrow) is visible on the outer side of this fascial layer. On the right, the tumor extends into the PPS through the pharyngobasilar fascia (white arrow); this site likely corresponds to the sinus of Morgagni

The TVS fascia allows to further subdivide the parapharyngeal space into two compartments: the prestyloid compartment, lying between the pterygoid muscles and the TVS fascia, and the retrostyloid compartment, just medial to the TVS fascia itself and including the carotid space (Maroldi et al. 1994; Nasser and Attia 1990). The PPS mainly contains fat tissue and loose connective tissue. In the prestyloid compartment ectopic minor salivary glands and vascular structures (pharyngeal ascending and internal ­ ­maxillary artery, pharyngeal venous plexus) are found. The retrostyloid PPS contains the internal carotid artery (ICA), the internal jugular vein (IJV), the cranial nerves IX–X–XI–XII, and the sympathetic plexus. Lymph nodes of the deep cervical chain, known as the jugulodigastric lymph nodes, are only found below the level of the posterior belly of the digastric muscle (Grégoire et al. 2003).

In the axial plane the prestyloid compartment of the PPS is recognized as a triangular fat-filled space (Fig. 1) with maximum width at the level of the soft palate. The pharyngobasilar fascia is sometimes visible on MRI as a hypointense line (Fig. 3). The sinus of Morgagni itself is not visible on MR, but the Eustachian tube, particularly at the torus tubarius, where it opens in the nasopharyngeal lumen, can be used as an anatomical landmark. The TVS, stylopharyngeal, and Charpy’s fascia cannot be routinely identified on MR. Their course can be mentally outlined on axial scans by using the tensor veli palatini muscle and styloid process as anatomical landmarks. The lateral fascial border of the PPS is not identified, but its cranial attachment lies just medial to foramen ovale (Fig. 1). MR is able to depict the course of the mandibular nerve—from Meckel’s cave, through the foramen ovale—and also the proximal segment of the main terminal branch, the inferior alveolar nerve, as it runs to the entrance of the mandibular canal. The ICA and IJV, just medial to the styloid process, generally show a flow void on a T1- and T2-weighted MR sequence.

2.2

Radiological Anatomy

Because of its higher contrast resolution, MRI is preferred over CT when dealing with pathology at the level of the PPS.

Parapharyngeal Space Neoplasms

3

Imaging Findings in Parapharyngeal Space Lesions

3.1

Primary Lesions of the Parapharyngeal Space

As indicated, primary lesions of the PPS are rare and most are benign. These lesions can be classified based on histology (mainly salivary gland tumors, neurogenic tumors, and paragangliomas), or according to the compartment of origin (prestyloid or retrostyloid).

3.1.1 Prestyloid Lesions The overwhelming majority of neoplasms in the prestyloid compartment are salivary gland tumors. Most other lesions in this compartment are related to anomalies of the branchial apparatus. Pleomorphic adenoma is the most frequent salivary tumor. Histologically, both epithelial and

a

Fig. 4  Axial T2-weighted (a) and gadolinium-enhanced T1-weighted spin-echo image (b) of a patient with a coincidentally discovered PPS tumor. The lesion cannot be separated from the deep lobe of the parotid gland (black arrow), and largely fills the prestyloid compartment; a thin layer of fat is still visible at the anteromedial margin of the tumor (black arrowheads, b). The pharyngeal wall (white

241

mesenchymal elements are found, hence the name “mixed tumor.” These tumors may arise from ectopic minor salivary glands, localized in the prestyloid compartment along the embryological growth path of the parotid gland. However, much more frequently they originate from the deep lobe of the parotid gland, extending exophytically into the PPS. The most reliable sign of a primary PPS tumor is the presence of a fat layer separating the tumor from the deep lobe of the gland. A deep parotid lobe tumor appears as a more or less dumbbell-­ shaped mass, connected to the parotid gland, potentially widening the stylomandibular tunnel, and displacing anteromedially the PPS fat (Fig. 4). Pleomorphic adenomas usually show the following MR appearance: hyperintense signal intensity on T2-weighted sequences, related to their myxoid component, and often pronounced enhancement with focal areas of hypointensity on T1-weighted sequences. Various signal inten-

b

arrowheads) and medial pterygoid muscle (white arrow) are displaced. The large vessels (curved arrow) are displaced posteriorly. The styloid process and styloid musculature produce an indentation in the posterior tumor margin (black arrowhead, a). Pleomorphic adenoma of the deep parotid lobe

242

Fig. 5  Axial plane T1-weighted spin-echo image of a patient with a right-sided peritonsillar swelling, rapidly increasing over a few hours’ time. A soft-tissue mass (asterisk) is seen in the prestyloid compartment of the PPS; the periphery of the mass, mainly on the tonsillar side, shows spontaneous hyperintensity, compatible with recent hemorrhage (arrows). Pathologic examination showed pleomorphic adenoma

sities may be seen in a pleomorphic adenoma, because of hemorrhage, calcifications, and necrosis (Fig. 5). Although pleomorphic adenoma is a benign tumor, it may recur locally if only a tumorectomy is performed, as the lesion has a thin and incomplete capsule. Most frequently, these relapses are multifocal and appear very hyperintense on T2-weighted sequences (Fig. 6). Malignant transformation of the epithelial component of a pleomorphic adenoma, called “carcinoma ex pleomorphic adenoma” or “malignant mixed tumor,” is the principal reason to remove all pleomorphic adenomas surgically. It manifests itself clinically by a sudden increase in tumor size, sometimes accompanied by pain and facial palsy. The radiological appearance may be relatively unremarkable compared to the “benign” mixed tumor, or it may show infiltration into and destruction of surrounding structures (Fig. 7).

R. Hermans

Salivary malignancies are not frequent, with mucoepidermoid and adenoid cystic carcinoma being the most common histologic types. CT and MR characteristics do not allow a reliable differentiation of a malignant from a benign lesion, although an overall low T2 signal intensity in a salivary gland tumor must raise the suspicion of a malignant tumor. The presence of infiltration into adjacent structures suggests malignancy (Fig. 8). Much rarer neoplasms in the prestyloid compartment are lipomas (Kakani et al. 1992; Smith et  al. 2006) and muscular neoplasms, such as rhabdomyomas and rhabdomyosarcomas. Rhabdomyosarcoma is, apart from lymphoma, the most common malignant head and neck neoplasms in children; the PPS is one of the sites of predilection of this tumor (Fig. 9). Anomalies of the second branchial apparatus represent a spectrum of manifestations, ranging from a fistula to an isolated cyst (Piccin et al. 2008). One end of the spectrum is a fistula between the anterior side of the sternocleidomastoid muscle and the pharyngeal wall at the level of the palatine tonsil. The other end of the spectrum is a cyst, potentially localized anywhere along this tract. The most typical localization of such a cyst is just below the mandibular angle, posterior to the submandibular salivary gland, lateral to the vessels, and anterior to the sternocleidomastoid muscle. As the trajectory of such a branchial apparatus anomaly runs through the PPS, such cysts may also occur within the PPS (Fig.  10) (Adams et al. 2016). Thickened, ­irregular walls may be seen in branchiogenic cysts that are or have been infected. Rarely, a cystic neoplasm originating from the deep parotid lobe, extending into the PPS, may mimic the appearance of a branchiogenic cyst (Fig. 11). Cystic lesions originating from the retropharyngeal space are sometimes confused with parapharyngeal lesions. In the context of a squamous cell carcinoma, a retropharyngeal liquefied nodular mass likely corresponds to a metastatic adenopathy (see Fig.  20). In the absence of such a

Parapharyngeal Space Neoplasms

a

243

b

c

Fig. 6  Axial T2-weighted spin-echo images (a–c) of a patient presenting with an infra-auricular swelling, 9 years after surgery for pleomorphic adenoma. Several bright nodular lesions (arrows) are seen, superficial and deep to

the sternocleidomastoid muscle, as well as in the prestyloid compartment of the PPS (arrow, c). Reoperation confirmed recurrent pleomorphic adenoma

R. Hermans

244

a

b

Fig. 7  Patient in her fifth decade, with a history of pleomorphic adenoma of the deep lobe of the left parotid gland. After initial resection, a recurrence was resected about 1 year ago. The patient now presents with periauricular pain and trismus, and a palpable mass in the preauricular and submandibular region. CT shows a large mass lesion in the parotid space, pre- and retrostyloid compartment of the PPS (arrows, a), extending into the masticator and buccinator space (arrowheads, a), and into the floor of the mouth, and peri-jugular and submandibular region

(arrows, b). Massive osteolysis of the left mandibular body is seen (arrowhead, b). Some adenopathies are seen in level II on the left side. The lower sections of this CT study also revealed lung metastases. Although repeat biopsies always showed pleomorphic adenoma, the clinical behavior and radiological appearance are compatible with malignant degeneration of this tumor. Palliative chemotherapy was initiated. The patient died of progressive disease 1 year later

history, an atypical cystic lesion should raise the possibility of a cystic retropharyngeal adenopathy as can be seen with papillary thyroid carcinoma (Fig.  12) (Lombardi et  al. 2004; Heimgartner and Zbaeren 2009). Another rare lesion that may be seen at this level is a retropharyngeal bronchogenic cyst (Fig. 13) (Jacob et al. 2007).

Schwannomas more frequently arise from the vagal nerve. Malignant degeneration has seldom been reported (Al Otieschan et  al. 1998). The typical MR appearance of a schwannoma is that of an ovoid mass with a slightly hyperintense signal on T2-weighted images; the lesion can be heterogenous because of areas of hemorrhage or cystic degeneration. After the administration of a contrast agent, marked enhancement may lead to a misdiagnosis of a hypervascular tumor. Actually, a schwannoma is a relatively hypovascular lesion and the enhancement is due to extravascular leakage through abnormally permeable vessels with poor venous drainage (Fig. 14). Rarely, flow voids may become visible in an extracranial schwannoma, corresponding to

3.1.2 Retrostyloid Lesions A retrostyloid tumor can easily be identified when an anteromedial displacement of the ICA is present. These lesions also displace the prestyloid PPS fat anteriorly. The most common primary lesions in this compartment are neurogenic tumors (17–25% of all PPS neoplasms) and paragangliomas (10–15%).

Parapharyngeal Space Neoplasms

Fig. 8  Patient in his fourth decade, presenting with a painful swelling in the left parotid region. Axial T2-weighted image shows tumoral lesion in the prestyloid compartment of the PPS, extending into the deep lobe of the parotid gland. The lesion appears not as bright as pleomorphic adenomas usually do (compare with Fig.  4). Also, there is some doubt about the relationship between the mass lesion and medial pterygoid muscle (arrowhead): subtle invasion cannot be excluded. Because of adherence to the surrounding structures, the surgical resection of the lesion was difficult. Histological examination revealed adenocarcinoma

dilated abnormal vessels (Kato et al. 2010); this may cause confusion with a paraganglioma, being a hypervascular tumor, where such flow voids correspond to feeding arteries or draining veins. If necessary, the differentiation can be made by obtaining a dynamic contrast-enhanced study: a schwannoma will nearly always show a slow, steady increase in enhancement, while paragangliomas show a rapidly increasing enhancement. Schwannomas originating from the cervical sympathetic chain are an exception to the rule that a retrostyloid tumor displaces the ICA anteromedially: as this chain is lying medial to the ICA (see Fig. 2), the vessels are expected to be displaced laterally. Such a sympathetic chain

245

schwannoma may therefore be confused with a retropharyngeal adenopathy (Fig.  15) (Yuen et  al. 2006; Saito et  al. 2007; Kahraman et  al. 2009). Paragangliomas (also called glomus tumors) arise from chemoreceptor cells, basically present at three different anatomical sites: at the level of the nodose ganglion of vagal nerve just below the skull base (vagal paraganglioma), at the carotid bifurcation (carotid paraganglioma), and at the level of the jugular foramen (jugular paraganglioma). A marked enhancement and, at MR, a “salt-and-pepper” pattern are quite typical: this is due to the presence of tortuous large-caliber vessels (detected as flow voids) within the mass (Fig. 16). Nevertheless, this pattern can be difficult to see or even absent, particularly in small lesions with predominantly small-caliber feeding vessels (Som and Curtin 1995). Carotid body tumors typically splay the internal and external carotid artery and may extend superiorly in the PPS. A vagal paraganglioma is centered in the retrostyloid compartment of the PPS; it only rarely reaches the carotid bifurcation. A jugular paraganglioma is centered in the jugular foramen, often eroding the surrounding bone and extending into the middle ear (Swartz et al. 1998). CT and MR allow an accurate diagnosis of glomus tumors in most cases. Several vascular anomalies may be encountered in the retrostyloid compartment. Sometimes the ICA shows a tortuous course, running behind the posterior pharyngeal wall: correct assessment avoids a disastrous biopsy of a misinterpreted submucosal pharyngeal lesion. An occlusion of ICA, or an IJV thrombosis, is sometimes encountered; MR shows a complex pattern of signal intensities of the intraluminal clot reflecting the different phases of hemoglobin degradation. MR is an accurate technique to diagnose dissection of the ICA.  Aneurysms of the extracranial part of the ICA occur; rarely these aneurysms cause ­cranial nerve palsy. The diagnosis of such an aneurysm is usually straightforward (Fig. 17).

R. Hermans

246

a

Fig. 9 Axial T2-weighted spin-echo images of an 11-year-old child, presenting with facial pain, mainly during mastication, decreased hearing on the right side, and weight loss. Large mass lesion (asterisk) centered on the right PPS, extending in the nasopharyngeal lumen (white

Fig. 10  Young child presenting with peritonsillar swelling. Axial T2-weighted spin-echo image shows a cystic mass in the prestyloid compartment of the PPS, displacing the pharyngeal wall. The cyst was resected and confirmed to be a branchiogenic cyst

b

arrows). The mass is displacing the muscles of mastication (black arrowhead). Invasion of the prevertebral space (white arrowhead) and apex of the right petrous bone (black arrow) is seen. Biopsy revealed embryonal rhabdomyosarcoma

Fig. 11  Axial T2-weighted spin-echo image of a 38-year-­ old patient, in whom coincidentally a cystic lesion was discovered in the prestyloid compartment of the PPS.  Preoperative hypothesis was branchiogenic cyst. Resection revealed a cystic Warthin’s tumor, exophytically growing from the deep parotid lobe. In retrospect, a relatively broad connection with the deep parotid lobe (asterisk) is seen

Parapharyngeal Space Neoplasms

a

Fig. 12  Asymptomatic, 42-year-old female patient, with coincidentally discovered peritonsillar swelling on right side. Axial T2-weighted (a) and gadolinium-enhanced T1-weighted (b) spin-echo images show largely cystic lesion, filling the prestyloid compartment of the PPS. The mass is centered between the prevertebral muscle (black arrowhead) and internal carotid artery (white arrowhead), a typical localization for a retropharyngeal adenopathy.

247

b

The internal carotid artery, as well as the internal jugular vein (white arrow), is displaced posterolaterally. The cystic mass contains a solid component, showing enhancement. A lesion showing such appearance, especially in a young female patient, should raise the possibility of metastatic papillary thyroid cancer, a diagnosis that eventually was confirmed

3.2

Fig. 13  T2-weighted spin-echo image of a 46-year-old patient, presenting with cough and globus feeling. Clinically a submucosal oropharyngeal swelling was seen, corresponding to a thin-walled cystic lesion anterior to the prevertebral muscles and medial to the internal carotid artery. Resection revealed a bronchogenic cyst

Secondary Lesions of the Parapharyngeal Space

To identify the site of origin of a lesion arising from a space neighboring the PPS, its relationships with the PPS fat and the large vessels must be carefully assessed. Masticator space neoplasms displace this fat plane posteromedially (Fig. 18) (Fernandes et al. 2013) while a tumor originating from the pharynx will usually infiltrate the retrostyloid compartment (Fig. 19), displacing the fat tissue laterally. A retropharyngeal lesion (most often a retropharyngeal adenopathy) displaces the fat of the PPS anterolaterally (Fig. 20). Large and aggressively behaving neoplasms, such as sometimes seen in sarcoma, nasopharyngeal cancer, or lymphoma, may show a trans-­ spatial growth pattern; in such circumstances, a precise diagnosis on imaging studies is not possible.

R. Hermans

248

a

Fig. 14 T2-weighted (a) and gadolinium-enhanced T1-weighted (b) spin-echo images, showing a heterogenous mass lesion in the retrostyloid compartment of the PPS, displacing the internal carotid artery (arrowhead) anteromedially, and the internal jugular vein (arrow) pos-

a

Fig. 15  Teenage patient, with a coincidently discovered parapharyngeal mass lesion; no clinical symptoms. Axial T2-weighted spin-echo image (a) shows hyperintense mass, displacing the internal carotid artery (arrowhead) and internal jugular vein (arrow) laterally. Compared to the axial plane T1-weighted image (b), the coronal

b

terolaterally. These findings are compatible with a schwannoma, most likely originating from the vagal nerve. As this lesion is causing few complaints, for the moment the patient is followed up conservatively

b

gadolinium-­enhanced T1-weighted image (c) shows considerable enhancement of this lesion. The main differential diagnosis was retropharyngeal adenopathy, or neurogenic tumor originating from the sympathetic chain. Surgical exploration showed a sympathetic chain tumor; pathological examination revealed ganglioneuroma

Parapharyngeal Space Neoplasms

249

c

Fig. 15 (continued)

a

Fig. 16  Axial T2-weighted (a) and gadolinium-enhanced T1-weighted spin-echo image (b) of a patient presenting with a submandibular swelling. A mass lesion is seen in the retrostyloid compartment of the PPS, separating the internal carotid artery (black arrowhead) and internal jugular vein (compressed, arrow). The lesion also displaces

Fig. 17  Axial contrast-enhanced CT image shows nodular lesion in the retrostyloid compartment of the PPS. At first sight one may think about retropharyngeal adenopathy. However, the narrowed lumen of the internal carotid artery (white arrowhead) is within the peripheral part of the mass, which also shows some peripheral calcifications (black arrowhead): these findings are compatible with an internal carotid artery aneurysm

b

the external carotid artery (white arrowhead) somewhat anteriorly, but is not centered on the carotid bifurcation. Within the tumor, several vessel-like signal voids are seen, indicating hypervascularity. These findings are compatible with a glomus vagale tumor; this diagnosis was confirmed after resection and pathological examination

R. Hermans

250

Fig. 18  Axial T2-weighted spin-echo image of a 22-year-­ old patient presenting with hearing loss on left side and increasing difficulties with mastication. A soft-tissue mass (black arrows) is seen on the deep side of mandible, extending into the PPS. The lesion involves the deep lobe of the parotid gland. Contrary to the patient shown in Fig.  4, the pterygoid musculature cannot be recognized anymore, suggesting that the point of origin is within the masticator space. Fluid is present in the mastoid, secondary to dysfunction of the Eustachian tube. Biopsy revealed rhabdomyosarcoma

Fig. 20  Axial contrast-enhanced CT image at the level of nasopharynx. Retropharyngeal adenopathy in a patient suffering from squamous cell cancer of the neck (unknown primary). A retropharyngeal adenopathy is typically situated between the prevertebral muscle (arrowhead) and the internal carotid artery (arrow); large adenopathies, such as in this case, displace the prestyloid fat anterolaterally

Secondary involvement of the parapharyngeal space may also be observed in benign multicompartmental lesions, such as hemangioma and lymphangioma. Parapharyngeal abscess is a rare event in the era of broad-spectrum antibiotics, generally secondary to head and neck infections of odontogenic, pharyngeal, tonsillar, otomastoideal, or salivary origin. Sometimes, a PPS tumor may mimic on imaging a parapharyngeal abscess at first sight; this issue can be solved by correlation with the clinical findings (Fig. 21).

4

Conclusion

Imaging has an important role in the characterization of mass lesions originating from the paraFig. 19  Axial gadolinium-enhanced T1-weighted spin-­ echo image of a patient suffering from nasopharyngeal pharyngeal space or its neighboring spaces. A carcinoma. The tumor extends posterolaterally (small good knowledge of the local anatomy is a prereqarrow) in the retrostyloid compartment of the PPS, and uisite, and in combination with the imaging charremains sharply demarcated from the prestyloid compartacteristics of the mass, a precise diagnosis, or at ment by the tensor veli palatini muscle (arrowhead) and tensor-vascular-styloid fascia (large arrow), although the least a limited list of differential diagnoses, can fat plane is somewhat pushed laterally be put forward in most cases.

Parapharyngeal Space Neoplasms

a

Fig. 21  Elderly patient presenting with right-sided throat pain. Clinical examination shows a swollen tonsil, and no fever. Axial contrast-enhanced CT image (a) shows a large fluid-filled mass lesion (asterisk) in the prestyloid compartment of the PPS; this may be confused with a peritonsillar abscess. However, a solid component (arrowhead) with some small calcifications is seen at its lateral side, suggest-

References Adams A, Mankad K, Offiah C, Childs L (2016) Branchial cleft anomalies: a pictorial review of embryological development and spectrum of imaging findings. Insights Imaging 7:69–76 Al Otieschan AA, Saleem M, Manohar MB, Larson S, Atallah A (1998) Malignant schwannoma of the parapharyngeal space. J Laryngol Otol 112:883–887 Farina D, Hermans R, Lemmerling M, Op de beeck K (1999) Imaging of the parapharyngeal space. J Belg Radiol 82:234–240 Fernandes T, Lobo JC, Castro R, Oliveira MI, Som PM (2013) Anatomy and pathology of the masticator space. Insights Imaging 4:605–616 Grégoire V, Levendag P, Ang KK et al (2003) CT-based delineation of lymph node levels and related CTVs in the node-negative neck: DAHANCA, EORTC, GORTEC, NCIC, RTOG consensus guidelines. Radiother Oncol 69:227–236 Heimgartner S, Zbaeren P (2009) Thyroid carcinoma presenting as a metastasis to the parapharyngeal space. Otolaryngol Head Neck Surg 140:435–436 Jacob JK, George S, Roy BR, Preethi S, Ranjith VT, Pappachan JM (2007) Retropharyngeal bronchogenic cyst. Otolaryngol Head Neck Surg 136:1025–1026

251

b

ing a tumor originating from the deep parotid lobe. The gadolinium-enhanced T1-weighted image (b) better shows the solid tumor component in the parotid gland (arrows). Transoral aspiration of the cystic component revealed hemorrhagic fluid. The patient was lost to follow-up; most likely, this lesion corresponds to a pleiomorphic adenoma containing hemorrhagic component in the PPS

Kahraman A, Yildirim I, Kilic MA, Okur E, Demirpolat G (2009) Horner’s syndrome from giant schwannoma of the cervical sympathetic chain: case report. B-ENT 5:111–114 Kakani RS, Bahadur S, Kumar S, Tandon DA (1992) Parapharyngeal lipoma. J Laryngol Otol 106:279–281 Kato H, Kanematsu M, Mizuta K et  al (2010) “Flow-­ void” sign at MR imaging: a rare finding of extracranial head and neck schwannoma. J Magn Reson Imaging 31:703–705 Lombardi D, Nicolai P, Antonelli AR, Maroldi R, Farina D, Shaha AR (2004) Parapharyngeal lymph node metastasis: an unusual presentation of papillary thyroid carcinoma. Head Neck 26:190–196 Luna-Ortiz K, Navarrete-Alemán JE, Granados-García M, Herrera-Gómez A (2005) Primary parapharyngeal space tumors in a Mexican cancer center. Otolaryngol Head Neck Surg 132:587–591 Maroldi R, Battaglia G, Maculotti P, Bondioni MP, Gheza G, Chiesa A (1994) Diagnostica per immagini delle lesioni dello spazio parafaringeo. In: Mosciaro O (ed) I Tumori parafaringei. Printed by Studio Forma, Verona, pp 69–103 Miller FR, Wanamaker J, Lavertu P, Wood BG (1996) Magnetic resonance imaging and the management of parapharyngeal space tumors. Head Neck 18:67–77

252 Mukherji SK, Castillo M (1998) A simplified approach to the spaces of the suprahyoid neck. Radiol Clin North Am 36:761–780 Nasser JG, Attia EL (1990) A conceptual approach to learning and organizing the surgical anatomy of the skull base. J Otolaryngol 19:114–121 Olsen KD (1994) Tumors and surgery of the parapharyngeal space. Laryngoscope 104:1–28 Pang KP, Goh CH, Tan HM (2002) Parapharyngeal space tumours: an 18 year review. J Laryngol Otol 116:170–175 Piccin O, Cavicchi O, Caliceti U (2008) Branchial cyst of the parapharyngeal space: report of a case and surgical approach considerations. Oral Maxillofac Surg 12:215–217 Saito DM, Glastonbury CM, El-Sayed IH, Eisele DW (2007) Parapharyngeal space schwannomas:

R. Hermans p­reoperative imaging determination of the nerve of origin. Arch Otolaryngol Head Neck Surg 133:662–667 Smith JC, Snyderman CH, Kassam AB, Fukui MB (2006) Giant parapharyngeal space lipoma: case report and surgical approach. Skull Base 12: 215–220 Som PM, Curtin HD (1995) Lesions of the parapharyngeal space. Role of MR imaging. Otolaryngol Clin North Am 28:515–542 Swartz JD, Harnsberger HR, Mukherji SK (1998) The temporal bone: contemporary diagnostic dilemmas. Radiol Clin North Am 36:819–854 Yuen HW, Goh CH, Tan TY (2006) Enlarged cervical sympathetic ganglion: an unusual parapharyngeal space tumour. Singap Med J 47:321–323

Malignant Lesions of the Masticator Space Bela Purohit and Robert Hermans

Contents

Abstract

1  Introduction

 253

2  Imaging Techniques

 256

3  General Imaging Features of MS Masses

 257

4  S  pecific Imaging Features of Primary MS Malignancies

 258

5  S  pecific Imaging Features of Secondary MS Malignancies  265 6  Post-treatment Imaging

 269

7  B  enign Lesions Mimicking MS Malignancies

 273

8  Conclusion

 279

References

 280

The masticator space (MS) is a deep facial space which contains the mandibular ramus, muscles of mastication and the mandibular branch (V3) of the trigeminal nerve. Malignant tumours of the MS are commonly of mesenchymal origin, namely osteosarcomas, chondrosarcomas and rhabdomyosarcomas. Besides these sarcomas, Non-Hodgkin’s lymphoma (NHL) may also arise in the MS.  Secondary malignancies of the MS include invasion from surrounding space tumours and occasionally metastases from distant primary sites. This chapter describes the clinico-radiological features of MS malignancies. The role of cross-sectional imaging with CT and MRI is highlighted along with a brief mention of certain advanced imaging techniques such as diffusion-weighted imaging (DWI) in the characterisation of MS lesions. Finally, post-treatment imaging and benign mimics of malignant MS masses is briefly reviewed.

1  Introduction B. Purohit (*) Department of Neuroradiology, National Neuroscience Institute, Singapore, Singapore e-mail: [email protected] R. Hermans University Hospitals, KU Leuven, Leuven, Belgium

The MS are paired suprahyoid neck spaces, one on each side of the face, extending from the angle of the mandible up to the parietal calvarium. Each space is defined by the superficial layer of deep cervical fascia. At the lower border of the

Med Radiol Diagn Imaging (2020) https://doi.org/10.1007/174_2020_229, © Springer Nature Switzerland AG Published Online: 25 July 2020

253

254

mandible, the superficial layer of deep cervical fascia splits into two layers and continues superiorly: the inner layer covers the medial pterygoid muscle and attaches to the skull base medial to the foramen ovale, whereas the outer layer covers the masseter muscle, passes lateral to the zygomatic arch and continues superiorly to encase the temporalis muscle. These two layers fuse at the anterior and posterior borders of the mandibular ramus, thus enclosing the space. The temporal fossa or suprazygomatic MS is the superior extent of the MS above the zygomatic arch whereas the infratemporal fossa or nasopharyngeal MS is the extension of the MS inferior to the skull base, between the pterygopalatine fossa medially and the zygomatic arch laterally. Since there is no fascial division between these two subspaces, pathological spaces can spread contiguously from one subspace to another (Chong and Fan 1996; Connor and Davitt 2004; Fernandes et  al. 2013; Razek 2014). In cross-section, the

Fig. 1 Graphic representation of MS anatomy demonstrating its contents and its relation to the parapharyngeal space (PPS) and parotid space (PS)

B. Purohit and R. Hermans

buccal space lies anterior to the infratemporal MS, the parotid space lies posteriorly and the parapharyngeal space lies posteromedially (Figs. 1 and 2) (Chong and Fan 1996; Faye et al. 2009). The MS contains the four muscles of mastication (masseter, temporalis, medial and lateral pterygoid). The V3 enters the masticator space through the foramen ovale, passes between the lateral and medial pterygoid muscles and finally enters the mandible through the mandibular foramen (Figs. 1 and 2). It provides motor innervation to the muscles of mastication and carries sensory information from the lower face, mandibular teeth, gums and anterior two-thirds of tongue (Czerny and Saat 2012; Fernandes et al. 2013; Razek 2014). Primary malignancies of the MS are rare. They are mostly of mesenchymal origin. These sarcomas may be of mandibular bony origin (e.g. osteosarcoma, chondrosarcoma) or may arise

Malignant Lesions of the Masticator Space

255

a

b

c

d

Fig. 2 Topographic anatomy of the MS.  The MS (enclosed by dotted margins) as seen on an axial T2W MR image (a). Skull-base attachments of the MS and surrounding spaces, namely, PS, PPS, pharyngeal mucosal space (PMS) and retropharyngeal space (RPS) as highlighted on a bone-window CT image (b). The foramen

ovale (arrow) opens into the MS. Coronal T1W MR image (c) demonstrates the contiguity between the suprazygomatic MS (thin arrow, enclosed by dotted margins) and infrazygomatic MS (thick arrow, enclosed by dotted margins). Also note the course of V3 (arrows) within the MS on coronal T1W MR image (d)

from muscles (rhabdomyosarcoma), fascia (fibrosarcoma) or from the V3 nerve sheath (malignant peripheral nerve sheath tumour). NHL is also a known primary of the MS. Secondary involvement of the MS from surrounding space malignancies (especially oral cavity, oropharynx, parotid space) is much more frequently seen. Finally, the MS may also be a site for metastasis from distant primary tumours

such as lung, breast and thyroid cancer (Chong and Fan 1996; Connor and Davitt 2004; Galli et al. 2010; Razek and Huang 2011; Czerny and Saat 2012; Fernandes et al. 2013; Razek 2014). Patients with MS malignancies usually present with facial swelling, pain and trismus. Trismus may be due to involvement of the temporomandibular joint or the muscles of mastication by pathology. Perineural tumour spread

256

(PNS) along V3 typically causes numbness/dysaesthesias along its course. The inherent deep-­ seated location of the MS further compounded by trismus often renders physical examination difficult. Hence cross-sectional imaging with CT and MRI plays a vital role in the detection and characterisation of MS lesions. The clinician needs to know the nature of the lesion in the MS, its localisation and extent, presence of neurovascular invasion, etc. so as to decide optimum therapeutic modalities. An early and correct differentiation of MS malignancy from the more commonly occurring infection is pivotal for treatment planning and favourable prognosis (Faye et al. 2009; Galli et al. 2010; Czerny and Saat 2012; Razek 2014).

2  Imaging Techniques MRI is the first-line study for the evaluation of suspected MS mass lesions, especially with strong suspicion of malignancy. CT is preferred in patients with suspected inflammatory disease or in patients with contraindications to MRI. CT is better than MR for detecting subtle erosions of the mandibular cortex and for excellent depiction of tumour matrix mineralisation. MRI has higher soft tissue contrast resolution whereby it better detects muscle and fascial invasion by tumours. It also provides better assessment of mandibular medullary disease and PNS (Fernandes et  al. 2013; Razek 2014). Multidetector CT for evaluation of MS lesions is ideally performed using iodinated intravenous contrast material. Imaging of the MS should include the suprazygomatic portion of the MS down to the submandibular space as well as the intracranial course of V3. Scans are typically acquired in the axial plane from the suprasellar region down to the base of the neck. Images are obtained in soft-tissue windows at 3 mm or less section thickness, without interslice gap. Coronal as well as sagittal reconstructions may be obtained and also bone window settings applied (Chong and Fan 1996; Galli et al. 2010; Czerny and Saat 2012; Razek 2014). MRI is performed with a dedicated head and neck coil with images mainly acquired in the

B. Purohit and R. Hermans

axial and coronal planes. Pre-contrast T1W, T2W, STIR or fat-supressed sequences are followed by post-contrast T1W sequences, the coronal plane being vital for assessing V3. Section thickness of 3–4 mm and matrix of 512 × 512 are recommended (Chong and Fan 1996; Galli et al. 2010; Czerny and Saat 2012; Razek 2014). In recent times, certain advanced imaging techniques such as DWI, perfusion MRI, MR spectroscopy (MRS) and fluorodeoxyglucose-­ positron emission tomography/CT (FDG-PET/ CT) have been utilised for their adjuvant role in the characterisation of deep neck space lesions, thereby adding to the diagnostic armamentarium (van Rijswijk et al. 2004; Yu et al. 2008; Razek et al. 2011; Razek and Huang 2011; Wang et al. 2010; Razek and Nada 2013). Multislice echoplanar DWI with quantitative ADC measurement has been described as a promising tool in the differentiation of MS malignancy from infection. Correlating with their high cellularity, malignant MS tumours show low ADC values as compared to benign tumours. In one study, the mean ADC value of MS malignancies (0.9  ×  10−3  mm2/s) was found to be significantly lower than that of MS infection ­ (1.59  ×  10−3  mm2/s) (Razek and Nada 2013). ADC values below 1.20 × 10−3 mm2/s and below 1.4 × 10−3 mm2/s have been described in two separate studies as optimal thresholds for differentiating MS malignancy from infection. The drawbacks of DWI include some overlap between ADC values of inflammatory conditions and malignant tumours and also instances of poor image quality due to susceptibility artefacts inherent to this region (Wang et al. 2010; Razek and Nada 2013). Literature search showed a single study describing the potential use of MRS in differentiating between infections and malignancies of the MS. The authors have described a significant difference in choline signal and choline/creatine ratios between malignant tumours and inflammatory lesions of the MS (Yu et al. 2008). However more studies are necessary to establish the clinical value of this technique. Dynamic contrast-enhanced MRI and dynamic susceptibility-weighted MRI also show promise

Malignant Lesions of the Masticator Space

for characterising soft tissue tumours. Malignant tumours, on account of their high vascularity and angiogenesis, tend to show earlier and faster uptake of contrast as compared to benign tumours. In fact, one study quotes that rapidity of tumour enhancement on dynamic contrast MRI may be one of the most robust parameters for identifying malignancy in soft tissue tumours, along with necrosis, cystic degeneration and lesion size (van Rijswijk et al. 2004). At dynamic susceptibility-weighted MRI, a threshold dynamic susceptibility contrast ratio of 30.7% has been described to show 84.6% accuracy for differentiating between benign and malignant head-neck tumours (Razek et al. 2011). The role of FDG-PET/CT in the evaluation of head neck malignancies is described in detail in the dedicated chapter on “Positron Emission Tomography in Head and Neck Cancer”. Suffice to say that FDG-PET is now widely used for the pretreatment staging/grading and post-treatment follow-up of head neck tumours like squamous cell carcinoma (SCC), sarcomas and lymphomas, which show high FDG uptake (Purohit et al. 2014). It must be kept in mind that although the above-mentioned adjuvant techniques like DWI

Fig. 3 Graphic representation of vector of displacement (arrow) of the fat-filled PPS by a mass in the MS

257

and FDG-PET/CT can be very useful in differentiating between benign and malignant tumours, they do not allow tissue-specific diagnosis and ultimately biopsy may be required. Tissue sampling remains the reference standard for the definitive diagnosis of deep neck space masses. MS masses are easily amenable to CT-guided biopsy using either a subzygomatic or paramaxillary approach (Gupta et  al. 2007; Razek and Huang 2011).

3  G  eneral Imaging Features of MS Masses The most important aim of imaging is to accurately depict the extent and epicentre of a mass lesion, i.e. if it primarily originates within the MS or infiltrates it from surrounding spaces. The anatomical origin of small lesions within the MS is easily determined by both CT and MR.  The origin of a large mass may be difficult to assess, the key rule to follow is that MS masses cause posteromedial displacement of the fat-filled parapharyngeal space (Fig. 3). However a very large or infiltrative neoplasm may completely efface

258

the parapharyngeal space (Chong and Fan 1996; Czerny and Saat 2012). Similar to the peripheral skeleton, malignant tumours arising within the mandible tend to show either moth-eaten or permeative bony destruction with a wide zone of transition. Benign lesions on the other hand show geographic destruction with well-demarcated margins/narrow zone of transition. Bulging or scalloping of the cortex without periosteal reaction indicates a benign bony lesion whereas cortical disruption with a ‘sunburst’ periosteal reaction or Codman triangle formation indicates malignancy (Wood et al. 1990; Ida et al. 1997; Lopes et al. 2007; Thariat et al. 2012; Avril et al. 2014). Dental radiographic findings suspicious for malignancy include destruction of unerupted tooth follicles, widening of periodontal ligament space, root resorption and widening/ resorption of inferior alveolar nerve canal (Chittaranjan et  al. 2014). Along with bone destruction, primary osseous tumours often show an associated soft tissue component, with or without a calcified matrix (Razek and Huang 2011; Thariat et  al. 2012; Krishna et  al. 2013; Razek 2014). Soft tissue sarcomas of the MS often show non-specific imaging findings. They usually appear as infiltrative masses. Interruption of surrounding fascia and mandibular erosion are key features indicating malignancy. Other radiological findings include inhomogeneous MR signal, intratumoural haemorrhage or necrosis, neurovascular invasion and marked rapid enhancement (Som et  al. 1997; Galli et  al. 2010; Fernandes et al. 2013; Razek and Huang 2011). Metastases to the MS may show lytic or sclerotic bony destruction without significant periosteal reaction and/or associated soft tissue masses (Dunfee et al. 2006; Agrawal and Nair 2007; Lin et al. 2009; Avril et al. 2014; Sahoo et al. 2013; Razek 2014). PNS along V3 is seen as thickening and increased contrast enhancement of the nerve and loss of fat pads at the foramen ovale, often with foraminal enlargement. PNS along V3 is best depicted on coronal T1W post-contrast sequence with fat saturation. PNS may lead to denervation atrophy of the masticator muscles which is seen as oedematous changes and increased contrast

B. Purohit and R. Hermans

enhancement in the acute phase followed by fatty atrophy (Connor and Davitt 2004; Czerny and Saat 2012; Moonis et al. 2012; Fernandes et al. 2013; Razek 2014). In addition to all the above-mentioned imaging features, pertinent clinical data like age, duration/rapidity of symptom onset, presence of pain, history of irradiation at site of tumour, history of known primary, etc. help to narrow down the list of differential diagnoses (Razek and Huang 2011).

4  S  pecific Imaging Features of Primary MS Malignancies Osteosarcoma is the most common primary bone tumour of the MS. Jaw osteosarcomas differ from osteosarcomas at other skeletal regions due to later development (usually in the second to fourth decades), high mortality associated with local disease but less propensity for distant metastases. They may arise de novo or secondary to previous irradiation. Extraskeletal osteosarcomas may also arise in the MS. Clinically jaw osteosarcomas present with pain, swelling, loosening of teeth or frank ulceration. Radiographic features include bony destruction, admixed with sclerosis, and an associated soft tissue mass often with osteoid/amorphous pattern of matrix mineralisation. Codman’s triangle formation and classic sunburst or ‘hair-on-end’ periosteal reactions are often seen (Fig.  4). Rare instances have been described of osteosarcomas mimicking benign cystic lesions of the mandible (Fig. 5) (Anil et al. 2012). On MRI, the involved medullary cavity and associated soft tissue mass show low to intermediate T1 signal, high T2/STIR signal and heterogeneous post-contrast enhancement. Calcification and new bone formation appear as signal voids (Ida et  al. 1997; Razek and Huang 2011; Kammerer et al. 2012; Thariat et al. 2012; Fernandes et  al. 2013; Chittaranjan et  al. 2014; Razek 2014). Ewing’s sarcoma (ES) is a small round cell tumour mainly affecting children and young adults. Only 1–4% cases occur in the head and neck with posterior mandibular predilection. An aggressive course with rapid growth and

Malignant Lesions of the Masticator Space

259

b

a

c

d

Fig. 4  Osteosarcoma of the left mandibular ramus in a 14-year-old male patient. Coronal T1W MR image (a) demonstrates hypointense signal in the involved bone and in the associated soft tissue mass (arrow). Note the characteristic sunburst periosteal reaction (thin arrows) within the tumour (thick arrow) on a coronal cone-beam CT reconstruction (b). Moderate contrast enhancement is

seen within the tumour (thick arrow) on post-contrast axial T1W MR image (c). Again, note the sunburst periosteal reaction seen as hypointense radiating spicules (thin arrow) within the tumour matrix. Volume-rendered CT reconstruction of the face (d) shows extensive destruction of the left mandibular ramus by the tumour

propensity to haematogenous metastases are dominant features of ES.  Clinically, patients often present with jaw swelling, pain, fever and leucocytosis, thereby causing suspicion for odontogenic infection. Classic radiographic findings of ES include a permeative/moth-eaten bony destruction with an aggressive periosteal reaction

(spiculated type more common than onion-peel type in the jaw) and a large associated soft tissue mass (Fig. 6). In children less than 5 years of age, these imaging features may also represent metastatic neuroblastoma or acute leukaemia. On CT, the soft tissue component of ES shows non-specific low-­intermediate attenuation

B. Purohit and R. Hermans

260

a

c

b

d

Fig. 5 A 12-year-old male patient with a well-­ circumscribed lytic lesion (arrow) in the right mandibular ramus on a panoramic radiograph (a). Coronal contrast-­ enhanced soft-tissue window CT image (b) and coronal bone-window CT image (c) show the lesion (arrow) to be expansile-lytic with no new bone formation or aggressive

periosteal reaction. The lesion shows bubbly appearance on volume-rendered CT reconstruction of the face (d). The diagnosis of a benign odontogenic cyst was made on the basis of imaging findings. Surprisingly it turned out to be a low-grade osteosarcoma on biopsy

without a calcified matrix. The MRI features include low T1 and high T2/STIR signal in the medullary cavity of the involved mandible, as well as an associated soft tissue mass. Both the bony and soft tissue component show post-contrast enhancement. Intratumoural haemorrhage/ necrosis, prominent intratumoural vascular channels and oedematous changes in the masticator muscles may be seen (Wood et al. 1990; Gorospe et al. 2001; Lopes et al. 2007; Krishna et al. 2013; Razek 2014; Deore et al. 2015). Chondrosarcoma is a slow-growing malignant cartilaginous tumour manifesting in the fourth

and fifth decades of life. Only 1–3% chondrosarcomas occur in the maxillofacial region. Clinically, mandibular chondrosarcomas present as a painless jaw swelling with expansion of the cortical plates, egg-shell crackling and often premature exfoliation of teeth. Radiographically, these tumours show endosteal scalloping and patchy radiolucent-sclerotic/ground glass appearance, the sclerosis being from chondroid matrix. Well-differentiated tumours demonstrate typical stippled and curvilinear chondroid calcifications within the tumour matrix. Aggressive tumours cause frank cortical destruction and often a

Malignant Lesions of the Masticator Space

sunburst periosteal reaction. CT shows low attenuation and MR shows characteristic bright T2 signal of the chondroid matrix. Low-grade tumours show a lobulated pattern with curvilinear septal and peripheral enhancement of fibrovascular t­issue and non-ossified cartilage, a pattern described as ‘ring and arc’. Higher grade variants do not have septations and show more

261

diffuse, heterogeneous enhancement with more extensive involvement of adjacent soft tissue (Murphey et al. 2003; Ollivier et al. 2004; Kundu et al. 2011; Razek 2014). Plasma cell neoplasms like plasmacytoma may involve the mandible causing lytic destruction without periosteal reaction. Soft tissue plasmacytomas can also occur in the MS and are seen

a

b

c

d

Fig. 6  Ewing sarcoma of the right mandibular ramus in a 12-year-old female patient. Note the lytic bony destruction with an associated soft tissue mass (thick arrow) and spiculated periosteal reaction (thin arrow) on axial non-­ contrast CT image (a). The mass shows mild hyperintense signal on axial T2W MR image (b), significant hyperintense signal on fat-saturated axial T2W MR image (c) and

moderate post-contrast enhancement (d). Restricted diffusion within the tumour on DWI/ADC images (e) indicates its malignant nature. Three years after surgery and chemotherapy, the patient developed recurrence in the form of meningeal metastasis (dotted arrow) as seen on this contrast-­enhanced coronal T1W MR image (f)

B. Purohit and R. Hermans

262

e

f

Fig. 6 (continued)

Malignant Lesions of the Masticator Space

a

263

b

c

Fig. 7  Fibrosarcoma of the left MS in a 22-year-old male patient. The infiltrative mass (arrow) appears mildly hyperintense on axial T2W MR image (a). Bony destruction of the left pterygoid plate and processes (thin arrow)

is seen on bone-window axial CT image (b). Contrast-­ enhanced fat-saturated coronal T1W MR image (c) depicts intracranial extension of the tumour into the left middle cranial fossa (dotted arrow)

as enhancing soft tissue masses, sometimes showing restricted diffusion due to high cellularity. Multiple myeloma manifests as classic ‘punched out’ lytic lesions, typically in the angle and ramus of the mandible, again without periosteal reaction. Multiple myeloma is associated with similar lesions in other long bones (Dunfee et  al. 2006; Czerny and Saat 2012; Fernandes et al. 2013; Razek 2014; Dayisoylu et al. 2016).

Intraosseous mucoepidermoid carcinomas and ameloblastic carcinomas are very rare primary mandibular tumours which may also cause expansile lytic bony destruction (Dunfee et  al. 2006; Fernandes et al. 2013; Avril et al. 2014). Rhabdomyosarcoma, synovial sarcoma, fibrosarcoma, malignant fibrous histiocytoma, etc. are some of the wide variety of soft tissue sarcomas which can arise in the MS (Fig.  7). Of these,

B. Purohit and R. Hermans

264

rhabdomyosarcoma is the most common and generally seen in children (Fig. 8). None of these tumours have specific imaging features. As

mentioned before, they usually manifest as infiltrative soft tissue masses with disruption of fascia and erosion of bone. However intact fascia does

a

b

c

d

Fig. 8  Rhabdomyosarcoma of the left MS in a 6-year-old male patient. The large poorly marginated tumour (arrow) shows intermediate signal on axial T2W MR image (a). Extensive infiltration of surrounding spaces is seen. Note associated left posterior cervical lymphadenopathy (dotted arrow). The tumour shows heterogeneous enhancement on contrast-enhanced axial T1W MR image (b) and restricted diffusion on DWI (c). Bony destruction of the left pterygoid plates (thin arrow) and left mandibular ramus (thick arrow) is seen on axial bone-window CT

image (d). The tumour showed very good response to chemotherapy, with drastic reduction in size. Only a small, rim-enhancing, centrally necrotic residual lesion (arrow) is seen on the contrast-enhanced axial T1W MR image (e) obtained after three cycles of chemotherapy. Two years post chemotherapy, the patient developed recurrent disease in the left MS (straight arrow) along with intracranial extension (stepped arrow), as seen on contrast-enhanced coronal T1W MR image (f)

Malignant Lesions of the Masticator Space

e

265

f

Fig. 8 (continued)

not exclude malignancy and biopsy is eventually required (Koga et  al. 2005; Gosau et  al. 2008; Park et al. 2009; Freiling et al. 2010; Razek and Huang 2011; Fernandes et al. 2013; Razek 2014). Malignant peripheral nerve sheath tumours are rare high-grade sarcomas that can arise de novo or from pre-existing benign neurofibromas or schwannomas. They are commonly associated with neurofibromatosis-1. The inferior alveolar nerve is one of the favoured sites in the head and neck. It is often difficult to differentiate malignant from benign nerve sheath tumours. Large tumour size >5  cm, rapid growth, illdefined infiltrative margins, heterogeneous MR signal and erosion of skull base foramina out of proportion to tumour volume suggest malignant nature (Gheisari and Roozbehi 2010; Fernandes et al. 2013; Razek 2014). NHL may involve the mandible or the muscles of mastication in adults as well as children (Fig.  9). Burkitt’s lymphoma typically involves the jaws in children (Fig. 10). Mandibular involvement by lymphoma may clinically mimic odontogenic infection, often with pain, jaw swelling and teeth loosening. Radiographically, ill-­ defined

lucencies are seen within the mandible often without frank cortical destruction. Associated soft tissue masses may be especially large in Burkitt’s lymphoma. On MRI, lymphomatous lesions appear hypointense on T1, intermediate on T2, bright on STIR with solid homogeneous enhancement and low ADC value (typically 0.6–0.8  ×  10−3  mm2/s). Necrosis is uncommon unless chemotherapy is administered. There may be associated neck nodal disease, extranodal lymphatic site involvement (Waldeyer’s ring) or multiple other extranodal extralymphatic site involvement (e.g. orbit, salivary glands) (Chong et al. 1998; Imaizumi et al. 2012; Fernandes et al. 2013; Avril et  al. 2014; Razek 2014; Bugshan et al. 2015; Derinkuyu et al. 2016).

5  S  pecific Imaging Features of Secondary MS Malignancies Tumour extension from adjacent anatomical spaces is the most common malignancy of the MS.  SCC of the oral cavity, oropharynx and malignant parotid gland tumours are the most

B. Purohit and R. Hermans

266

a

b

c

Fig. 9  NHL of the left MS in a 30-year-old male patient. The tumour (arrow) involves the left mandibular ramus with marrow infiltration, cortical destruction and surrounding soft tissue infiltration as seen on contrast-­ enhanced axial CT image (a). An enhancing nodule is seen within the left parotid gland (dotted arrow), likely

intra-parotid lymphadenopathy. There is also associated left submandibular lymphadenopathy (dotted arrow) on a lower section image (b). Axial bone-window CT image (c) shows widening and erosion of the left mandibular canal (arrow) suspicious for PNS along the left inferior alveolar nerve

Malignant Lesions of the Masticator Space

267

a

c

Fig. 10  Burkitt’s lymphoma of the mandible in a 7-year-­ old male patient. The tumour causes lytic destruction of the right mandibular ramus with an associated soft tissue mass (arrow) as seen on contrast-enhanced axial (a) and coronal soft-tissue window CT image (b) and coronal

b

d

bone-window CT image (c). After 1 month of chemotherapy, there was almost complete resolution of the tumour, as depicted on contrast-enhanced axial and coronal soft tissue window images (d, e) and axial bone-window CT image (f)

B. Purohit and R. Hermans

268

e

f

Fig. 10 (continued)

common extrinsic tumours involving the MS. SCC arising from the gingival and buccal space usually spreads along the ramus of the mandible with bony destruction or may extend directly into the MS through the fat between the medial pterygoid muscle and the ramus, with or without bony destruction. Also, these tumours may extend perineurally along V3. SCC of the retromolar trigone easily reaches the MS via the pterygomandibular raphe. Tumour involvement of the MS in oral cavity SCC is staged as very advanced local disease (T4b) as per the recent eighth edition of AJCC (Fig. 11). Parotid malignancies extending into the soft tissue of the MS are staged as T3 whereas mandibular involvement is staged as T4. Parotid gland adenoid cystic carcinomas are especially notorious for PNS along V3 (Fig.  12) (Wei et  al. 2007; Galli et  al. 2010; Czerny and Saat 2012; Fernandes et al. 2013; Razek 2014; Huang and O’Sullivan 2017). Nasopharyngeal carcinoma (NPC) involves the MS in almost 20% cases. It spreads through the sinus of Morgagni, a natural defect in the pharyngobasilar fascia, to invade the parapharyngeal space and MS. In more advanced cases, tumours

breach the pharyngobasilar fascia and destroy pterygoid plates. NPC extension into the medial and lateral pterygoid muscles is classified as T2, infiltration of the pterygoid plates is classified as T3 and extensive soft tissue infiltration beyond the lateral pterygoid muscle is classified as T4 (Fig.  13) (Hyare et  al. 2010; Razek and King 2012; Fernandes et al. 2013; Razek 2014; Huang and O’Sullivan 2017). Some authors describe poorer overall survival in NPC patients with lateral MS involvement as compared to medial MS involvement (Zhang et al. 2014; Luo et al. 2014). Metastatic lesions in the MS usually involve the posterior body and ramus of mandible (Fig. 14), but may also occur entirely within soft tissue (Fig.  15). The most common primary sites include lungs, breasts, kidneys, prostate and thyroid. The majority of metastatic lesions are osteolytic with ill-defined margins (often without periosteal reaction). Soft tissue extension is common. Osteoblastic and mixed lyticsclerotic patterns are seen in prostate and breast cancers, respectively. (Dunfee et  al. 2006; Agrawal and Nair 2007; Lin et al. 2009; Sahoo et al. 2013; Fernandes et al. 2013; Razek 2014; Avril et al. 2014).

Malignant Lesions of the Masticator Space

269

a

b

c

d

Fig. 11  Secondary invasion of the MS from oral cavity and oropharyngeal SCC in three separate cases, all indicating T4 disease. First case (a) depicts SCC of the right maxillary tuberosity (asterisk) infiltrating into the right MS (arrow) on contrast-enhanced axial CT image. Another case (b) of extensive SCC arising in the left oropharyngeal

wall (asterisk) extending into the left MS (arrow) on contrast-enhanced axial CT image. Another case (c, d) of SCC of the left retromolar trigone (asterisk) infiltrating into the left MS (arrow) and also showing PNS along the left inferior alveolar nerve (dotted arrow) on contrastenhanced fat-saturated axial T1W MR images (c, d)

6  Post-treatment Imaging

surgery, chemotherapy or chemoradiotherapy, or a combination of these. Post-treatment imaging is crucial for assessment of residual/ recurrent disease, as well as for detecting certain treatment-­ related complications, like

The treatment options for MS malignancies depend mainly upon the histopathology and extent of the tumour. These options include

B. Purohit and R. Hermans

270

a

Fig. 12  Secondary invasion of the MS from adenocarcinoma of the right parotid gland in a 55-year-old male patient. Contrast-enhanced axial (a) and coronal (b) T1W MR images demonstrate the parotid tumour (asterisk)

a Fig. 13  Secondary invasion of the MS from NPC in two separate cases. First case (a) depicts an aggressive NPC (asterisk) widening the right pterygopalatine fossa (thin arrow) with extensive involvement of the right MS (thick arrow) on contrast-enhanced fat-saturated axial T1W MR image. Another case (b) of left-sided NPC (asterisk)

b

infiltrating into the right MS and showing PNS along right V3 (arrow). The nerve is significantly thickened with increased enhancement. Note the denervation atrophy of the right masticator muscles (dotted arrow)

b infiltrating the left MS (arrow) and extending perineurally via left V3 to the left foramen ovale (dotted arrow) on contrast-­enhanced fat-saturated coronal T1W MR image. Note necrotic left retropharyngeal lymphadenopathy (arrowhead)

Malignant Lesions of the Masticator Space

a

271

b

c

Fig. 14  Bony metastases to the mandible from small-cell carcinoma of the lung in a 43-year-old male patient. Extensive bony destruction of the left mandibular ramus is seen with an associated heterogeneously enhancing soft

tissue mass (arrow) on axial T1W (a) and post-contrast axial T1W MR image (b). Note the irregular cortical destruction (arrow) on axial bone-window CT image (c)

osteoradionecrosis of the mandible (Czerny and Saat 2012; Razek 2014). MRI is the favoured modality for evaluating post-treatment response. Substantial reduction of both soft tissue and bony tumour volume, with decreased T2 signal and resolution of contrast enhancement, indicates good histological response to therapy (i.e. evolution into fibrosis/scar tissue) (Hermans 2008; Gorospe et  al. 2001; Razek 2014).

Quantitative ADC has also been described as a useful parameter for differentiating between residual/recurrent tumour and post-­ operative/ post-radiation changes. ADC value of 1.30  ×  10−3  mm2/s has been described as a threshold below which lie residual/recurrent tumours (Razek et al. 2007). FDG-PET/CT also aids in differentiating residual/recurrent tumour from fibrosis (Purohit et al. 2014).

B. Purohit and R. Hermans

272

a

b

c

d

Fig. 15  Metastasis to the right masseter muscle from primary melanoma of the foot in a 60-year-old female patient. (a) Axial T1W MR image shows a well-­ circumscribed hypointense mass (arrow) involving the right masseter muscle. The mass shows high FDG uptake

on axial FDG-PET image (b). After 1 year, the metastatic mass (arrow) showed interval increase in size, as seen on contrast-enhanced axial (c) and coronal (d) T1W MR images. Note appearance of brain parenchymal metastasis (dotted arrow)

Osteoradionecrosis of the mandible may occur after irradiation. Imaging findings include areas of mixed osteolysis-sclerosis within the irradiation portal. Fragmentation and sloughing of necrotic bone may occur. Adjacent masticator muscles may show T2 hyperintensity, diffuse enhancement or mass-like thickening,

thereby mimicking tumour recurrence or osteomyelitis. CT is valuable in detecting cortical disruption, trabecular disorganisation and possible pathological fractures (Hermans 2003; Bharatha et al. 2012; Razek 2014). Radiation-induced sarcomas can arise within a high-dose radiation field, approximately

Malignant Lesions of the Masticator Space

5–10  years later after radiation. They may have varied histologies including osteosarcoma, malignant fibrous histiocytoma, etc. New appearance of a rapidly growing large/heterogeneous destructive mass (within the radiation field and after sufficient latent period) that displays different signal intensity from the previous primary tumour should raise suspicion for a radiation-­induced sarcoma. The presence of calcification or ossification within these tumours is a strong corroborative finding (Bharatha et al. 2012; Razek 2014). Bisphosphonate-related osteonecrosis of the jaw is characterised by non-healing exposed jaw bones in patients who have undergone bisphosphonate treatment for osteoporosis or bone metastasis. CT shows osteolytic as well as osteosclerotic lesions, bone fragmentation, sequestration and periosteal reaction. Imaging findings almost mimic osteoradionecrosis, but the clinical context is different (Morag et  al. 2009; Razek 2014; Avril et al. 2014).

273

a

b

7  B  enign Lesions Mimicking MS Malignancies Sometimes benign conditions may mimic MS malignancies on imaging. A Stafne’s cyst, also called as salivary gland inclusion defect, is often detected incidentally as a radiolucent defect (usually less than 2 cm in size) at the lingual cortex of the posterior mandible. It contains fat or salivary gland tissue. Panoramic radiographs and CT suffice for its diagnosis. Odontogenic tumours, such as keratocystic odontogenic tumours and ameloblastomas, often extend into the ramus of the mandible. They commonly cause soap-bubble-­ like cortical expansion and well-circumscribed osteolysis of the mandible (Fig. 16). Radiographic and clinical findings in conjunction help to identify the odontogenic origin of these lesions (Dunfee et al. 2006; Avril et al. 2014; Czerny and Saat 2012). Langerhans cell histiocytosis (LCH), typically seen in children and young adults, can cause lytic destruction of the mandible along with an enhancing soft tissue mass, thereby mimicking a malignant tumour (Fig. 17). These confounding lesions

c

Fig. 16  Odontogenic keratocyst in a 27-year-old male patient. A well-circumscribed lytic lesion (arrow) detected in the right mandibular ramus on panoramic radiograph (a). The non-enhancing lesion (arrow) shows typical expansile lytic appearance with scalloped cortices on contrast-­enhanced axial CT image (b) and axial bone-­ window image (c)

B. Purohit and R. Hermans

274

a

d

b

e

c

Fig. 17  A 20-year-old female patient detected with a well-circumscribed lytic lesion (arrow) in the left mandibular ramus on a panoramic radiograph (a). Contrastenhanced axial CT image (b) shows lytic destruction of the left mandibular ramus with an associated mildly enhancing soft tissue mass (arrow). A lamellated periosteal reaction (dotted arrow) is seen on axial bone-window

CT image (c). The involved medullary cavity and soft tissue mass (arrow) both show hypointense signal on axial T1W MR image (d) and heterogeneous enhancement on contrast-enhanced axial T1W MR image (e). Imaging differential diagnoses included sarcoma, lymphoma and Langerhans cell histiocytosis (LCH). This lesion turned out to be LCH on biopsy

Malignant Lesions of the Masticator Space

a

275

b

c

Fig. 18  Ossifying fibroma of the right mandibular ramus in a 35-year-old female patient. Axial non-contrast CT image (a) shows a large mixed lytic-sclerotic lesion (arrow) involving the right mandibular ramus. Axial bone-­ window CT image (b) shows focal cortical disruptions

(thick arrow) and also a pathological fracture through the lesion (thin arrow). The tumour matrix shows foci of ossification but no aggressive periosteal reaction. Volume-­ rendered CT reconstruction of the face (c) shows the bubbly appearance of the lesion

often end up being biopsied (Can et  al. 2005; Dunfee et al. 2006; Avril et al. 2014; Christopher et al. 2018). Fibro-osseous diseases like fibrous dysplasia and ossifying fibromas sometimes mimic malignancy, especially when they are associated with extensive cortical destruction, matrix calcifica-

tions and large soft tissue masses on imaging (Figs. 18 and 19). A biopsy may be required for histological diagnosis (Jung et  al. 1999; Singer et  al. 2004; Liu et  al. 2010; Kushchayeva et  al. 2018). Osteomyelitis of the mandible may clinically and radiologically mimic malignancy (Fig.  20).

B. Purohit and R. Hermans

276

a

b

d

d

Fig. 19  Intrinsic lesion of the right mandibular ramus in a 27-year-old female patient. Axial T2W MR image (a) and contrast-enhanced axial T1W MR image (b) show a heterogeneously enhancing mass (thick arrow) involving the right mandibular ramus and adjacent soft tissue. Note the hypointense areas (thin arrow) within the lesion, which corresponded to faint ossified foci on axial bone-­

window CT image (c). There are cortical disruptions and also suspicious spiculated-type periosteal reaction at the anterior aspect of the lesion (dotted arrow) on a lower section (d). Imaging differential diagnoses included osteosarcoma and aggressive fibro-osseous tumour. This lesion turned out to be fibrous dysplasia on biopsy

Malignant Lesions of the Masticator Space

277

a

b

c

d

Fig. 20  Chronic left-sided jaw swelling and trismus in a 34-year-old male patient. Axial T2W MR image (a) shows a destructive lesion in the left mandibular ramus (arrow) with involvement of adjacent soft tissue. Contrast-­ enhanced coronal T1W MR image (b) shows heterogeneous enhancement both within the involved medullary cavity and associated soft-tissue mass (arrow). Axial bone-window CT image (c) shows cortical erosions of the left mandibular ramus (arrow). However sagittal (d) and coronal (e) cone-beam CT reconstructions better

demonstrate the irregular lytic destruction (arrow) of the mandibular ramus admixed with patchy sclerosis. Note the involvement of the left mandibular canal (dotted arrow). No aggressive periosteal reaction seen. Note high 99mTc-­ MDP uptake in the left mandibular ramus on SPECT images (f). Biopsy confirmed imaging diagnosis of chronic osteomyelitis with scar tissue formation. The patient improved with antibiotics and hyperbaric oxygen therapy

B. Purohit and R. Hermans

278

e

f

Fig. 20 (continued)

This is especially so in chronic osteomyelitis which may cause sunburst periosteal reaction of the mandible. Laboratory data, clinical history along with imaging findings help to reach the correct diagnosis (Dunfee et al. 2006; Schuknecht and Valavanis 2003; Razek 2014; Avril et al. 2014). A vascular malformation within the mandible can cause osteolysis and associated soft tissue

mass. Vascular malformations may also arise within the soft tissue of the MS. The detection of multiple vascular channels within the lesion points to the correct diagnosis (Fig.  21). A biopsy should be avoided in these situations to avoid torrential bleeding (Dunfee et  al. 2006; Czerny and Saat 2012; Fernandes et  al. 2013; Razek 2014).

Malignant Lesions of the Masticator Space

a

279

b

c

Fig. 21  Right cheek swelling in a 14-year-old female. Axial T1W MR image (a) shows a hypointense mass involving the right masseter muscle (thick arrow) with heterogeneous T2 hyperintense signal on axial T2W MR image (b). Few serpiginous vascular channels (thin

arrows) are seen in the subcutaneous fat, anteromedial to the mass on both images (a) and (b). Time-of-flight source MR angiogram image (c) shows heterogeneous arterial phase enhancement within the mass (arrow) confirming the diagnosis of a high-flow arteriovenous malformation

8  Conclusion

along V3 is a crucial imaging finding with important clinical implications. Awareness of the ­pertinent imaging features of MS malignancies along with corroborative clinical data helps to narrow the differential diagnosis and aid in appropriate management.

In conclusion, MS space malignancies may be primary or secondary in nature. Adjacent space malignancies infiltrating the MS are commoner than primary tumours arising in the MS.  PNS

280

References Agrawal AR, Nair N (2007) Unusual metastasis of poorly differentiated thyroid carcinoma to the masticator space. Clin Nucl Med 32:516–518 Anil S, Krishnan AP, Rajendran R (2012) Osteosarcoma of the mandible masquerading as a dental abscess: report of a case. Case Rep Dent 2012:635062 Avril L, Lombardi T, Ailianou A et al (2014) Radiolucent lesions of the mandible: a pattern-based approach to diagnosis. Insights Imaging 5:85–101 Bharatha A, Yu E, Symons S et al (2012) Pictorial essay: early and late-term effects of radiotherapy in head and neck imaging. Can Assoc Radiol J 63:119–128 Bugshan A, Kassolis J, Basile J (2015) Primary diffuse large B-cell lymphoma of the mandible: case report and review of the literature. Case Rep Oncol 8:451–455 Can IH, Kurt A, Ozer E et  al (2005) Mandibular manifestations of Langerhans cell histiocytosis. Oral Oncol 41:174–177 Chittaranjan B, Tejasvi MLA, Geetha P (2014) Intramedullay osteosarcoma of the mandible: a clinicoradiologic perspective. J Clin Imaging Sci 4(Suppl 2):6 Chong VFH, Fan YF (1996) Pictorial review: radiology of the masticator space. Clin Radiol 51:457–465 Chong J, Som PM, Silvers AR et  al (1998) Extranodal non-Hodgkin’s lymphoma involving the muscles of mastication. Am J Neuroradiol 19:1849–1851 Christopher Z, Binite O, Henderson-Jackson E et al (2018) Langerhans cell histiocyosis of bone in an adult: a case report. Radiol Case Rep 13:310–314 Connor SEJ, Davitt SM (2004) Masticator space masses and pseudomasses. Clin Radiol 59:237–245 Czerny C, Saat R (2012) Malignant lesions of the masticator space. In: Hermans R (ed) Head and neck cancer imaging, medical radiology, diagnostic imaging. Springer-Verlag, Berlin, Heidelberg Dayisoylu EH, Ceneli O, Coskunoglu EZ (2016) Solitary plasmacytoma of the mandible: an uncommon entity. Iran Red Cerscent Med J 18:e22932 Deore S, Dandekar R, Mahajan A et  al (2015) Ewing’s sarcoma of the mandible: a case report presenting as odontogenic infection. J Oral Maxillofac Surg Med Pathol 27:741–745 Derinkuyu BE, Boyunaga O, Oztunali C et  al (2016) Imaging features of Burkitt lymphoma in pediatric patients. Diagn Interv Radiol 22:95–100 Dunfee BL, Sakai O, Pistey R et al (2006) Radiologic and pathologic characteristics of benign and malignant lesions of the mandible. Radiographics 26:1751–1768 Faye N, Lafitte F, Williams M et al (2009) The masticator space: from anatomy to pathology. J Neuroradiol 36:121–130 Fernandes T, Lobo JC, Castro R et al (2013) Anatomy and pathology of the masticator space. Insights Imaging 4:605–616

B. Purohit and R. Hermans Freiling NJ, Merks JH, Saeed P et al (2010) Imaging findings in craniofacial childhood rhabdomyosarcoma. Paediatr Radiol 40:1723–1738 Galli F, Flor N, Villa C et al (2010) The masticator space. Value of computed tomography and magnetic resonance imaging in the localisation and characterisation of lesions. Acta Otorhinolaryngol Ital 302:94–99 Gheisari R, Roozbehi A (2010) Malignant peripheral nerve sheath tumour of the infratemporal fossa. J Craniofac Surg 21:596–598 Gorospe L, Fernandez-Gil A, Garica-Raya P et al (2001) Ewing’s sarcoma of the mandible: radiologic features with emphasis on magnetic resonance appearance. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 91:728–734 Gosau M, Draenert FG, Winter WA et  al (2008) Fibrosacroma of the childhood mandible. Head Face Med 4:21 Gupta S, Henningsen JA, Wallac MJ et  al (2007) Percutaneous biopsy of head and neck lesions with CT guidance: various approaches and relevant anatomic and technical considerations. Radiographics 27:371–390 Hermans R (2003) Imaging of mandibular osteoradionecrosis. Neuroimaging Clin N Am 13:579–604 Hermans R (2008) Posttreatment imaging in head and neck cancer. Eur J Radiol 66:501–511 Huang SH, O’Sullivan B (2017) Overview of the 8th edition TNM classification for head and neck cancer. Curr Treat Options Oncol 18:40 Hyare H, Wisco JJ, Alusi G et al (2010) The anatomy of nasopharyngeal carcinoma spread through the pharyngobasilar fascia to the trigeminal mandibular nerve on 1.5 T. Surg Radiol Anat 32:937–944 Ida M, Tetsumura A, Kurabayashi T et al (1997) Periosteal new bone formation in the jaws. A computed tomographic study. Dentomaxillofac Radiol 26:169–176 Imaizumi A, Kuribayashi A, Watanabe A et  al (2012) Non-Hodgkin’s lymphoma involving the mandible: imaging findings. Oral Surg Oral Med Oral Pathol Oral Radiol 113:e33–e39 Jung SL, Choi KH, Park YH et  al (1999) Cemento-­ ossifiying fibroma presenting as a mass of the parapharyngeal and masticator space. Am J Neuroradiol 20:1744–1746 Kammerer PW, Shabazfar N, Makoie NV et  al (2012) Clinical, therapeutic and prognostic features of osteosarcoma of the jaws-experience of 36 cases. J Craniomaxillofac Surg 40:541–548 Koga C, Harada H, Kusukawa J et  al (2005) Synovial sarcoma arising in the mandibular bone. Oral Oncol Extra 41:45–48 Krishna KBB, Thomas V, Kattoor J et al (2013) A radiological review of Ewing’s sarcoma of the mandible: a case report with one year follow-up. Int J Clin Pediatr Dent 6:109–114 Kundu S, Pal M, Paul RR (2011) Clinicopathologic correlation of chondrosarcoma of mandible with a case report. Contemp Clin Dent 2:390–393

Malignant Lesions of the Masticator Space Kushchayeva YS, Kushchayev SV, Glushko TY et  al (2018) Fibrous dysplasia for radiologists: beyond ground glass bone matrix. Insights Imaging 9:1035 Lin TY, Chou YY, Hsiao FC et al (2009) Lung cancer metastatic to the masticator space. Onkologie 32:349–351 Liu Y, You M, Wang H et  al (2010) Ossifying fibromas of the jaw bone: 20 cases. Dentomaxillofac Radiol 39:57–63 Lopes SLPC, de Almeida SM, Costa ALF et  al (2007) Imaging findings of Ewing’s sarcoma in mandible. J Oral Sci 49:167–171 Luo DH, Yang J, Qiu HZ (2014) A new T classification based on masticator space involvement in nasopharyngeal carcinoma: a study of 742 cases with magnetic resonance imaging. BMC Cancer 14:653 Moonis G, Cunnane M, Emerick K et al (2012) Patterns of perineural tumour spread in head and neck cancer. Magn Reson Imaging Clin N Am 20:435–446 Morag Y, Morag-Hezroni M, Jamadar D et  al (2009) Bisphosphonate reated osteonecrosis of the jaw: a pictorial review. Radiographics 29:1971–1986 Murphey MD, Walker EA, Wilson AJ et al (2003) Imaging of primary chrondrosarcoma: radiologic-pathologic correlation. Radiographics 23:1245–1278 Ollivier L, Vanel D, Leclere J (2004) Imaging of chondrosarcomas. Cancer Imaging 4:36–38 Park SW, Kim HJ, Lee JH et al (2009) Malignant fibrous histiocytoma of the head and neck: CT and MR imaging findings. Am J Neuroradiol 30:71–76 Purohit BS, Ailianou A, Dulguerov N et al (2014) FDGPET/CT pitfalls in oncological head and neck imaging. Insights Imaging 5:585–602 Razek AAK (2014) Computed tomography and magnetic resonance imaging of lesions at masticator space. Jpn J Radiol 32:123–137 Razek AAKA, Elsorogy LG, Soliman NY et  al (2011) Dynamic susceptibility contrast perfusion MR imaging in distinguishing malignant from benign head neck tumors: a pilot study. Eur J Radiol 77:73–79 Razek AA, Huang BY (2011) Soft tissue tumors of the head and neck: imaging-based review of the WHO classification. Radgiographics 31:1923–1954 Razek AA, Kandeel A, Soliman N et  al (2007) Role of diffusion-weighted echo-planar MR imaging in differentiation of residual/recurrent head and neck tumours

281 and post-treatment changes. Am J Neuroradiol 28:1146–1152 Razek AAK, King A (2012) MRI and CT of nasopharyngeal carcinoma. Am J Roentgenol 198:11–18 Razek AAK, Nada N (2013) Role of diffusion-weighted MRI in differentiation of masticator space malignancy from infection. Dentomaxillofac Radiol 42:20120183 Sahoo NK, Rangan NM, Kakkar S et al (2013) Masticator space metastasis from a male breast carcinoma: a case report. J Oral Maxillofac Surg Med Pathol 25:160–163 Schuknecht B, Valavanis A (2003) Osteomyelitis of the mandible. Neuroimaging Clin N Am 13:605–618 Singer SR, Mupparapu M, Rinnagio J (2004) Clinical and radiographic features of chronic monostotic fibrous dysplasia of the mandible. J Can Dent Assoc 70:548–552 Som PM, Curtin HD, Silvers MD (1997) A re-evaluation of imaging criteria to assess aggressive masticator space tumours. Head Neck 19:335–341 Thariat J, Julieron M, Brouchet A et  al (2012) Osteosarcomas of the mandible: are they different from other tumor sites? Crit Rev Oncol Hematol 82:280–295 van Rijswijk CS, Geirnaerdt MJ, Hogendoorn PC et  al (2004) Soft tissue tumours: value of static and dynamic gadopentate dimeglumine – enhanced MR imaging in prediction of malignancy. Radiology 233:493–502 Wang P, Yang J, Yu Q et  al (2010) Evaluation of solid lesions affecting masticator space with diffusion-­ weighted MR imaging. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 109:900–907 Wei Y, Xiao J, Zou L (2007) Masticator space: CT and MRI of secondary tumor spread. Am J Roentgenol 189:488–497 Wood RE, Nortje CJ, Hesseling P et  al (1990) Ewing’s tumour of the jaw. Oral Surg Oral Med Oral Pathol 1:120–127 Yu Q, Yang J, Wang P (2008) Malignant tumors and chronic infections in the masticator space: preliminary assessment with in vivo single-voxel 1H-MR spectroscopy. Am J Neuroradiol 29:716–719 Zhang GY, Huang Y, Ci XY et al (2014) Prognostic value of grading masticator space involvement in nasopharyngeal carcinoma according to MR imaging findings. Radiology 273:136–143

Neoplasms of the Sinonasal Cavities Davide Farina, Davide Lombardi, Giovanni Palumbo, and Marco Ravanelli

Contents

Abstract

1

Introduction 

 283

2

Normal Radiological Anatomy 

 284

3

Indications for Imaging Studies 

 286

I maging Appearance and Extension Patterns of Sinonasal Neoplasms   287 4.1  Appearance of the Tumor Mass on CT and MRI   287 4.2  Extension Toward Surrounding Structures   288 4

5 T  umor Types  5.1  E  pithelial Tumors  5.2  N  on-epithelial Tumors  6

Treatment Monitoring 

References 

 293  294  301  311  314

The sinonasal cavities are a complex framework of air cavities, variable in number and size, located in an overall small anatomical area. The high variety of tissues present in this region accounts for the extremely wide range of benign and malignant tumors, only a minority of which display a typical imaging appearance. However, imaging plays a pivotal role in treatment planning, either allowing to select tumors amenable to purely endoscopic resection or in advanced malignant tumors, detecting tumor invasion in crucial anatomic areas (such as the orbit, anterior skull base, and masticator space). Detailed knowledge of the anatomy is the key to exploit all the information provided by imaging, not only at staging but also during follow-up. A pattern of expected anatomic changes is predictable based on the treatment strategy and serves as a guide to detect recurrent tumors.

1  Introduction

D. Farina (*) · D. Lombardi · G. Palumbo · M. Ravanelli Department of Medical and Surgical Specialties, Radiological Sciences and Public Health, University of Brescia – ASST Spedali Civili Brescia, Brescia, Italy e-mail: [email protected]

Paranasal sinus cancer is an overall rare disease accounting for 3–5% of all head and neck tumors and less than 1% of all cancers (Bossi et al. 2016; Siddiqui et  al. 2017) with an incidence of one case every 100,000 and a peak in the seventh to eighth decade. The overall incidence is substantially stable in the Western world (Youlden et al.

Med Radiol Diagn Imaging (2020) https://doi.org/10.1007/174_2020_236, © Springer Nature Switzerland AG Published Online: 19 June 2020

283

284

2013). The maxillary sinus is the most common site of origin (60–70%) followed by nasal cavity (20–30%) and ethmoid (10–15%); frontal and sphenoid sinus are very rarely the primary tumor site (Siddiqui et  al. 2017; Samant and Kruger 2007). Tumor histology is heterogeneous, reflecting the presence of different tissues into the sinonasal tract. There are geographical differences in the epidemiology of paranasal sinus cancer both in terms of overall incidence (Gras Cabrerizo et al. 2007) and in terms of relative incidence of histotypes. Squamous cell carcinoma is the most frequent histotype; however, the proportion of adenocarcinoma in Europe is much higher than in the United States (in men 0.26 vs. 0.06 per 100,000, respectively) (Rampinelli et  al. 2018). According to the International Agency for Research on Cancer (IARC), some occupational risk factors contribute to carcinogenesis. In 1995, wood dust was classified as human carcinogen correlated to the development of sinonasal adenocarcinoma; this has also produced effects on legislation, as the European Union set limits to exposure to wood dusts in workplaces. In addition, a strong evidence of carcinogenic effect was demonstrated for nickel compounds, radium, and leather dust (Charbotel et al. 2014); the evidence for formaldehyde is less robust. Although the major risk factors are occupational, some correlation between tobacco and paranasal cancer (squamous cell carcinoma in particular) was observed (Thompson 2006; Bossi et al. 2016). As in the early stages, the symptoms produced by paranasal tumors are aspecific and similar to those observed in chronic inflammatory conditions (nasal obstruction, nasal discharge, facial pain, occasionally epistaxis), the diagnosis is generally delayed. However, when unilateral or persistent after medical treatment, such symptoms should alert the clinician. More specific signs and symptoms may be seen when tumor breaches the bony framework of the paranasal cavities: on inspection, facial deformity may be observed as the result of anterior maxillary sinus wall destruction. Headache and neurologic deficits, proptosis, epiphora, trismus, loosening of teeth may suggest tumor invasion of the skull base, orbit, masticator space or alveolar process

D. Farina et al.

of maxillary bone, respectively. Perineural tumor spread may induce pain, paresthesia and, when reaching as far as the cavernous sinus, diplopia; not infrequently, however it is asymptomatic.

2  Normal Radiological Anatomy The paranasal sinuses are a complex framework of interconnected air cavities pneumatizing the midface, lined by a mucosal layer that hosts exocrine and salivary glands. Thus, the anatomy may be subdivided in two paired compartments (anterior and posterior) based on the physiologic pathways along which the secretions are drained out of the paranasal cavities. The anterior compartment consists of anterior ethmoid cells, frontal, and maxillary sinus; the posterior compartment is composed of posterior ethmoid cells and sphenoid sinus. The separation between the two compartments is outlined by the basal lamella of the middle turbinate, i.e., the attachment of the turbinate to the lateral nasal wall or medial orbital wall. From the oncologic perspective, however, impairment of drainage pathways by tumor has much less important implications than the relationships of the neoplasm with the outer boundaries of the sinonasal tract, the transgression of which may result in invasion of relevant anatomic structures. Four neighbors—namely the orbit, anterior cranial fossa, pterygopalatine fossa, and infratemporal fossa—deserve special attention and careful assessment on imaging studies. A bone layer separates the orbit from the sinonasal cavities, thin and frail on the medial wall (as the name, lamina papyracea, suggests) and thicker on the roof and floor. Along the inner surface of the orbit, the periosteum of the seven orbital bones fuses in a continuous connectival layer, the periorbita, which offers a barrier to tumor spread. On the anterior part of the lamina papyracea and, caudally, along the anterior part of the medial maxillary sinus wall, courses the nasolacrimal duct, easily seen on all planes because of its straight and vertical course reaching the inferior meatus.

Neoplasms of the Sinonasal Cavities

285

Cranially, the anterior skull base demarcates the sinonasal cavities from the anterior cranial fossa. As it happens in the orbit, the separation between these two anatomic regions is composed of bone layers with different thickness. The paired and symmetric cribriform plates, located on the midline, are the thinnest part, perforated by multiple olfactory fila. The thin vertical lamina of the ethmoid connects the cribriform plate to the thicker ethmoid fovea and this laterally continues in the orbital roof, provided by the frontal bone. Posterior to the cribriform plate the anterior skull base is a flat bone surface, named ethmoid and sphenoid planum (Fig. 1). Along the inner surface of the anterior skull base, the dura mater is firmly attached to the bone.

The pterygopalatine fossa is a slit-like space located in between the posterior maxillary sinus wall and the pterygoid plates, connected with the nasal fossa via the sphenopalatine foramen. Interspersed in the fat tissue filling the fossa are the sphenopalatine artery (terminal branch of the external carotid artery) and a network of neural structures. The maxillary nerve courses horizontally in the upper part of the fossa, crosses the inferior orbital fissure and ends in its terminal branch, i.e., the infraorbital nerve. The infraorbital nerve courses in a canal along the floor of the orbit to reach the infraorbital foramen; the maxillary nerve provides two branches, namely the greater and lesser palatine nerves, which run vertically in the pterygopalatine fossa to reach the upper alveo-

Fig. 1  The anterior part of anterior skull base has an undulated shape, composed of the thin cribriform plate (cp) and thicker ethmoid fovea (ef), connected by the vertical lamina of the ethmoid; the posterior part, sphenoethmoid planum (pl), is a flatter surface. The pterygopalatine fossa (ppf) is the thin space between the posterior maxillary sinus wall and pterygoid plate; the foramen rotundum (fr), through which the maxillary nerve (V2) courses,

opens in the upper tract. The pterygopalatine fossa is in direct continuity with the inferior orbital fissure and communicates indirectly with the superior orbital fissure (sof), through which the oculomotor (III), trochlear (IV), ophthalmic (V1), and abducens (VI) nerve course. All these nerves converge in the cavernous sinus, where they can be discriminated applying submillimetric 3D T2 MRI sequences. nld nasolacrimal duct

286

lar process and teeth. In the cranial part of the pterygopalatine fossa, the palatine nerves reach the pterygopalatine (or sphenopalatine) ganglion where they are joined by sensory and parasympathetic fibers coming, respectively, from the sphenopalatine nerve (branch of the maxillary nerve) and from the vidian nerve. The latter is formed by the greater superficial and deep petrosal nerve and reaches the ganglion after coursing into the vidian canal embedded in the pterygoid root. Just above the inferior orbital fissure, and freely communicating with it, is the superior orbital fissure, located between the ala minor and ala major of the sphenoid: this fissure is crossed by the oculomotor (III), the trochlear (IV), the abducens (VI), and the ophthalmic nerve (V1) as they run from the cavernous sinus to the orbit (Fig. 1). Finally, the pterygopalatine fossa opens laterally in the infratemporal fossa, a fat pad interposed between the posterolateral maxillary sinus wall and the masticator space, bordered cranially by the skull base and caudally in continuity with the buccal space.

3  Indications for Imaging Studies Clinical examination and endoscopy of the nasal cavities should be the first diagnostic step in the work-up of patients complaining of the symptoms mentioned above, even more so if symptoms are unilateral and/or protracted and refractory to first-line medical treatment. However, endoscopy allows to inspect only the mucosal surface of the nasal cavities, middle and inferior turbinate and septum (lower part), thus allowing detection of superficial and/or exophytic lesions and providing indirect or no information on sinus cavities. Imaging is therefore an essential complement in most patients. The selection of the technique mostly depends on clinical suspect. If chronic rhinosinusitis or nasal polyposis are suspected, computed tomography (CT) is the choice. The inherent contrast resulting from the different density of air, bone, and mucosa allows to examine the sinus cavities and drainage pathways properly, thus collecting all the information

D. Farina et al.

needed to confirm the diagnosis and plan the treatment. In recent years, cone beam CT (CBCT) has added to the list of cross-sectional imaging techniques. Compared to CT, CBCT offers higher spatial resolution (slice thickness 0.1–0.2 mm vs. 0.5–0.7 mm), which permits more detailed analysis of subtle bone changes, exposing to lower radiation dose. If clinical and endoscopic findings provide direct or indirect evidence of a mass lesion, ­magnetic resonance (MRI) should be preferred for its higher contrast resolution. Combining conventional 2D, 3D, and diffusion-weighted (DWI) sequences, it is possible to demonstrate the lesion (and in some case define its nature), differentiate the lesion from retained secretions, and stage it, defining its extent to orbit, anterior skull base, pterygopalatine and infratemporal fossa and the cavernous sinus. Nodal metastases are overall infrequent (with the exception of skin tumors of the nasal vestibule and tumors arising in the lower part of the nasal cavity and maxillary sinuses). However, it must be emphasized that cross-sectional imaging (CT, MRI) plays a pivotal role in the detection of retropharyngeal nodes, blinded at ultrasound and physical examination. In a small number of cases, a neoplasm may be an incidental finding in a routine CT scan performed in the suspect of chronic inflammatory disease. As the densities of thickened mucosa and most neoplasms can be indistinguishable on a plain CT, it is essential to highlight some clues that help in the differential diagnosis. Unilateral disease should always alert the reporting radiologist: inflammatory disease is more frequently bilateral although not necessarily symmetrically represented. In a cohort of 250 patients, Eckhoff et  al. (2019) found a significant association between unilateral sinonasal disease and the final diagnosis of benign neoplasm and malignancy. Unilateral opacification of the anterior ethmoid, maxillary, and frontal sinus may be the effect of a neoplasm occupying the middle meatus. Bone changes may also be indicative: remodeling (i.e., bending, demineralization, or sclerosis) suggests chronic inflammatory disease, a slow-growing expansile process (like a mucocele) or a benign

Neoplasms of the Sinonasal Cavities

Fig. 2  CBCT performed for nasal obstruction. On the left side, a soft tissue lesion (asterisk) grows in between the nasal septum and the middle turbinate, laterally displaced (arrowheads). The middle meatus is obstructed by the displaced turbinate, opacification of the maxillary sinus and anterior ethmoid are secondary to the impairment of their drainage pathways. The roof of the ethmoid is focally eroded (arrow). The nasal fossa and paranasal sinuses are normally aerated on the left side. Biopsy proved adenocarcinoma

neoplasm. Conversely, bone destruction suggests a malignancy, although small bone fragments may also be seen embedded within benign neoplasms (Fig. 2).

4  Imaging Appearance and Extension Patterns of Sinonasal Neoplasms 4.1  A  ppearance of the Tumor Mass on CT and MRI Most benign and malignant neoplasms appear on both CT and MRI as solid, moderately enhancing masses; malignancies are also characterized by an aggressive pattern of growth, manifesting as transgression of the bony boundaries of the sinonasal cavities. In the majority of cases, CT density and MRI signal intensities do not allow consistent tissue characterization. However, some general rules

287

may orient the differential diagnosis. Sparse calcifications may be found in a variety of epithelial and non-epithelial tumors, whereas a calcified matrix (easily recognized on CT but sometimes showing a confusing MRI pattern) directs the diagnosis toward a fibro-osseous lesion (Fig. 3). The majority of tumors display an intermediate signal on T2 images; hyperintensity on such sequences rules out quite reliably epithelial tumors and lymphoma and shifts the differential diagnosis toward rarer neurogenic, chondrogenic, or minor salivary gland tumors. Spontaneous T1 hyperintensity is quite specific of mucosal melanoma, but the sensitivity of this sign is limited because not all melanoma subtypes display such pattern. The actual utility of DWI is controversial. A systematic review (Munhoz et al. 2018) analyzed 16 different papers investigating the utility of DWI in sinonasal pathology: the ADC value of malignancies ranged between 0.05 × 10−3 mm2/s and 1.46 × 10−3 mm2/s, whereas the highest ADC value of benign tumors was 1.94 × 10−3 mm2/s. However, many studies found substantial overlap between the ADC of benign or tumor-like lesions and the range of values of malignancies. Moreover, although the variety of benign and malignant histotypes included in some of the 16 studies was high, the systematic review could not encompass the entire spectrum of differential diagnoses that are found in this anatomic area. As a large part of sinonasal lesions can be biopsied quite easily, DWI should not be considered a problem solver but rather a complement of the MRI protocol, allowing to improve the quality of the diagnostic hypothesis. In a limited number of cases, a reliable diagnosis can be obtained based on the pattern of CT or MRI findings (site of origin, growth pattern, signal/density): inverted papilloma is a paradigmatic example (see below). Cross-sectional imaging also plays a role in discriminating tumor from the retained secretions accumulating into sinus cavities blocked by the mass. On contrast-enhanced CT scans, the ­density of the mass is higher than the hypodense fluid trapped into the sinuses. On MRI, the T2 sequence plays a key role, displaying fluid with

D. Farina et al.

288

a

b

Fig. 3 (a, b) CT (a) and TSE T2 (b) show a large well-­ defined mass (asterisk) encroaching the ethmoid planum (arrow). The densely calcified matrix, which suggests a

fibro-osseous lesion, is more obviously demonstrated by CT. Pathology confirmed osteoma

signal intensity markedly higher than the vast majority of tumors. The discrimination may be more complicated when the secretions undergo dehydration. CT may be misleading because the difference in density between inspissated mucus and tumor is much lower. On MRI, the T2 signal intensity of dehydrated secretions is lower than fluid and in some cases very similar to the signal of tumor; the high signal of such secretions on plain T1, however, clears any uncertainty in most cases (Fig. 4).

The interface between the orbit and the ethmoid cells is made of three juxtaposed layers: the epithelium investing the sinus cavities, the thin bone of the lamina papyracea, and the periorbit. CT demonstrates even the finest bone changes but fails at demonstrating the periorbit, thus missing an essential element for the diagnosis of orbital invasion. On MRI, bone and periorbit cannot be distinguished, both being hypointense on all sequences. However, the demonstration of a continuous hypointense linear interface separating the tumor from the intraorbital fat tissue indicates no orbital invasion with very high negative predictive value (Fig.  5) (Maroldi et  al. 1997; Kim et al. 2006). On opposite, a gap in this hypointense layer, as well as an unsharp interface between tumor and fat tissue suggests periorbital invasion and infiltration of the extraconal fat tissue, though with lower positive predictive value (Fig. 6). For long time, these findings were considered the cornerstone for the correct treatment planning: in fact, orbital preservation was traditionally considered feasible unless imaging demonstrated periorbital invasion. In recent years, however, the indications for orbital preservation have been progressively enlarged (McCary et al. 1996; Tiwari et al. 1998; Castelnuovo et al. 2015). The refinement of endoscopic techniques allows clearer intraoperative differentiation

4.2  Extension Toward Surrounding Structures The pretreatment planning of sinonasal malignancies is largely influenced by the involvement of anatomical sites neighboring the sinonasal complex. Two distinct patterns of growth may be described based on the site of origin of the mass, i.e., the nasoethmoidal pattern and the maxillary sinus pattern, which pose different challenges, particularly for surgical treatment.

4.2.1  Nasoethmoidal Pattern Nasoethmoid tumors tend to grow laterally and cranially involving the orbit and the anterior skull base, respectively.

Neoplasms of the Sinonasal Cavities

289

a

b

c

d

Fig. 4 (a, d) The analysis of signal on TSE T2 (a, b) and on T1 (c, d) allows to discriminate the ethmoid tumor mass (dotted line) from the retained secretions in the frontal and sphenoid sinuses (asterisks). Depending on the degree of dehydration, entrapped fluid within the sinus

cavities may display T2 signal ranging from hyper (right frontal sinus) to hypointense (left sphenoid sinus). The inversely increasing T1 signal helps to discriminate dessicated secretions from solid tissue

between invaded and normal tissues; moreover, even relatively large orbital defects can be reconstructed with good functional and cosmetic results. Consequently orbital exenteration is reserved to cases with extensive invasion of extraconal fat, infiltration of extraocular muscles, intraconal orbital fat, optic nerve and bulb, skin overlying the eyelids, and proximal lacrimal pathways (Lisan et al. 2016; Turri-Zanoni et al. 2019). Massive invasion of the orbital apex is still considered a contraindication to surgery and bears a significantly poorer prognosis. The evolution of surgical management of the orbit has shifted the attention on MRI scans from the periorbit to the extraconal fat tissue, extrinsic ocular muscles, and apex. The application of sequences maximizing the signal gap between tumor and fat—such as plain T1 and fat-suppressed T2— enables the demonstration of fat tissue invasion at the orbital apex. Muscular infiltration detection

has pitfalls: the differentiation between compression and infiltration may difficult, particularly when the muscle is deformed by the pressure exerted by the tumor; moreover, the assessment of the inferior oblique muscle may be more difficult than the other extrinsic muscles, due to its thinness and course (Fig. 6). The interface between the nasal fossae and the anterior cranial fossa is composed of mucosa (investing the ethmoid roof), bone of the anterior skull base, dura, cerebrospinal fluid (CSF), and leptomeninges. Along the intracranial surface of the cribriform plate are the olfactory bulbs and tracts, surrounded by CSF. As for the orbit, CT provides detailed representation also of the thinnest bony elements, like the cribriform plate. MRI, however, is superior in depicting the soft tissue elements in the interface. On MRI bone encroachment is shown as interruption of the hypointense linear signal of the

290

D. Farina et al.

a

cortical bone: this is better depicted on T2 sequences in which the hypointensity is contrasted by the hypersignal of CSF (Fig.  7a, b) (Ishida et al. 2002). When bone is interrupted, contrast-enhanced sequences provide information on the relationships between tumor and dura. Dural reaction, manifesting as linear enhancement around the area of contact with tumor, may occur as a reactive, inflammatory change (Fig.  7c, d). Tumor cells, in fact, release proangiogenic factors, which ultimately increase the vascularization of the dura. On the other hand, however, tumor also release metalloproteinases, responsible for extracellular matrix degradation and dural invasion. Some MRI signs correlate with dural infiltration: the interruption of the enhancing line representing the dura, the nodular appearance of the enhancing dura, and dural thickness greater than 2  mm (Eisen et  al. 1996; McIntyre et  al. 2012) (Fig. 7e, f). Edema and nodular enhancement of brain parenchyma and leptomeningeal enhancement suggest brain infiltration. The assessment of intracranial extent of the tumor has implications on prognosis (which is significantly worsened by dural invasion) and on surgical planning. In the last decades, endoscopic

b

Fig. 5 (a, b) Ethmoid chondrosarcoma. The lesion is in contact with the inferomedial orbital wall which, on CT (a), shows a large erosion (white arrows). On MRI, however, TSE T2 sequence (b) demonstrates the persistence of a hypointense linear interface (black arrows) between tumor and orbital fat, representing an intact periorbita

a

b

Fig. 6 (a, c) Orbital invasion. Squamous cell carcinoma (a, b): both T2 (a) and GE T1 (b) with contrast show an undulated, nodular interface between the tumor and the extraconal fat tissue. The hypointense line representing

c

the periorbita is completely effaced on T2, the tumor reaches the medial rectus muscle (arrow). Sarcoma (c): the lesion erodes the orbital floor encasing the inferior oblique muscle (arrow, on the contralateral side)

Neoplasms of the Sinonasal Cavities

291

a

b

c

d

e

f

Fig. 7 (a–f) Anterior skull base invasion. In (a) the tumor contacts the cribriform plane. The thin hypointense signal of bone may be seen (arrow); above it, the olfactory bulb is surrounded by CSF: findings indicate no invasion of the skull base. In (b) the olfactory bulb is completely encased by tumor tissue, which effaces the signal of CSF. (c–d)

Intracranial invasion: the tumor is capped by a thin and continuous dural layer, which enhances after contrast administration (white arrows): findings indicate no dural infiltration. (e, f) The dura is thicker and exhibits irregular, nodular pattern (black arrows) with multiple interruptions, suggesting infiltration

292

resection proved to be feasible in the great majority of nasoethmoid tumors when no extrasinusal extension is seen. When the skull base is encroached or focally involved and/or in presence of dural infiltration, a transnasal craniectomy with anterior skull base reconstruction may be required (Villaret et al. 2010; Schreiber et al. 2018; Mattavelli et al. 2019). Transnasal craniectomy may be performed also in selected cases of scarcely aggressive malignancies with focal brain infiltration (Mattavelli et al. 2019). Macroscopic dural involvement over the orbital roof and/or frontal sinus bony wall invasion generally require the combination of the endoscopic approach with a subfrontal craniectomy (cranio-endoscopic approach) (Villaret et  al. 2010). Massive brain infiltration with edema, bilateral orbit infiltration, optic nerve and/or chiasm involvement, internal carotid artery encasement should still be considered contraindications to any kind of surgical approach.

4.2.2  Maxillary Sinus Pattern Tumor arising from or invading the maxillary sinus may erode its bony walls gaining access to the orbit, pterygopalatine and infratemporal fossa, buccal space, and premaxillary fat pad. In some cases (particularly in adenoid cystic carcinoma and lymphoma) bone invasion occurs in a more subtle pattern, referred to as permeative pattern (Xing et al. 2019). Such pattern consists of tumor growth on the inner and outer surface of maxillary sinus walls, with subtle moth-eaten cortical erosions but no gross destruction. In other areas, such as the central skull base, the medullary space of spongiotic bone may be invaded with only minimal cortical erosions. MRI depicts permeative bone invasion pattern better than CT, particularly when it involves spongiotic bone. The combination of plain and contrast-enhanced T1 sequences reveals any abnormal signal of the medullary bone. The pterygopalatine fossa contains a complex network of nerve fibers (maxillary and vidian nerve, palatine and sphenopalatine nerves) along which perineural tumor spread may occur. Perineural spread is defined as tumor growth within the perineural space of named nerves,

D. Farina et al.

i.e., in the virtual space between neural axons and the surrounding perineurial tissue. Various chemical neurotropic factors, produced by tumor cells and by the local microenvironment, promote and modulate this pattern of growth (Brown 2016). Perineural spread of sinonasal tumor mostly occurs along the maxillary nerve and its branches, although the facial nerve may also be involved, via the interconnections provided by the vidian nerve. Tumor growth tends to occur along the longitudinal axis of the nerve, rather than concentrically around its circumference, and retrogradely—i.e., from the tumor site toward the skull base—rather than anterogradely (Badger and Aygun 2017). Perineural tumor growth enlarges the nerve, thus abnormal nerve thickening and enhancement are key findings on MRI (Fig. 8). Enlarged nerve segments may remodel, widen, and erode bone foramina, fissures, and canals; such findings, better depicted by CT, manifest at a later stage. Retrograde spread may result in intracranial extension into the cavernous sinus, Meckel’s cave or, even further, along the cisternal segment of the trigeminal nerve. Enlargement, increased convexity and enhancement of the cavernous sinus, suggest invasion, although the latter, particularly on MSCT, can be difficult to discriminate from physiologic venous enhancement (Maroldi et al. 1997; Maroldi et al. 2008). Both CT and MRI clearly demonstrate the effacement of the Meckel’s cave, due to the high contrast between the solid tumor tissue and the fluid content of the cave. Perineural spread occurring along a motor branch may also manifest indirectly, with paralysis and atrophy of the corresponding muscles. Overall, MRI is superior to MSCT in detecting perineural spread, demonstrating sensitivity up to 100% (Hanna et  al. 2007; Gandhi et  al. 2011). However, there are pitfalls, impacting on the specificity of the technique. Nerve enhancement may be seen in several non-neoplastic lesions such as inflammation, ischemia, infarct, trauma, demyelination, and axonal degeneration. The enhancement of segmental venous plexuses surrounding nerve structures may mimic

Neoplasms of the Sinonasal Cavities

293

a

b

c

d

Fig. 8 (a–d) Perineural spread. Adenoid cystic carcinoma of the maxillary sinus breaching the posterior wall and invading the pterygopalatine fossa (ppf). Perineural

spread is demonstrated along the maxillary nerve (V2), the vidian nerve (vid), and the palatine nerves (pn)

perineural spread, as it happens, for example, with the maxillary nerve in the foramen rotundum or with the geniculate ganglion. The differential diagnosis of perineural spread may also include inflammatory diseases (granulomatosis with polyangitis, IgG4-related disease) and primary neoplasms (meningiomas or schwannomas growing into skull base foramina). Finally, skip lesions, i.e., focal perineural tumor deposit intercalated between normal nerve segments, may result in understaging of tumor, when not adequately encompassed by the field of view of the scan.

5  Tumor Types The 4th edition of the WHO classification of paranasal tumors (Slootweg et  al. 2017) introduced significant changes, describing new e­ntities (namely NUT carcinoma, biphenotypic sarcoma, and seromucinous hamartoma) and emerging entities, but omitting many histotypes that can occur elsewhere in the body (Thompson and Franchi 2018). The following discussion will maintain the structure of the 4th WHO classification, although focusing on most common entities and on those with peculiar characteristics at imaging.

D. Farina et al.

294

5.1  Epithelial Tumors

a

5.1.1  Benign Epithelial Tumors 5.1.1.1  Sinonasal Papillomas Schneiderian papillomas are epithelial tumors originating from the schneiderian membrane of the mucosa that lines the nasal cavity and paranasal sinuses; three subtypes are described, namely the inverted (IP) and exophitic papilloma (EP)— each accounting for nearly half of cases—and the very rare oncocytic papilloma (OP). CT findings of papillomas are non-specific: most often they appear as polypoid soft tissue masses with lobulated margins (when an air interface allows to appreciate the peripheral borders). Remodeling and sclerosis of adjacent bone structures are frequent occurrences; tiny intralesional calcifications (entrapped bone fragments) may also be seen (Fig. 9). EP almost exclusively arises from the nasal septum, whereas the site of origin of IP may be found more commonly in the lateral nasal wall, ethmoid, or maxillary sinus (Vorasubin et al. 2013). On MRI, papillomas display a characteristic columnar pattern (Fig. 9), either on TSE-T2 or on SE-T1 sequences: this reflects the histologic architecture of the lesion, i.e., juxtaposition of several layers of epithelium (intermediate signal on both sequences) and stroma (hyperintense and vividly enhancing). Such pattern has a very high positive predictive value (95.8%) (Ojiri et  al. 2000; Maroldi et al. 2004). Literature data on the imaging differentiation of the three subtypes is very limited: in a small series including OP and IP (Yang et  al. 2018), cystic changes, high T1-signal, and the low incidence of osteitis were significantly associated to OP. Although benign, papillomas may be associated with atypia and dysplasia; malignant transformation is described for IP and OP in around 5–15% of cases (Vorasubin et al. 2013) (Fig. 10), but not for EP. Loss of columnar pattern and low ADC values (below 1.47 × 10−3 mm2/s) may predict the presence of squamous cell carcinoma. IP has a high propensity to recur (4–22% of cases) (Lombardi et  al. 2011). Several risk factors are described in the literature and a compos-

b

Fig. 9 (a, b) Inverted papilloma. On CT (a), a unilateral polypoid lesion is seen in the right ethmoid, with multiple nodular calcifications. Sagittal TSE T2 (b) shows the typical columnar pattern radiating from the attachment site of the lesion, on the nasal septum

ite staging system has been proposed to predict the likelihood of recurrence based on intraoperative findings. Three combinations of factors increase the risk of recurrence: presence of dysplasia and incomplete resection; dysplasia and frontal sinus involvement; incomplete resection and frontal sinus involvement (Lee et al. 2019). Incomplete resection and recurrence are a matter of debate. According to Lee et al. (2019), incomplete resection may occur due to a presurgical misdiagnosis of nasal polyposis or to an intentional decision of the surgeon (when the lesion is close to vital structures). According to Lund (2000), “the term recurrence merely indicates residual disease in the majority of cases and is directly related to the surgical approach and the ‘care’ with which the IP is

Neoplasms of the Sinonasal Cavities

295

sinonasal epithelium (more commonly nasal cavity and maxillary sinus) with no connection to gnathic sites. The histopathologic appearance resembles its more common jaw counterpart although sinonasal ameloblastoma is predominant in males, occurs later (5th–6th vs. 2nd–3rd decade) (Barrena et  al. 2019) and shows more aggressive behavior. On cross-sectional imaging, ameloblastoma appears as a mixed solid and cystic (uni- or multilocular) lesion, remodeling and destructing adjacent bone structures. Imaging findings are not specific, in particular the MRI appearance of the plexiform variant of ameloblastoma may mimic the columnar pattern of inverted papilloma: this is due to the presence of thin septa and papillary projections of solid tissue within the lesion, both enhancing after contrast administration (Fig. 11). Surgery is the treatment of choice; local recurrence is very frequent, mainly due to bone invasion. In some recurrent tumors, radiation therapy can be an option. Distant metastasis may rarely occur even in the absence of cytological signs of malignancy of the primary lesion.

Fig. 10  The maxillary sinus is occupied and expanded by a lesion which exhibits two distinct patterns. The superomedial part of the lesion, showing a characteristic columnar pattern (black arrow), proved to be inverted papilloma at pathology. The inferolateral part, invading the alveolar process and showing solid pattern (white arrow), was confirmed to be squamous cell carcinoma

removed.” In this regard, it must be emphasized that at the site of insertion of the IP, in some cases, a focal reactive sclerosis of bone is present, manifesting as a spur. Meticulous reporting of this finding, which may be depicted by both CT and MRI, is essential: during surgery, accurate subperiosteal dissection of the mucosa and drilling of the bone at the site of insertion achieves better tumor control (Lombardi et al. 2011). 5.1.1.2  Sinonasal Ameloblastoma Sinonasal ameloblastoma is a benign but locally aggressive neoplasm primarily arising from the

5.1.1.3  Salivary Gland Adenomas WHO classification includes pleomorphic adenoma, myoepithelioma, and oncocytoma. Most arise in the submucosa of the nasal septum although oncocytoma may also be found in the nasolacrimal duct. Myoepithelioma is exceedingly rare. Imaging findings of sinonasal pleomorphic adenoma are not different from the equivalent in major salivary glands. 5.1.1.4  Respiratory Epithelial Adenomatoid Hamartoma (REAH) and Seromucinous Hamartoma (SH) REAH and SH are relatively new disease entities, introduced in the WHO classification in 2005 and 2017, respectively. Both arise in the respiratory epithelium that lines the olfactory cleft and consist of the proliferation of submucosal glands in the epithelial lining. REAH and SH share similar pathologic features and imaging appearance: the most common presentation is a soft tissue mass expanding the

D. Farina et al.

296

a

b

Fig. 11 (a, b) The maxillary sinus is occupied by a lesion displaying hyper T2 signal (a) and heterogeneous contrast enhancement (b), resembling the columnar pattern of inverted papilloma. The involvement of the alveolar

process (arrowheads), however, offers a clue to include an odontogenic lesion in the differential diagnoses. Pathology demonstrated plexiform ameloblastoma

olfactory cleft on both sides, with no bone erosion (Dean et al. 2019). On sagittal reformations, the inferior border of the lesions exhibits crescentic shape. Signs and symptoms are non-specific, overlapping with chronic rhinosinusitis, thus the differential diagnosis with nasal polyposis may be challenging. The site of origin of REAH and SH may also suggest olfactory neuroblastoma and adenocarcinoma: the absence of anterior skull base erosion is a crucial clue, as it directs the differential diagnosis toward a benign entity and allows to plan a purely endoscopic surgical approach. Recurrence after surgery is rare.

to 60% of all malignancies (Thompson 2006). Four variants are described, namely keratinizing (KSCC), non-keratinizing (NKSCC), spindle cell, and lymphoepithelial (Thompson and Franchi 2018). The maxillary sinus is the most common site of origin (around 60% of cases) followed by nasal cavity and ethmoid (10–20% each), but there are geographical differences in distribution; frontal and sphenoid sinus are rarely the primary site. Peak incidence is in the sixth and seventh decades, with male predominance (2:1). Risk factors include occupational exposure (nickel, chlorophenols, textile dust, chromium), smoking, and sinonasal papilloma. Interestingly, human papillomavirus (HPV) was found in some cases, particularly in NKSCC, although the role in the pathogenesis and the clinical significance of this correlation are not as clear as for oropharyngeal SCC (Thompson and Franchi 2018).

5.1.2  Malignant Epithelial Tumors 5.1.2.1  Squamous Cell Carcinoma Squamous cell carcinoma is the most frequent sinonasal malignant histotype, accounting for up

Neoplasms of the Sinonasal Cavities

a

297

b

Fig. 12 (a, b) Squamous cell carcinoma. A solid lesion arising from the mucosa of the posterolateral wall protrudes in the cavity of the right maxillary sinus. The mass erodes the posterolateral wall and infiltrates the fat tissue

of the infratemporal fossa. Reactive inflammatory thickening of the mucosa investing the roof and anteromedial sinus wall (arrows) is detected; a fungus ball (asterisk) is retained in the sinus

Surgery is the mainstay of treatment, followed by post-operative chemoradiation (CRT); induction chemotherapy may offer a benefit to patients with advanced disease, in particular in terms of orbital preservation. Evidences of the literature, however, are controversial (Khoury et al. 2019). At imaging, SCC exhibits no distinctive feature (Fig. 12). The prognosis is overall poor (5-year survival rate is reported to be 50%, recurrence rate is 56%), correlating with the site of origin (maxillary sinus tumors have worse outcome), with the subtype (KSCC worse than NKSCC) and with the stage at the diagnosis.

aggressive nature of the tumor (Thompson and Franchi 2018; Abdelmeguid et al. 2019). On cross-sectional imaging, SNUC lacks distinctive features: usually, it appears as a solid mass arising from the ethmoid or from the upper part of nasal cavities, simultaneously involving multiple sites and with high propensity to invade the orbit and the skull base; calcifications are usually absent (Raghavan and Phillips 2007). The treatment is a combination of surgery, chemotherapy, radiation therapy; three modality schemes provided better results than bimodal. The prognosis is overall poor with median survival 22 months (Chambers et al. 2015). Sinonasal neuroendocrine carcinoma (SNEC) (Fig. 13) is a highly aggressive tumor with neuroendocrine differentiation, subclassified as small and large cell type. SNEC presents at advanced stage, frequently with local and distant metastases. Treatment principles are similar to SNUC (Rischin and Coleman 2008).

5.1.2.2  Sinonasal Undifferentiated Carcinoma and Neuroendocrine Carcinoma Sinonasal undifferentiated carcinoma (SNUC) is a rare and highly aggressive epithelial tumor with uncertain histogenesis (possibly the Schneiderian epithelium) and no squamous or glandular differentiation. Median age of presentation is in the fifth to sixth decade. The high rate of dural invasion (50%), orbital invasion (30%), and nodal metastatization (10–30%) at presentation denote the highly

5.1.2.3  Adenocarcinoma The term adenocarcinoma includes all glandular-­ type malignancies, excluding those of minor salivary gland origin. Two major forms are described: the intestinal-type adenocarcinoma

D. Farina et al.

298

a

b

Fig. 13 (a, b) Sinonasal neuroendocrine carcinoma. A large ethmoid mass protrudes within the frontal sinus (black arrow) and invades the anterior skull base. The dura

is thickened and discontinued (white arrow). Edema (arrowheads) suggests brain infiltration

(ITAC), characterized by histologic features resembling colonic adenocarcinoma, and the non-ITAC (Stelow and Bishop 2017). Both have male predominance, with median age at presentation in the fifth to sixth decade; however, risk factors, clinical history, and prognosis, are quite different. There is strong correlation between ITAC and occupational exposure to softwood dusts and leather dusts. In woodworkers, the risk is estimated to increase by 900 times, time interval for tumor development is about 40 years. The most common sites of origin of ITAC are the ethmoid and nasal cavity, the maxillary sinus is less commonly involved. The 5-year survival ranges between 53 and 83%, mostly influenced by local recurrences (Rampinelli et al. 2018). Non-ITAC is further subdivided in high and low-grade variant, both unrelated to occupational risk factors. Maxillary sinus is more commonly involved, particularly by the high-grade form. The prognosis is significantly worse for high-­ grade tumors (Bignami et al. 2018). Noticeably, on imaging studies ITAC shows a mixed solid-fluid pattern due to the mucus pro-

duced by tumor cells (Fig. 14). This may complicate the discrimination between the lesion and the retained secretions (or mucoceles, due to sinus blockage). Meticulous attention should therefore be paid to the presence of any solid component within a mucus collection, combining the information provided by pre- and postcontrast sequences and by diffusion-weighted sequences. 5.1.2.4  Salivary Gland-Type Carcinomas This group includes a number of histotypes (like adenoid cystic carcinoma, acinic cell carcinoma, mucoepidermoid carcinoma, carcinoma ex-­ pleomorphic adenoma, and oncocytic carcinoma) that may arise from the minor salivary glands scattered in the sinonasal cavities (Lombardi et al. 2006). Among these, adenoid cystic carcinoma (ACC) has peculiar characteristics that reflect on its appearance at imaging. Sinonasal ACC accounts for only 10–25% of all head and neck ACC, but it bears a worse prognosis (5-year survival 62% as opposed to a cumulative 90% for all other sites). The reason for this different natural history is unclear; however,

Neoplasms of the Sinonasal Cavities

299

a

b

c

d

e

f

Fig. 14 (a, f) Adenocarcinoma. There is substantial difference in MRI pattern between intestinal (a–c) and non-­ intestinal (d–f) type. ITAC shows diffuse T2 hyperintensity (a), incomplete enhancement (b) and minimal restriction

on DWI (c) as a result of mucus production by tumor cells. Non-ITAC, conversely, shows solid pattern (d) and homogeneous enhancement (e). Bone destruction (arrows in f) may be found in both

localization in the frontal sinus and poor differentiation proved to be tumor-related negative factors (Mays et al. 2018; Trope et al. 2019). Nodal metastases are uncommon (5%), whereas distant metastasis (mostly to the lung, liver, and brain) occur in more than 50% of patients during their clinical history (Maroldi et al. 2004). Surgery is the mainstay of treatment:

in a recent review of 793 patients, surgery was the just treatment in only 20% of cases, and it was combined with radiation therapy (with or without chemotherapy) in around 50% of cases. However, in recent years, evidence has shown that, in selected unresectable cases, chemoradiation may be used with a curative intent (Mays et al. 2018).

D. Farina et al.

300

Three subtypes of adenoid cystic carcinoma are described, namely tubular, cribriform, and solid, which may have slightly different imaging appearance. Two imaging findings, however, characterize ACC: the submucosal and subperiosteal pattern of growth and the perineural spread (Fig. 15). Submucosal growth into the soft tissues manifests with infiltrative pattern: muscles, vessels, and fat tissue are effaced and/or encased, often with very limited mass effect. Subperiosteal growth results in the presence of tumor tissue on both surfaces of a cortical bone with only focal destruction; similarly, invasion of medullary bone may occur with minimal changes of the cortical bone. These findings are much better demonstrated on MRI: plain SE-T1 and 3D gradient-echo (GE) T1 with fat suppression are key sequences because they enhance the tumor-to-background contrast. Nonetheless, significant discrepancies between the tumor map provided by MRI and the actual tumor spread demonstrated during intraoperative mapping are a frequent occurrence. The permeative growth pattern of ACC increases the probability of microscopic diffusion below the threshold of MRI detection. Perineural spread requires adjustment of the field of view of the examination, in order to track the entire course of the trigeminal nerve (most

a

b

Fig. 15 (a–c) Adenoid cystic carcinoma. The large ethmoid mass displays biphasic MRI pattern: solid in the posterosuperior part, involving the posterior ethmoid; mixed in the anterior part involving anterior ethmoid and

common path of perineural spread for sinonasal ACC), as well as the greater superficial petrosal nerve, coursing along the edge of the petrous bone and connecting the vidian nerve to the facial nerve. The combination of high contrast and spatial resolution (submillimetric voxel size) offered by 3D GE T1 is useful to map this pattern of tumor growth. 5.1.2.5  N  uclear Protein in Testis Midline Carcinoma (NUT) Carcinoma and Teratocarcinosarcoma Recently introduced by the 2017 WHO classification, NUT carcinoma is an aggressive malignancy with no specific imaging appearance. Although the number of cases of NUT carcinoma reported in the literature is limited, orbital, and skull base invasion are quite frequent, emphasizing its aggressive behavior. The diagnosis requires the identification of NUTM1 gene rearrangement. 5.1.2.6  Staging of Sinonasal Carcinomas The staging of sinonasal neoplasms is based on endoscopic and imaging findings. Two different staging systems have been formulated for ethmoid tumors and maxillary sinus tumors (Brierley et al. 2016), reflecting the two main patterns of tumor growth described in Table 1.

c

nasal fossa. Pathology proved coexistence of tubular pattern (posterior part of the lesion) and cribriform pattern (anterior). CISS after contrast demonstrates perineural tumor spread along the vidian nerve (arrows)

Neoplasms of the Sinonasal Cavities

301

Table 1  Staging of sinonasal neoplasms (AJCC) Tx Tis T1 T2

T3

T4 T4a T4b

Maxillary sinus Nasal cavity and ethmoid Primary tumor cannot be assessed Carcinoma in situ Tumor limited to maxillary sinus mucosa with no Tumor restricted to 1 subsite, with or without bone erosion or destruction of bone invasion Tumor invading 2 subsites in a single region or Tumor causing bone erosion or destruction extending to involve an adjacent region within the including extension into the hard palate or middle nasoethmoidal complex, with or without bone nasal meatus, except extension to posterior wall of invasion maxillary sinus and pterygoid plates Tumor extends to invade the medial wall or floor of Tumor invades any of the following: the orbit, maxillary sinus, palate, or cribriform plate  –  Bone of the posterior wall of maxillary sinus   – Subcutaneous tissues  – Floor or medial wall of orbit, pterygoid fossa, ethmoid sinuses Moderately advanced or very advanced local disease Moderately advanced local disease; tumor invades any of the following: anterior orbital contents, skin of nose or cheek, minimal extension to anterior cranial fossa, pterygoid plates, sphenoid, or frontal sinuses Very advanced local disease; tumor invades any of the following: orbital apex, dura, brain, middle cranial fossa, cranial nerves other than maxillary division of trigeminal nerve (V2), nasopharynx, or clivus

5.2  Non-epithelial Tumors 5.2.1  Neuroectodermal Tumors 5.2.1.1  Olfactory Neuroblastoma Olfactory neuroblastoma (ONB), also referred as esthesioneuroblastoma, mostly arises from the nasal vault although ectopic locations in the lower part of the nasal fossa and in the maxillary sinus are reported (Thompson 2009; Holmes et al. 2016; Bell 2018). The most accepted theory suggests that this malignancy originates from the basal neural cells of the olfactory epithelium. The incidence is quite even across ages, but it is very low below the age of 20 and peaks around the sixth decade (Thompson 2009; Abdelmeguid 2018). Rarely, ONB may manifest with paraneoplastic syndromes like hypertension and hyponatremia, due to ectopic hormone secretion. ONB is clinically staged according to Kadish classification (Kadish et  al. 1976) as limited to the nasal cavity (stage A), involving nose and paranasal sinuses (stage B) and extending beyond the nasal cavity and paranasal sinuses (stage C). This classification survives after four decades; in

1993, it was expanded setting a stage D for tumors with nodal or distant localizations at presentation (Morita et  al. 1993; McFadden and Vieira-Brock 2016). On MSCT, ONB may exhibit spontaneous hyperdensity (reflecting high cellularity) and a variable amount of calcifications. MRI pattern is composed of intermediate T2 signal and T1 hypointensity; cystic areas may be found, capping the intracranial part of the tumor. Both iodine and paramagnetic contrast enhancement is strong (Fig. 16), sometimes making the differential diagnosis with vascular tumors difficult. In some cases, when ONB extends intracranially, it shows a characteristic dumb-bell shape with the waist across the skull base. Treatment is multimodal in most cases as frequently ONB is diagnosed in advanced stages. Surgical resection is classically obtained by open craniofacial resection although the indications for endoscopic or cranio-endoscopic resection are rapidly expanding. Surgery is supplemented with post-operative radiation treatment. A combination of chemo- and radiotherapy is applied in unresectable tumors.

D. Farina et al.

302

a

b

Fig. 16 (a, b) Olfactory neuroblastoma. The ethmoid mass shows intermediate TSE T2 signal (a) and homogeneous contrast enhancement (b). The lesion invades the

skull base through a large gap on both sides of the cribriform plate (black arrows). The intracranial part of the lesion displays cystic changes (white arrows)

5.2.1.2  E  wing Sarcoma (ES)/Peripheral Neuroectodermal Tumor (PNET) ES/PNET is a high-grade small round blue cell tumor mainly occurring in young subjects in their first to second decade, with male predominance (1.5:1). Head and neck involvement accounts for 10% of cases, the maxillary sinus is the most common sinonasal site, skull base and jaws may also be affected (Fig. 17). ES/PNET has a quite aggressive growth pattern, orbital and intracranial invasion are common at presentation. Multimodal treatment achieves 50–60% survival rate at 5 years (Thompson 2017).

On imaging studies, melanoma appears as a solid soft tissue mass with local invasion and relatively high tendency to nodal spread. Three peculiar aspects may help to characterize this histotype. Hyperintense signal on plain T1 sequences is quite specific although exhibited mainly by the melanotic variant. This pattern is explained by the paramagnetic properties of melanin as well as by intralesional hemorrhage (Kim et  al. 2000) (Fig. 18). MRI shows intralesional flow voids, representing the rich vascular network of the lesion. This finds indirect confirmation in the frequent episodes of epistaxis in patients affected by melanoma (Kim et al. 2000). The desmoplastic variant of mucosal melanoma may show perineural spread (Raghavan and Phillips 2007). Surgery (open or endoscopic approach) is the mainstay of treatment. Despite multimodal treatment, which includes immunotherapy, the prognosis of mucosal melanoma is very poor (Roth

5.2.1.3  Mucosal Melanoma Sinonasal melanoma arises from melanocytes migrated during embryonal life from the neural crest to the mucosa investing the nasal cavity and sinuses. Nasal fossae (septum and lateral wall) are the most common sites of origin; the maxillary sinus is the most frequent site among sinus cavities. The disease generally affects patients older than 65 years (Ascierto et al. 2017).

Neoplasms of the Sinonasal Cavities

a

b

Fig. 17 (a, b) Ewing sarcoma. On CT (a) a lobulated soft tissue mass is seen in the sphenoethmoid recess (white arrows), an unusual site for inflammatory polyps. MRI (b) is performed 3 weeks after CT: the enhancement pattern is consistent with solid nature; the marked progression heralds the aggressive nature of the lesion. Normal contralateral sphenoethmoid recess (arrowhead)

et al. 2010; Ascierto et al. 2017); because of this poor prognosis, the specific staging system of sinonasal melanoma does not consider the T1 and T2 stages.

5.2.2  Soft Tissue Tumors 5.2.2.1  Benign Soft Tissue Tumors The WHO classification includes leiomyoma, hemangioma, schwannoma, and neurofibroma. Head and neck leiomyomas account for less than 1% of all leiomyomas; among these, sinonasal localization is exceedingly rare; imaging findings are non-specific (Yang et al. 2018).

303

Sinonasal hemangiomas are classified as cavernous and capillary, the latter including lobular capillary hemangioma (also referred to as pyogenic granuloma). The most frequent sites of origin are the anterior nasal septum and the inferior turbinate. Repeated microtrauma and hormonal imbalances are indicated as pathogenetic factors for lobulary capillary hemangioma. On MSCT, hemangioma appears as a soft tissue density mass that may cause bone remodeling (Kim and Kwon 2017; Takaishi et al. 2017). MRI pattern consists of T2 hyperintensity and spontaneous T1 hypointensity. Bright enhancement is obtained after contrast agent application (Maroldi and Nicolai 2004) (Fig. 19). The differential diagnosis should include other highly enhancing lesions, such as juvenile angiofibroma (JAF), glomangiopericytoma, and highly vascularized metastases (i.e., arising from kidney, thyroid, lung, and breast tumors). Endoscopic resection is feasible, also for large-sized lesions, without significant blood loss; consequently, preoperative embolization is generally not necessary. No recurrence was observed in a series of 40 lobulary capillary hemangiomas treated endoscopically and followed up, on average, for 5 years (Puxeddu et al. 2006). Peripheral nerve sheath tumors are overall rare entities, generally benign in nature. Schwannoma more commonly affects the ethmoid followed by the nasal septum and maxillary sinus, usually as a solitary lesion. Neurofibroma may occur in the nasal vestibule and maxillary sinus, generally in patients affected by neurofibromatosis type I. 5.2.2.2  Malignant Soft Tissue Tumors The 4th edition of WHO classification has introduced a new entity in this category, namely the biphenotypic sinonasal sarcoma (BSNS). Formerly termed low-grade sarcoma with dual neural and myogenic features, BSNS tends to occur more commonly in the ethmoid or superior part of nasal fossae. Although imaging characteristics are aspecific, central necrosis and hyperostosis of the adjacent bone have been reported (Oren et al. 2019).

D. Farina et al.

304

a

b

Fig. 18 (a, b) Melanoma. The maxillary sinus is partially occupied by a soft tissue mass (m) which shows some areas of spontaneous hyperintensity on plain T1 (arrows

in a). Hyperintense material is retained within the sinus cavity (asterisk), consistent with blood due to superficial bleeding of the lesion

Rhabdomyosarcoma (Fig.  20) is the most common sinonasal malignancy in children. It is a mesenchymal neoplasm with skeletal muscle differentiation of which four subtypes are described, namely embryonal, alveolar, pleomorphic, and spindle cell. Rhabdomyosarcoma is characterized by aggressive local behavior, thus at the time of diagnosis it is often a large mass invading adjacent structures. On MRI, T2 signal is hyperintense to muscle, contrast enhancement is heterogeneous; necrosis, calcifications, and hemorrhage are rare (Hagiwara et al. 2001). The list of malignant soft tissue tumors also includes synovial sarcoma, more common in younger patients, second to fourth decade, but mainly in the parapharyngeal space and hypopharynx; fibrosarcoma, leiomyosarcoma (Fig.  21), and undifferentiated pleomorphic

sarcoma, which are more typically observed in the adult; angiosarcoma, which has no age predilection and most frequently arises from the skin; malignant peripheral nerve sheath tumor, associated to neurofibromatosis type I in 25–30% of cases. 5.2.2.3  Borderline/Low-Grade Soft Tissue Tumors Solitary fibrous tumor, also referred to as hemangiopericytoma, is an overall rare soft tissue tumor, most frequently occurring in the chest and abdomen; in the head and neck, sinonasal tract, oral cavity, and orbit may be affected (Davanzo et al. 2018). On both MSCT and MRI, solitary fibrous tumor appears as an indistinct soft tissue mass with moderate to bright contrast enhancement; bone changes (remodeling, destruction) are

Neoplasms of the Sinonasal Cavities

305

a

b

c

d

Fig. 19 (a–d) Hemangioma. In the proximity of the superolateral extremity of the frontal sinus, frontal bone is expanded by a TSE T2-hyperintense (a) enhancing (c, d)

soft tissue lesion, with an ADC value of 1.5 × 10−3 mm2/s (b). US (not shown) confirmed the high degree of vascularization of the lesion

generally also found (Fig. 22). Distant metastases are rare (Kim et al. 2005). Wide surgical excision is the treatment of choice; recurrences, mostly due to incomplete resection, range in the literature between 7 and 40% and may occur up to

15 years after initial treatment; such natural history warrants prolonged clinical and imaging follow-up. Sinonasal glomus tumor arises from perivascular glomus-like myoid cells present in the nasal

D. Farina et al.

306

a

b

Fig. 20 (a, b) Rhabdomyosarcoma. In an 8-year-old boy complaining with unilateral nasal obstruction, a lesion is found in the right nasal vestibule, displaying high TSE T2

a

signal (a) and incomplete enhancement (b). Multiple bilateral adenopathies were also present (not shown)

b

Fig. 21 (a, b) A leiomyosarcoma is seen in the left nasal fossa in a patient who, 20 years before, had been treated with RT for undifferentiated carcinoma of the

nasopharynx. Arrows point at residual submucosal scar tissue in the nasopharyngeal vault

cavity and paranasal sinuses. It should not be confused with the more common head and neck paraganglioma although there is substantial overlapping of imaging features. Glomus tumor

appears as a well-defined but locally aggressive soft tissue mass, with intense contrast enhancement and multiple peripheral and intralesional vessels (Razek and Huang 2011).

Neoplasms of the Sinonasal Cavities

a

307

c

b

Fig. 22 (a–c) Solitary fibrous tumor. Coronal T2 (a, b) shows a soft tissue mass with intermediate signal and sparse flow voids (arrows) filling the ethmoid and right nasal cavity. The lesion compresses the inferior turbinate

(black arrowhead) obstructing the distal opening of the nasolacrimal duct. A small dacriocystocele is observed proximally (white arrowhead). After contrast (c), intense enhancement is obtained

The list of borderline/low-grade soft tissue tumors also includes desmoid-type fibromatosis and epithelioid hemangioendothelioma.

common. On MRI, a dense osteoma may even go undetected if the signal void of air into the sinus cavity surrounds the low signal intensity of the lesion. Ossifying fibroma is a locally aggressive neoplasm originating from mesenchymal blast cells, with a peak of incidence in the third to fourth decade and with female predominance; jaw and mandible are involved more frequently than paranasal sinuses. Several pathologic variants exist, accounting for the variety of patterns that may be displayed on cross-sectional scans. A peripheral calcified shell is almost constantly seen on CT, whereas the density of the inner part of the lesion is more variable, multiloculated, and sometimes lytic. On MRI, the calcified rim shows hypointense T2 signal and tends to enhance after contrast administration; the pattern of the inner core is rather unpredictable, ranging from hypo- to hyperintense on T2 with variable enhancement.

5.2.3  Osseous and Cartilaginous Tumors 5.2.3.1  Benign Fibro-osseous Tumors The differential diagnosis of benign fibro-­osseous lesions includes three main entities, namely osteoma, ossifying fibroma, and fibrous dysplasia. Osteoma is the most common benign tumor in the paranasal sinus, often found incidentally in patients in their sixth to seventh decade; it has a slight male predominance. Frontal sinus is the most common site of origin, followed by the ethmoid. CT density and MRI signal reflect the degree of mineralization of the matrix of the lesion: uniformly sclerotic pattern, characterized by homogenous hyperdensity on CT, is the most

D. Farina et al.

308

Fibrous dysplasia is a non-neoplastic slowly progressive bone disorder characterized by expansion of the involved segment, due to replacement of medullary bone with fibrous tissue and immature woven bone. Fibrous dysplasia may involve one or multiple bone segments. In the paranasal sinus, the polyostotic form is more frequent; in 71% of cases it involves the ethmoid, followed by the sphenoid (43%), frontal (33%), and maxillary bone (29%). On CT, the bone segments involved by fibrous dysplasia are expanded, with preserved cortical layer. Density of the medullary bone

ranges from low to ground glass, depending on the balance between fibrous and osseous tissue (Anschuetz et al. 2017). MRI appearance is more protean, with variable signal intensities and contrast enhancement patterns, which sometimes may complicate the differential diagnosis. Given the varied signal pattern of ossifying fibroma and the mutable balance between osseous and fibrous tissue in fibrous dysplasia, the differential diagnosis between the two entities may be in some case extremely difficult on both CT and MRI (Fig. 23).

a

b

c

d

Fig. 23 (a–d) Coronal CT and TSE T2 sequence in two different patients (a, b and c, d) affected by expansile nasoethmoid lesions characterized by mixed pattern, alternating dense and hypointense solid tissue (probably fibrotic) and hypodense, T2 hyper material (fluid). Both

lesions do not display an aggressive pattern of growth. Although very similar, the two lesions obtained different diagnoses, namely fibrous displasia (a, b) and ossifying fibroma (c, d)

Neoplasms of the Sinonasal Cavities

309

5.2.3.2  Osteosarcoma Head and neck osteosarcoma affects the jaw and mandible more frequently than the sinonasal tract; ethmoid origin is anecdotal (Gonzalez et al. 2016; Erdur et al. 2018). Although the pathogenesis is not clarified, specific risk factors have been identified, including preexisting bone abnormalities—among which Paget’s disease, giant cell tumor, osteogenesis imperfecta, osteoblastoma, and others—and previous radiation therapy. Sinonasal osteosarcoma is mostly high-­ grade; however, distant metastases (to other bones, lymph nodes, and lungs) are less frequent than for high-grade tumors arising from long bones. On MSCT, osteosarcoma appears as a soft tissue mass containing coarsely calcified new bone, which exhibits aggressive pattern of growth, resulting in mixed sclerotic and lytic lesions. On MRI, the calcified osteoid matrix shows hypointense T2 signal (Fig.  24); medullary bone invasion is more accurately depicted by MRI (Vlychou et al. 2007). Treatment consists of wide surgical resection, followed by chemotherapy in high-grade lesions; radiation therapy is indicated as completion treatment when clear surgical margins are not achieved. Local recurrence is the most common pattern of failure, emphasizing the role of MRI and MSCT during follow-up.

5.2.4  Hematolymphoid Tumors

5.2.3.3  Chondrosarcoma Chondrosarcoma generally affects long bones, ribs, and pelvis; sinonasal localization is overall uncom-

5.2.4.1  Lymphoma Lymphoma is the third most common tumor of the paranasal sinuses, accounting for 12–15% of

a

b

Fig. 24 (a–c) Osteosarcoma. The mass lesion, centered in the pterygoid root, invades the maxillary and sphenoid sinus and infiltrates the cavernous sinus (arrowhead on the unaffected side). The dura of the middle cranial fossa is

mon. In a systematic review of 161 cases published in the literature, nasal septum was by far the most common site of origin (42.8%); there was a slight female predominance (1.27:1) with an average age of 42 years (Khan et al. 2013). Sinonasal chondrosarcoma is generally a high-­grade lesion, different from laryngeal lesions that are more commonly low grade (Yamamoto et al. 2002). MSCT appearance consists of a soft tissue matrix with scattered calcifications, which may represent a helpful hint for the differential diagnosis. MRI pattern consists of bright T2 signal, T1 hypointensity and non-homogenous contrast enhancement described as “rings and arcs” (Momeni et  al. 2007): peripheral margins and intralesional septa display enhancement, whereas cartilage and mucoid matrix do not (Fig. 25). The different pathologic grades are indistinguishable on CT and MRI. As chondrosarcomas are slow-growing but locally aggressive tumors, invasion, and destruction of adjacent bone structures are common imaging findings. Surgery is the treatment of choice. Local recurrences are frequent (around 50% of cases), whereas distant metastasis is quite rare.

c

thickened but shows no interruption (arrows). The densely calcified matrix, shown by CT (a), is mirrored by the hypointense signal on both TSE T2 (b) and contrast-­ enhanced T1 (c)

310

D. Farina et al.

a

DLCBL manifests as a solid soft tissue mass remodeling and sometimes destructing the adjacent bone structures; invasion of the orbit or of the deep spaces of the suprahyoid neck is common. Density and signal intensity may be indistinguishable from squamous cell carcinoma although in lymphoma necrosis is uncommon. Usually, lymphoma displays intense restriction on ADC maps of DWI sequences, which generally unravels the differential diagnosis (Weber et al. 2003; Chain and Kingdom 2007) (Fig. 26a, b). Bone destruction is more extensive in ENKTL (Fig. 26c, d). When it involves the midline structures, the list of differential diagnoses should be enlarged to aggressive infections (bacterial and fungal), aggressive inflammatory diseases (such as Wegener’s granulomatosis, sarcoidosis, polyarteritis nodosa, and systemic lupus), and cocaine-induced midline destructive lesions.

b

Fig. 25 (a, b) Chondrosarcoma. Coronal CT (a) shows a soft tissue mass containing irregular calcifications. On MRI (b) the sphenoethmoid lesion shows irregular enhancement, manifesting as sparse dot-like and linear areas. Such pattern, coupled with the hyper T2 signal (see Fig. 5, related to the same patient), is highly suggestive of chondrosarcoma

cases; most common presentations are diffuse large B-cell lymphoma (DLBCL) and extranodal NK/T-cell lymphoma (ENKTL). DLBCL is mostly found into the sinonasal cavities and bears more favorable prognosis as compared to ENKTL, which typically involves the nasal cavity, producing destruction of midline bone structures (nasal septum, palate) (Peng et al. 2014). EBV plays a role in the pathogenesis of ENKTL although the mechanism is not fully elucidated.

5.2.4.2  Extramedullary Plasmacytoma Extramedullary plasmacytoma (EP) is a malignant neoplastic proliferation of B-cells, which, different from multiple myeloma and solitary bone plasmacytoma, occurs outside bone and manifests as a soft tissue mass. EP is mostly found in the head and neck, the sinonasal cavities being the most common site of origin. In a review of 175 cases reported in the literature (D’Aguillo et al. 2014), there was male predominance (69%) and overall favorable outcome, with almost 72% of patients showing no evidence of disease or alive with disease over a median follow-up of 39  months. Imaging findings (soft tissue mass, bone destruction) are completely non-specific (Fig. 27).

5.2.5  Metastasis Rarely the nose and paranasal sinus are the site of a distant metastasis, mostly of renal cell, breast, or lung carcinoma. Signs and symptoms are similar to those caused by benign and malignant lesions. Imaging appearance is non-specific although intense enhancement is expected in hypervascular metastases, such as those

Neoplasms of the Sinonasal Cavities

311

a

b

c

d

Fig. 26 (a–d) Lymphoma. (a, b) Diffuse large B-cell lymphoma (DLBCL) manifests as a solid, densely cellular mass with marked restriction of water diffusivity and no necrosis. (c, d) The main sign of extranodal NK/T-cell

lymphoma (ENKTL) is represented by necrosis of midline structures, with a large perforation of the nasal septum (arrows) and a small necrotic area in the palate (arrowhead)

produced by renal cell carcinoma and melanoma. Maxillary sinus (33%) is the most commonly involved site followed by sphenoid sinus (22%), ethmoid (14%), and frontal sinus (9%); simultaneous involvement of more than one cavity occurs in as many as 22% of cases (Prescher and Brors 2001).

6  Treatment Monitoring Two concepts guide the interpretation of imaging studies in the follow-up of paranasal cancer. First, the anatomy is changed by the demolition and reconstruction entailed by the various surgical procedures and by the changes at the

D. Farina et al.

312

a

b

Fig. 27 (a–c) Plasmacytoma. A soft tissue mass is seen in the left nasal fossa (asterisk): the intermediate signal intensity on TSE T2 (a) suggests hypercellularity, confirmed by the low ADC value (0.7  ×  10−3  mm2/s). The

a

c

enhancement pattern (c) is homogeneous. Nasal packing (arrow) was positioned along the lower border of the lesion after biopsy

b

Fig. 28 (a, b) Follow-up scans acquired, respectively, 6  months and 24  months after endoscopic resection of adenocarcinoma of the ethmoid with skull base reconstruction. The comparison between the scans shows

complete re-epithelization (arrows in b) and thinning of the duraplasty (d), along with regression of the inflammatory thickening of the dura (arrows in a)

tissue level produced by radiation therapy. Thus, it is essential to collect all the details on the treatment performed. The demolition observed after surgery may be wider than expected based on the pretreatment scan: endoscopic resection, in fact, requires opening sagittal or medial corridors to reach the anterior and central skull base, respectively (Roxbury et  al. 2017). When the anterior skull base is resected, a separation between the nasal fossae and the neurocranium has to be restored

(Fig. 28). This is generally obtained positioning three layers of heterologous material (two intracranially and one extracranially) (Mattavelli et al. 2017). A rotation flap of nasal mucosa may also be harvested, to protect the nasal side of the reconstruction (Learned et al. 2014). Large surgical gaps need to be filled by flaps. This can be obtained, for example, by rotating the temporalis muscle inferomedially to fill the cavity produced by maxillectomy or orbital resection. Free flaps, harvested from a distant donor

Neoplasms of the Sinonasal Cavities

a

Fig. 29 (a, b) Follow-up scans acquired 6 months after maxillectomy and orbital resection for squamous cell carcinoma. The axial TSE T2 and fat-sat GE T1 with contrast

a

313

b

demonstrate the lateral thigh free flap harvested to fill the surgical gap in the left orbit. Arrows point at the vascular pedicle anastomosed to a branch of the temporal artery

b

Fig. 30 (a, b) Follow-up scans acquired 12 months after maxillectomy and orbital resection reconstructed with a free flap of the rectus abdominis. The combination of TSE

T2 (a) and STIR (b) allows to discriminate the muscular (m) and fatty (f) part of the flap from the submucosal recurrence (r)

site and transferred with their own vascular pedicle, are also frequently used (Fig. 29). Radiation therapy may induce chronic inflammatory changes in the mucosa investing the sinus cavities, in the subcutaneous fat tissue, in cranial nerves and in muscles (masticator space, muscular flaps). Particularly in muscles and nerves such inflammation manifests with enhancement patterns which may mislead the interpretation. The second concept is that the materials and tissues used in the reconstruction undergo changes during time; consequently the pattern of

normal post-operative findings varies during the time course of follow-up, often in predictable ways (Fig. 28). Awareness of the pattern of expected changes on cross-sectional scans in the different time points of follow-up is crucial: any inconsistency in such pattern may indicate a complication or a recurrence. Complications are mostly seen in the immediate post-operative course, or in the short term. Tension pneumocephalus and intraorbital inflammatory collections are among the most serious

314

and are generally investigated with CT (DelGaudio and Ingley, 2010). CSF leak may occur when surgery is extended to the anterior skull base; not infrequently, however, such complication is detected and repaired intraoperatively. Mucocele may develop in the mid- to long term in sinus cavities blocked by inflammatory adhesions. Given its expansile nature, it may be confused for a recurrent tumor; in most cases, however, the pattern of signals (fluid content, possibly with spontaneously high T1 signal) clarifies the diagnosis. Local recurrences are frequently deeply seated, in areas blinded at endoscopy. MRI is superior to CT because the high contrast resolution offered by conventional sequences is boosted by the application of DWI. Although the anatomy may be distorted at the bone/air interface, careful comparison between b1000 images and ADC map provides an added value. Recurrent tumors mostly manifest as nodular lesions with signal pattern matching that of the primary, emphasizing the utility of the pretreatment scan also during follow-up (Fig.  30). When the primary violates the dura, intracranial relapses may manifest as leptomeningeal spread (linear, gyriform enhancement) (Vellin et  al. 2007) or as multiple extraaxial nodules, intracranial nerve enhancement, superficial brain lesions (Maroldi et al. 2005).

References Abdelmeguid AS (2018) Olfactory neuroblastoma. Curr Oncol Rep 20:7. https://doi.org/10.1007/ s11912-018-0661-6 Abdelmeguid AS, Bell D, Hanna EY (2019) Sinonasal undifferentiated carcinoma. Curr Oncol Rep 21:26. https://doi.org/10.1007/s11912-019-0776-4 Anschuetz L, Buchwalder M, Dettmer M et al (2017) A clinical and radiological approach to the management of benign mesenchymal sinonasal tumors. ORL J Otorhinolaryngol Relat Spec 79:131–146. https://doi. org/10.1159/000468945 Ascierto PA, Accorona R, Botti G et  al (2017) Mucosal melanoma of the head and neck. Crit Rev Oncol Hematol 112:136–152. https://doi.org/10.1016/j. critrevonc.2017.01.019 Badger D, Aygun N (2017) Imaging of perineural spread in head and neck cancer. Radiol Clin N Am 55:139– 149. https://doi.org/10.1016/j.rcl.2016.08.006

D. Farina et al. Barrena BG, Phillips BJ, Moran KJ, Betz SJ (2019) Sinonasal ameloblastoma. Head Neck Pathol 13:247– 250. https://doi.org/10.1007/s12105-018-0933-3 Bell D (2018) Sinonasal neuroendocrine neoplasms: current challenges and advances in diagnosis and treatment, with a focus on olfactory neuroblastoma. Head Neck Pathol 12:22–30. https://doi.org/10.1007/ s12105-018-0887-5 Bignami M, Lepera D, Volpi L et al (2018) Sinonasal non-­ intestinal-­type adenocarcinoma: a retrospective review of 22 patients. World Neurosurg 120:e962–e969. https://doi.org/10.1016/j.wneu.2018.08.201 Bossi P, Farina D, Gatta G et  al (2016) Paranasal sinus cancer. Crit Rev Oncol Hematol 98:45–61. https://doi. org/10.1016/j.critrevonc.2015.09.009 Brierley JD, Gospodarowicz MK, Wittekind C (2016) TNM classification of malignant tumours, 8th edition Brown IS (2016) Pathology of perineural spread. J Neurol Surg B Skull Base 77:124–130. https://doi.org/10.105 5/s-0036-1571837 Castelnuovo P, Turri-Zanoni M, Battaglia P et al (2015) Endoscopic endonasal management of orbital pathologies. Neurosurg Clin N Am 26:463–472. https://doi. org/10.1016/j.nec.2015.03.001 Chain JR, Kingdom TT (2007) Non-Hodgkin’s lymphoma of the frontal sinus presenting as osteomyelitis. Am J Otolaryngol 28:42–45. https://doi.org/10.1016/j. amjoto.2006.06.010 Chambers KJ, Lehmann AE, Remenschneider A et  al (2015) Incidence and survival patterns of sinonasal undifferentiated carcinoma in the United States. J Neurol Surg B Skull Base 76:94–100. https://doi.org /10.1055/s-0034-1390016 Charbotel B, Fervers B, Droz JP (2014) Occupational exposures in rare cancers: a critical review of the literature. Crit Rev Oncol Hematol 90:99–134. https:// doi.org/10.1016/j.critrevonc.2013.12.004 D’Aguillo C, Soni RS, Gordhan C et al (2014) Sinonasal extramedullary plasmacytoma: a systematic review of 175 patients. Int Forum Allergy Rhinol 4:156–163. https://doi.org/10.1002/alr.21254 Davanzo B, Emerson RE, Lisy M et  al (2018) Solitary fibrous tumor. Transl Gastroenterol Hepatol 3:94. https://doi.org/10.21037/tgh.2018.11.02 Dean KE, Shatzkes D, Phillips CD (2019) Imaging review of new and emerging sinonasal tumors and tumor-like entities from the fourth edition of the World Health Organization classification of head and neck tumors. AJNR Am J Neuroradiol 40:584–590. https://doi. org/10.3174/ajnr.A5978 DelGaudio JM, Ingley AP (2010) Treatment of pneumocephalus after endoscopic sinus and microscopic skull base surgery. Am J Otolaryngol 31:226–230 Eckhoff A, Cox D, Luk L, et  al (2019) Unilateral versus bilateral sinonasal disease: considerations in differential diagnosis and workup. Laryngoscope. 28108. https://doi.org/10.1002/lary.28108 Eisen MD, Yousem DM, Montone KT et  al (1996) Use of preoperative MR to predict dural, perineural, and

Neoplasms of the Sinonasal Cavities venous sinus invasion of skull base tumors. AJNR Am J Neuroradiol 17:1937–1945 Erdur ZB, Gözen ED, Inan HC et al (2018) Osteosarcoma of the ethmoid sinus. J Craniofac Surg 29:e487–e488. https://doi.org/10.1097/SCS.0000000000004502 Gandhi MR, Panizza B, Kennedy D (2011) Detecting and defining the anatomic extent of large nerve perineural spread of malignancy: comparing “targeted” MRI with the histologic findings following surgery. Head Neck 33:469–475. https://doi.org/10.1002/hed.21470 Gonzalez ME, Raghavan P, Cho B et  al (2016) Primary osteogenic osteosarcoma of the ethmoid sinus in an adolescent: case report. J Radiol Case Rep 10:1–9. https://doi.org/10.3941/jrcr.v10i2.2298 Gras Cabrerizo JR, Sarandeses García A, Montserrat I, Gili JR, Orús Dotú C (2007) Revision of carcinomas in paranasal sinuses. Acta Otorrinolaringol Esp 58:266–275 Hagiwara A, Inoue Y, Nakayama T et  al (2001) The “botryoid sign”: a characteristic feature of rhabdomyosarcomas in the head and neck. Neuroradiology 43:331–335. https://doi.org/10.1007/s002340000464 Hanna E, Vural E, Prokopakis E et al (2007) The sensitivity and specificity of high-resolution imaging in evaluating perineural spread of adenoid cystic carcinoma to the skull base. Arch Otolaryngol Head Neck Surg 133:541–545. https://doi.org/10.1001/ archotol.133.6.541 Holmes M, Su SY, Bell D (2016) Ectopic primary olfactory neuroblastoma of the maxillary sinus. Ann Diagn Pathol 22:45–48. https://doi.org/10.1016/j. anndiagpath.2016.04.001 Ishida H, Mohri M, Amatsu M (2002) Invasion of the skull base by carcinomas: histopathologically evidenced findings with CT and MRI.  Eur Arch Otorhinolaryngol 259:535–539. https://doi. org/10.1007/s00405-002-0491-9 Kadish S, Goodman M, Wang CC (1976) Olfactory neuroblastoma. A clinical analysis of 17 cases. Cancer 37:1571–1576. https://doi. org/10.1002/1097-0142(197603)37:33.0.co;2-l Khan MN, Husain Q, Kanumuri VV et  al (2013) Management of sinonasal chondrosarcoma: a systematic review of 161 patients. Int Forum Allergy Rhinol 3:670–677. https://doi.org/10.1002/alr.21162 Khoury T, Jang D, Carrau R et al (2019) Role of induction chemotherapy in sinonasal malignancies: a systematic review. Int Forum Allergy Rhinol 9:212–219. https:// doi.org/10.1002/alr.22229 Kim HJ, Lee HK, Seo JJ et  al (2005) MR imaging of solitary fibrous tumors in the head and neck. Korean J Radiol 6:136–142. https://doi.org/10.3348/ kjr.2005.6.3.136 Kim HJ, Lee TH, Lee HS et  al (2006) Periorbita: computed tomography and magnetic resonance imaging findings. Am J Rhinol 20:371–374. https://doi. org/10.2500/ajr.2006.20.2889 Kim JS, Kwon SH (2017) Sinonasal hemangioma: diagnosis, treatment, and follow-up of 37 patients at a

315 single center. J Oral Maxillofac Surg 75:1775–1783. https://doi.org/10.1016/j.joms.2016.12.044 Kim SS, Han MH, Kim JE et al (2000) Malignant melanoma of the sinonasal cavity: explanation of magnetic resonance signal intensities with histopathologic characteristics. Am J Otolaryngol 21:366–378. https://doi. org/10.1053/ajot.2000.18865 Learned KO, Adappa ND, Lee JYK et  al (2014) MR imaging evolution of endoscopic cranial defect reconstructions using nasoseptal flaps and their distinction from neoplasm. AJNR Am J Neuroradiol 35:1182–1189 Lee JJ, Roland LT, Licata JJ et  al (2019) Morphologic, intraoperative, and histologic risk factors for sinonasal inverted papilloma recurrence. Laryngoscope 130(3):590–596. https://doi.org/10.1002/lary.28078 Lisan Q, Kolb F, Temam S et al (2016) Management of orbital invasion in sinonasal malignancies. Head Neck 38:1650–1656. https://doi.org/10.1002/hed.24490 Lombardi D, Piccioni M, Farina D et al (2006) Oncocytic carcinoma of the maxillary sinus: a rare neoplasm. Eur Arch Otorhinolaryngol 263:528–521. https://doi. org/10.1007/s00405-005-0004-8 Lombardi D, Tomenzoli D, Buttà L et  al (2011) Limitations and complications of endoscopic surgery for treatment for sinonasal inverted papilloma: a reassessment after 212 cases. Head Neck 33:1154–1161. https://doi.org/10.1002/hed.21589 Lund VJ (2000) Optimum management of inverted papilloma. J Laryngol Otol 114:194–197. https://doi. org/10.1258/0022215001905319 Maroldi R, Farina D, Battaglia G et  al (1997) MR of malignant nasosinusal neoplasms. Frequently asked questions. Eur J Radiol 24:181–190. https://doi. org/10.1016/s0720-048x(97)01183-2 Maroldi R, Farina D, Palvarini L et  al (2004) Magnetic resonance imaging findings of inverted papilloma: differential diagnosis with malignant sinonasal tumors. Am J Rhinol 18:305–310 Maroldi R, Ambrosi C, Farina D (2005) Metastatic disease of the brain: extra-axial metastases (skull, dura, leptomeningeal) and tumour spread. Eur Radiol 15:617–626. Epub 2004 Dec 31 Maroldi R, Farina D, Borghesi A et al (2008) Perineural tumor spread. Neuroimaging Clin N Am 18:413–429, xi. https://doi.org/10.1016/j.nic.2008.01.001 Maroldi R, Nicolai P (2004) Imaging in treatment planning for sinonasal diseases. Springer Mattavelli D, Ferrari M, Bolzoni Villaret A et  al (2019) Transnasal endoscopic surgery in selected nasal-­ ethmoidal cancer with suspected brain invasion: indications, technique, and outcomes. Head Neck 41:1854–1862. https://doi.org/10.1002/hed.25621 Mattavelli D, Schreiber A, Ferrari M et  al (2017) Three-layer reconstruction with iliotibial tract after endoscopic resection of sinonasal tumors. World Neurosurg 101:486–492. https://doi.org/10.1016/j. wneu.2017.02.066 Mays AC, Hanna EY, Ferrarotto R et al (2018) Prognostic factors and survival in adenoid cystic carcinoma of the

316 sinonasal cavity. Head Neck 40:2596–2605. https:// doi.org/10.1002/hed.25335 McCary WS, Levine PA, Cantrell RW (1996) Preservation of the eye in the treatment of sinonasal malignant neoplasms with orbital involvement. A confirmation of the original treatise. Arch Otolaryngol Head Neck Surg 122:657–659. https://doi.org/10.1001/archo tol.1996.01890180063015 McFadden LM, Vieira-Brock PL (2016) The persistent neurotoxic effects of methamphetamine on dopaminergic and serotonergic markers in male and female rats. Toxicol Open Access 2:116. https://doi. org/10.4172/2476-2067.1000116 McIntyre JB, Perez C, Penta M et  al (2012) Patterns of dural involvement in sinonasal tumors: prospective correlation of magnetic resonance imaging and histopathologic findings. Int Forum Allergy Rhinol 2:336– 341. https://doi.org/10.1002/alr.21022 Momeni AK, Roberts CC, Chew FS (2007) Imaging of chronic and exotic sinonasal disease: review. AJR Am J Roentgenol 189:S35–S45. https://doi.org/10.2214/ AJR.07.7031 Morita A, Ebersold MJ, Olsen KD et  al (1993) Esthesioneuroblastoma: prognosis and management. Neurosurgery 32:706–714. ; discussion 714–715. https://doi.org/10.1227/00006123-199305000-00002 Munhoz L, Abdala Júnior R, Abdala R, Arita ES (2018) Diffusion-weighted magnetic resonance imaging of the paranasal sinuses: a systematic review. Oral Surg Oral Med Oral Pathol Oral Radiol 126:521–536. https://doi.org/10.1016/j.oooo.2018.07.004 Ojiri H, Ujita M, Tada S, Fukuda K (2000) Potentially distinctive features of sinonasal inverted papilloma on MR imaging. AJR Am J Roentgenol 175:465–468. https://doi.org/10.2214/ajr.175.2.1750465 Oren N, Vaysberg A, Ginat DT (2019) Updated WHO nomenclature of head and neck lesions and associated imaging findings. Insights Imaging 10:72. https://doi. org/10.1186/s13244-019-0760-4 Peng KA, Kita AE, Suh JD et  al (2014) Sinonasal lymphoma: case series and review of the literature. Int Forum Allergy Rhinol 4:670–674. https://doi. org/10.1002/alr.21337 Prescher A, Brors D (2001) Metastases to the paranasal sinuses: case report and review of the literature. Laryngorhinootologie 80:583–594. https://doi. org/10.1055/s-2001-17835 Puxeddu R, Berlucchi M, Ledda GP et al (2006) Lobular capillary hemangioma of the nasal cavity: a retrospective study on 40 patients. Am J Rhinol 20:480–484. https://doi.org/10.2500/ajr.2006.20.2878 Raghavan P, Phillips CD (2007) Magnetic resonance imaging of sinonasal malignancies. Top Magn Reson Imaging 18:259–267. https://doi.org/10.1097/ RMR.0b013e31815711b7 Rampinelli V, Ferrari M, Nicolai P (2018) Intestinal-type adenocarcinoma of the sinonasal tract: an update. Curr Opin Otolaryngol Head Neck Surg 26:115–121. https://doi.org/10.1097/MOO.0000000000000445

D. Farina et al. Razek AA, Huang BY (2011) Soft tissue tumors of the head and neck: imaging-based review of the WHO classification. Radiographics 31:1923–1954. https:// doi.org/10.1148/rg.317115095 Rischin D, Coleman A (2008) Sinonasal malignancies of neuroendocrine origin. Hematol Oncol Clin North Am 22(1297–1316):xi. https://doi.org/10.1016/j. hoc.2008.08.008 Roth TN, Gengler C, Huber GF, Holzmann D (2010) Outcome of sinonasal melanoma: clinical experience and review of the literature. Head Neck 32:1385– 1392. https://doi.org/10.1002/hed.21340 Roxbury CR, Ishii M, Blitz AM et  al (2017) Expanded endonasal endoscopic approaches to the skull base for the radiologist. Radiol Clin N Am 55:1–16 Samant S, Kruger E (2007) Cancer of the paranasal sinuses. Curr Oncol Rep 9:147–151. https://doi. org/10.1007/s11912-007-0013-4 Schreiber A, Ferrari M, Mattavelli D et  al (2018) Unilateral endoscopic resection with transnasal craniectomy for sinonasal intestinal-type adenocarcinoma: a bi-institutional case-control study on 54 patients. Oral Oncol 87:89–96. https://doi.org/10.1016/j. oraloncology.2018.10.027 Siddiqui F, Smith RV, Yom SS et al (2017) ACR appropriateness criteria® nasal cavity and paranasal sinus cancers. Head Neck 39:407–418. https://doi.org/10.1002/ hed.24639 Slootweg PJ, Chan JKC, Stelow EB et al (2017) Tumours of the nasal cavity, paransal sinuses and skull base. In: WHO classification of head and neck tumours, 4th edn. IARC, Lyon, pp 14–61 Stelow EB, Bishop JA (2017) Update from the 4th edition of the World Health Organization classification of head and neck tumours: tumors of the nasal cavity, paranasal sinuses and skull base. Head Neck Pathol 11:3–15. https://doi.org/10.1007/ s12105-017-0791-4 Takaishi S, Asaka D, Nakayama T et al (2017) Features of sinonasal hemangioma: a retrospective study of 31 cases. Auris Nasus Larynx 44:719–723. https://doi. org/10.1016/j.anl.2017.01.012 Thompson LD (2017) Small round blue cell tumors of the sinonasal tract: a differential diagnosis approach. Mod Pathol 30:S1–S26. https://doi.org/10.1038/ modpathol.2016.119 Thompson LDR (2009) Olfactory neuroblastoma. Head Neck Pathol 3:252–259. https://doi.org/10.1007/ s12105-009-0125-2 Thompson LDR (2006) Sinonasal carcinomas. Curr Diagn Pathol 12:40–53. https://doi.org/10.1016/j. cdip.2005.10.009 Thompson LDR, Franchi A (2018) New tumor entities in the 4th edition of the World Health Organization classification of head and neck tumors: nasal cavity, paranasal sinuses and skull base. Virchows Arch 472:315–330. https://doi.org/10.1007/s00428-017-2116-0 Tiwari R, van der Wal J, van der Waal I, Snow G (1998) Studies of the anatomy and pathology of the orbit in

Neoplasms of the Sinonasal Cavities carcinoma of the maxillary sinus and their impact on preservation of the eye in maxillectomy. Head Neck 20:193–196. https://doi.org/10.1002/(sici)10970347(199805)20:33.0.co;2-w Trope M, Triantafillou V, Kohanski MA et  al (2019) Adenoid cystic carcinoma of the sinonasal tract: a review of the national cancer database. Int Forum Allergy Rhinol 9:427–434. https://doi.org/10.1002/ alr.22255 Turri-Zanoni M, Lambertoni A, Margherini S et al (2019) Multidisciplinary treatment algorithm for the management of sinonasal cancers with orbital invasion: a retrospective study. Head Neck 41:2777–2788. https:// doi.org/10.1002/hed.25759 Vellin JF, Achim V, Sinardet D et  al (2007) Rapidly developing leptomeningeal carcinomatosis following anterior skull base surgery: a case report. Auris Nasus Larynx 34:565–567 Villaret AB, Yakirevitch A, Bizzoni A et  al (2010) Endoscopic transnasal craniectomy in the management of selected sinonasal malignancies. Am J Rhinol Allergy 24:60–65. https://doi.org/10.2500/ajra.2010.24.3397 Vlychou M, Ostlere SJ, Kerr R, Athanasou NA (2007) Low-grade osteosarcoma of the ethmoid sinus. Skelet Radiol 36:459–462. https://doi.org/10.1007/ s00256-006-0231-0

317 Vorasubin N, Vira D, Suh JD et  al (2013) Schneiderian papillomas: comparative review of exophytic, oncocytic, and inverted types. Am J Rhinol Allergy 27:287– 292. https://doi.org/10.2500/ajra.2013.27.3904 Weber AL, Rahemtullah A, Ferry JA (2003) Hodgkin and non-Hodgkin lymphoma of the head and neck: clinical, pathologic, and imaging evaluation. Neuroimaging Clin N Am 13:371–392. https://doi.org/10.1016/ s1052-5149(03)00039-x Xing Z, Huang H, Xiao Z et  al (2019) CT, conventional, and functional MRI features of skull lymphoma: a series of eight cases in a single institution. Skelet Radiol 48:897–905. https://doi.org/10.1007/ s00256-018-3085-3 Yamamoto S, Motoori K, Takano H et  al (2002) Chondrosarcoma of the nasal septum. Skelet Radiol 31:543–546. https://doi.org/10.1007/s00256-002-0535-7 Yang B, Li J, Dong J (2018) MR imaging and CT features of oncocytic papilloma of the sinonasal tract with comparison to inverted papilloma. Br J Radiol 91:20170957. https://doi.org/10.1259/ bjr.20170957 Youlden DR, Cramb SM, Peters S et al (2013) International comparisons of the incidence and mortality of sinonasal cancer. Cancer Epidemiol 37:770–779. https://doi. org/10.1016/j.canep.2013.09.014

Parotid Gland and Other Salivary Glands Tumors Frédérique Dubrulle, Ophélie Guillaud, Mathieu Nobile, and Christophe Jandeaux

Contents

Abstract

1

Introduction 

 320

2

Anatomy 

 320

3

Imaging Issues 

 321

4 P  arotid Benign Tumors  4.1  P  leomorphic Adenoma or Benign Mixed Tumor  4.2  Warthin Tumor (Adenolymphoma)  4.3  Other Benign Tumors  4.4  Congenital Tumors  4.5  Cystic Tumors 

 323  328  330  330  334

5 P  arotid Malignant Tumors  5.1  H  istologic Classification  5.2  I maging Findings 

 336  336  337

6

Strategy in Difficult Cases 

 340

7 7.1  7.2  7.3 

 seudo-Tumors of the Parotid Gland  P Sjögren’s Syndrome  Sarcoidosis  Intraparotid Lymph Nodes 

 344  344  346  346

8 8.1  8.2  8.3 

 umors of the Other Salivary Glands  T Minor Salivary Glands Tumor  Submandibular Gland Tumors  Sublingual Gland Tumor 

 347  347  347  349

9

Conclusion 

 350

References 

 323

 350

F. Dubrulle (*) Collège de Médecine des Hôpitaux de Paris, Paris, France Imagerie, Hôpital Huriez, CHU Lille, Lille, France O. Guillaud · M. Nobile · C. Jandeaux Imagerie, Hôpital Huriez, CHU Lille, Lille, France e-mail: [email protected]

Parotid gland masses are most frequently benign tumors. In order to determine the most appropriate therapy, MR imaging is the best imaging tool to evaluate tumor topography, locoregional extension, relationship with the facial nerve, and even its nature, benign or malignant. Diffusion-weighted sequence (with ADC) and dynamic contrast-enhanced sequences (DCE-MRI) with time-signal intensity curve (TIC) analysis, used together, are helpful for improving diagnostic performance in tumors without morphological signs of malignancy. This chapter reviews the imaging characteristics of parotid gland and other salivary glands tumors.

Abbreviations ADC Apparent diffusion coefficient CT Computed tomography DCE-MRI Dynamic contrast-enhanced MR imaging DTI Diffusion tensor imaging DWI Diffusion-weighted sequence FNA Fine needle aspiration HIV Human immunodeficiency virus MR Magnetic resonance

Med Radiol Diagn Imaging (2020) https://doi.org/10.1007/174_2020_230, © Springer Nature Switzerland AG Published Online: 08 July 2020

319

F. Dubrulle et al.

320

MRI PET ROI SUV TICs Tpeak WR

1

Magnetic resonance imaging Positron emission tomography Region of interest Standardized uptake value Time-signal intensity curves Time to peak enhancement Washout ratio

Introduction

The parotid is the largest salivary gland. It is located in the parotid space. The parotid gland can be affected by a variety of pathologic processes, especially neoplasic. Parotid tumors represent less than 3% of all head and neck tumors and are most frequently benign. These tumors require surgery in most cases and imaging is essential in the workup of these lesions that depends on tumor nature (benign or malignant), localization, size, and extension.

2

Anatomy

The parotid space is a paired lateral suprahyoid neck space surrounded by the superficial layer of the deep cervical fascia. This space extends from the external auditory canal and the mastoid tip superiorly to the angle of the mandible below. It contains the parotid gland, intra- or extra-­ parotid lymph nodes. The gland contains about 20 intraglandular lymph nodes which are considered normal if their transverse diameter is less than 8 mm; this space also contains extracranial branches of the facial nerve and vessels: the external carotid artery and the retromandibular vein just behind the mandibular ramus in the parotid gland (Fig. 1). The facial nerve exits the skull base via the stylomastoid foramen and continues within the parotid gland from its posterior and superior part to its anterior and inferior part, lateral to the retromandibular vein. It then divides into superior temporo-facial branches and inferior cervical branches. By convention, the facial nerve is used as a reference plane within the gland to separate

Fig. 1  Anatomy of the parotid space. The parotid space (surrounded in red) is made of the parotid gland, the retromandibular vein (blue circle), the external carotid artery (red circle), and intra- and extra-parotid lymph nodes. It also contains the facial nerve, which is not directly seen on classical imaging; it is used as a reference plane (green line) to separate the parotid gland into an external superficial lobe and an internal deep lobe

the external superficial lobe and the internal deep lobe but there is no true anatomic division (Fig. 1). The facial nerve plane is not seen routinely with classical imaging and its course can only be estimated (Harnsberger 2011). The parotid space is directly lateral to the parapharyngeal space (prestyloid space), without real anatomic division between these two spaces. The deep portion of the parotid gland bulges in the prestyloid parapharyngeal space in which the deep lobe tumors can extend (Fig. 2). The parapharyngeal space also contains fat tissue, accessory salivary glands, and a prestyloid branch of the mandibular nerve (V3). Anterior to the parotid space is the masticator space which contains the pterygoid muscles. Posterior to the parapharyngeal space is the carotid space (retrostyloid vascular space) (Fig. 2). There is an anatomic division between these different spaces (the masticator space, the carotid space, and the parapharyngeal space). Benign

Parotid Gland and Other Salivary Glands Tumors

Fig. 2  Lateral and internal to the parotid space is the anterior part of the parapharyngeal space (or prestyloid compartment) (both surrounded in red); there is no real anatomic division between these two spaces. The most internal part of the deep parotid lobe bulges in the prestyloid compartment. Posterior (retrostyloid) part of the parapharyngeal space, also called the carotid space (surrounded in green), separated from the anterior parapharyngeal space by the styloid process (indicated in yellow), as well as muscles and a fascial layer originating from it (see also chapter ‘Parapharyngeal Space Neoplasms’). Masticator space (surrounded in blue)

tumors will respect these anatomic limits whereas they will be infiltrated by malignant tumors. The parotid gland also contains salivary ducts. The main parotid duct, or Stensen’s duct emerges from the anterior part of the parotid gland, runs over the masseter muscle and in the superficial cervical fascia, and then abruptly courses medially to pierce the buccinator muscle, forming a nearly 90° angle with this muscle, terminating in the buccal mucosa at the level of the second upper molar (Fig. 3).

3

Imaging Issues

When a patient presents with a palpable mass of the parotid space, the radiologist has to answer several key questions, which are essential to the

321

Fig. 3  The main parotid salivary duct canal or Stensen’s duct (in red) emerges from the anterior part of the parotid gland, runs over the masseter muscle and in the superficial cervical fascia, then forms a 90° angle (arrow), pierces the buccinator muscle to terminate in the oral cavity at level of the second upper molar

head and neck surgeon in order to determine the best therapy (Harnsberger 2011; Shah 2002; Vogl et al. 1999). –– Is the mass intra- or extraparotid? Small, intraparotid masses are easy to identify. For large and deep lobe masses, the knowledge of the different cervical spaces is essential. The pattern of displacement of the parapharyngeal space has to be analyzed. –– Is the parotid space mass single or multiple? Unilateral or bilateral? Multiple lesions are suggestive for specific tumors: for example, bilateral tumors are suggestive of Warthin tumor, multiple small cystic formations suggest Sjogren’s syndrome. –– Does the tumor show benign or malignant characteristics? The surgical approach will depend on these characteristics. If malignancy is suspected, is there evidence of perineural spread along the facial nerve or along the branches of the trigeminal nerve? In that case, the therapeutic attitude will be different.

322

–– Is tumor limited to the superficial lobe of the parotid? A superficial parotidectomy is sufficient for a benign, well-circumscribed, superficial lobe lesion. On the other hand, a superficial lesion extended in the deep lobe requires a total parotidectomy (O’Brien 2003). –– What is relationship of the mass to facial nerve? The facial nerve is not seen on classical imaging but its intraparotid course is known (plane between the stylomastoid foramen and the lateral border of the retromandibular vein). –– Is it possible to suspect the histologic type of a benign tumor? A pleomorphic adenoma of the parotid gland will require surgery. Warthin tumor in elderly patients may be followed up clinically and by imaging. In order to answer all these key questions, MR imaging is the primary modality of choice for parotid gland tumors (Joe and Westesson 1994; Shah 2004). The classical morphologic sequences used are: axial and coronal turbo spin-echo T2-weighted, pre-contrast axial spin-echo T1-weighted with a slice thickness of 2–3  mm; at least one T2 sequence has to be performed without fat suppression to analyze the real T2 signal of the tumor (with fat suppression, all lesions present hyper T2 signal). A sequence with fat saturation after contrast administration is useful to better visualize potential perineural extension along the facial nerve, extension to the deep spaces, or intracranial extension. This sequence is also useful to analyze the tumor enhancement (Tassart et al. 2020). Two functional sequences are useful to be associated with the morphologic sequences: –– Diffusion-weighted sequence (DWI) with calculation of an apparent diffusion coefficient (ADC) is useful to better characterize tumors (Ikeda et  al. 2004; Habermann et  al. 2005; Yerli et al. 2007; Eida et al. 2007). The ROI (region of interest) is placed manually by the radiologist and positioned in a soft tissue region of the lesion, avoiding cystic, hemor-

F. Dubrulle et al.

rhagic, or calcified areas. To improve the reproducibility, a measure of the ADC tumor/ ADC normal parotid gland ratio is recommended by some authors (Tassart et al. 2020; Espinoza and Halimi 2013). –– Dynamic contrast-enhanced MR imaging (DCE-MRI) is a sequence measuring gadolinium-­enhanced signal intensity pixel by pixel over time: a T1-weighted sequence is repeated every 10–15  s during 5  min after bolus injection of gadolinium. The resulting time-signal intensity curves (TICs) are a key element in lesion characterization. This TICs involves careful placing of the ROI over the lesion to avoid cystic, hemorrhagic or calcified areas. In accordance with previous studies by Yabuuchi et  al. (2003), the time to peak enhancement (Tpeak) was determined from the TICs and the washout ratio (WR). The TICs were then divided into four groups (Table 1): type A (gradual enhancement with Tpeak > 120 s without WR); type B (early enhancement with Tpeak    30%); type C (early enhancement with Tpeak 2 mm through the capsule), although only macroscopic ENE qualifies as positive on TNM staging. At present, clinical ENE may be supported by imaging findings, which have been described previously, but must be evident on physical examination to qualify as positive on TNM staging. This is called “clinically overt ENE” and includes evidence of invasion of skin, infiltration of musculature/dense tethering to adjacent structures, or dysfunction of a cranial nerve, the brachial plexus, the sympathetic trunk or the phrenic nerve. The rationale for this restricted definition is that clinical ENE must be unquestionable in order to prevent inadvertent upstaging as this could deny patients the chance of curative treatment (Lydiatt et al. 2017). Nevertheless, clinical assessment is inaccurate and likely to underestimate ENE. Moreover pathological designation of ENE is frequently not possible as many patients are treated by (chemo) radiotherapy rather than surgery and, besides, pathological detection of ENE is imperfect for several reasons (Wreesmann et al. 2016). Therefore, there is scope for imaging to play a greater role in terms of classifying ENE optimally for staging purposes. To this end, there is some promising preliminary evidence for the utility of imaging-based ENE grading systems, using straightforward definitions such as absent (grade 0), infiltration of fat (grade 1) and infiltration of muscle/skin (grade 2) (Ai et al. 2019).

8

I mpact of Nodal Imaging on Patient Management

8.1

Detection of Metastatic Nodes

To understand the impact of imaging for identifying or excluding metastatic nodes, it is worth outlining the treatment strategies for the neck with (N+) and without (N0) metastatic nodal disease.

421

N+ neck: For patients with metastatic nodal disease, definitive surgical treatment may entail a radical or modified radical neck dissection on the involved side(s). Both operations remove lymph nodes in levels I–V and differ only in terms of whether extra-lymphatic structures (spinal accessory nerve, sternocleidomastoid muscle, internal jugular vein) are also removed (Robbins et  al. 2002). In recent years, selective neck dissections (SNDs), which preserve at least one nodal group in levels I–V, have become established as effective oncological procedures that produce less morbidity than radical procedures. SNDs are indicated for early node-positive disease (localised, small-volume nodes that lack ENE) (Andersen et  al. 2002; Schmitz et  al. 2009; Traynor et  al. 1996). Identification of sites of metastatic nodes is also important for planning intensity-modulated radiotherapy (IMRT) (Gregoire et al. 2018; Biau et al. 2019). Metastatic nodes in close proximity to radiosensitive structures including the parotid glands, oesophagus, posterior cranial fossa and spinal cord may pose challenges for radiotherapy field planning, and neoadjuvant chemotherapy may be used to shrink the tumour bulk before commencing concurrent chemoradiotherapy. N0 neck: Depending on the primary site, up to 40% of necks in patients with HNSCC that have no palpable or radiologically detectable nodal metastases (clinically N0) harbour occult nodal metastases (van den Brekel et  al. 1996). Management of these patients is controversial and varies between “watchful waiting” with careful monitoring and prophylactic treatment of the neck by radiotherapy or neck dissection. It is generally accepted that the clinically N0 neck should be treated electively if the risk of occult nodal metastases in the individual exceeds 20% (Weiss et al. 1994; Pfister et al. 2000; Pillsbury 3rd and Clark 1997). Proponents of prophylactic neck surgery argue that selective neck dissection (e.g. excision of nodes in levels I, II, III) can be performed at the same time as surgery for the primary tumour with minimal additional morbidity, and is associated with improved outcomes. In support of this strategy, one early study of clinical N0 oral HNSCC documented a 12% regional (nodal) recurrence rate for patients who received

K. S. S. Bhatia and A. D. King

422

a prophylactic neck dissection compared to 33% for patients who received no prophylactic neck treatment. Furthermore, successful surgical salvage was possible in only 24% of regional failures (Kligerman et  al. 1994). More recently a large prospective randomised trial of T1 or T2 lateralised oral cavity SCCs with clinically N0 necks found a significantly improved disease-­ free survival (69.5% versus 49.5%) and 3-year survival (80% versus 67.5%) for patients undergoing prophylactic selective neck dissection (with further neck treatment as appropriate if found to be node positive) compared to watchful waiting with close imaging surveillance using US (followed by therapeutic treatment for recurrences) (D'Cruz et al. 2015). However, this trial did not account for the depth of tumour invasion in selecting patients due to its historic nature (TNM classifications did not include depth of invasion for T staging until the eighth edition); thus it is possible that some enrolled patients had “high-risk” tumours by current standards, which would not be eligible for watchful waiting. Proponents of watchful waiting contend that it can avoid debilitating morbidity such as long-­ term shoulder pain as well as costs of neck surgery, especially as surgery is ultimately unnecessary for the majority of patients with clinically N0 necks. If watchful waiting is to be used, appropriate selection criteria are important. Early T-stage tumours in the glottis, parotid gland and paranasal sinuses are suitable as these have a low risk of occult nodal metastases (4  mm depth of invasion in oral tongue cancers) and high grade or at high-risk sites including the oropharynx, hypopharynx, supraglottis and subglottis usually undergo prophylactic neck treatment by surgery or radiotherapy (Shear et  al. 1976; Martinez-­ Gimeno et al. 1995; Pentenero et al. 2005; Hegde et al. 2017). Based on the treatment stratifications described above, nodal imaging has the most impact on patient management in the following situations: 1. Distinguishing between patients with an N0 and an N+ neck: In this situation, finding even

one single metastatic node in an otherwise normal neck could change management from a conservative watchful waiting policy to surgery/radiotherapy, or change the surgical approach from a limited to a more extensive neck dissection. Conversely, proving that a single indeterminate node is reactive rather than metastatic could have the opposite effect on management. In this respect, US-FNAC is an invaluable diagnostic technique for lymph nodes with equivocal malignant features on imaging that will alter treatment (Dirix et al. 2006; Bussels et al. 2006). 2. Identification of unexpected nodes, i.e. those outside common pathways including in the contralateral neck; retropharyngeal, parotid, occipital, buccal, facial, anterior jugular and paratracheal nodes; and nodes adjacent to sites which may be preferentially spared during radiotherapy such as the salivary glands: In these situations, the radiotherapy fields or neck dissections may need to be extended, or the treatment modality changed from one to the other.

8.2

Extranodal Extension and Infiltration of Adjacent Structures

Extranodal extension of tumour beyond the perinodal fascia into adjacent tissues can influence treatment options including resectability (Table 2) (Figs. 6, 9 and 10). Invasion of the internal or common carotid artery usually precludes neck dissection as the initial treatment due to high intraoperative risks and the generally poor prognosis of this group of patients who often have locoregionally advanced and disseminated disease (Freeman 2005). Imaging has an imperfect accuracy for carotid invasion although as a general rule tumour directly contacting the vessel less than 180° or greater than 270° of the vessel circumference indicates that vascular invasion is absent or present, respectively (Steinkamp et al. 1999; Yousem et  al. 1995; Yoo et  al. 2000). In addition, tumour fixation to the artery may be detected by palpation during US (Gritzmann et al. 1990; Mann et al. 1994). Frank invasion of

Neck Nodal Disease

the prevertebral muscles by primary or nodal tumour typically signifies that disease is surgically incurable. CT and MRI can reliably exclude prevertebral invasion if there is an identifiable retropharyngeal fat plane between the tumour and the prevertebral muscles. Conversely, loss of retropharyngeal fat plane, with or without minor abnormalities in the adjacent prevertebral muscles, is non-specific as this finding can represent muscle invasion but is more commonly due to peritumoural inflammation. Tumour invading the external carotid artery or its branches does not preclude surgery because these vessels can be sacrificed, reflecting the rich arterial anastomoses in the head and neck. Similarly, the ­sternocleidomastoid muscle, internal jugular vein and skin can all be sacrificed if there is suspected involvement of these structures on imaging.

8.3

Identification of Patients at High Risk for Distant Metastases

Several nodal features at diagnosis are associated with a higher risk of developing distant metastases including size >6  cm, >3 nodes involved, ENE and nodes in the lower neck/supraclavicular fossa (Alvi and Johnson 1997; de Bree et  al. 2000; Houghton et  al. 1998; Leemans et  al. 1993). Under these circumstances, whole-body PET CT may be indicated to search for distant metastases (Brouwer et al. 2005, 2006).

9

Treatment Assessment

9.1

Prediction of Treatment Response to (Chemo) Radiotherapy

TNM staging is the most widely used system for predicting treatment response and outcome. Nodal markers of poor outcome include large size, multiple nodes, involvement of more inferior neck nodes and ENE. ENE based on pathological examination indicates biologically aggressive disease, and is associated with poorer outcomes including survival, as described previ-

423

ously (Myers et  al. 2001; Larsen et  al. 2009). However, at present there are insufficient data on the association of radiologically detected ENE and HNSCC outcome. A few CT studies have documented that appreciable pretreatment nodal necrosis is an independent risk factor for regional failure after chemoradiotherapy (Grabenbauer et al. 1998; Munck et al. 1991; Wang et al. 1996). Nevertheless, in practice it is common for large markedly necrotic nodes, especially if due to HPV-related HNSCC, to demonstrate a complete response to chemoradiotherapy. The limitations of anatomical imaging for predicting response on pretreatment imaging or from early intra-treatment change in size have stimulated research into functional imaging biomarkers. DWI is the most promising, using ADC or the parameter D (diffusion coefficient) from IVIM, whereby a poor response is associated with a high pretreatment ADC/D or a lower early rise in ADC/D 2–3 weeks after the start of treatment. The rise in ADC/D in responding tumours may reflect a combination of necrosis, apoptosis and inflammation (Fig.  12) (Kim et  al. 2009; Vandecaveye et al. 2010; Dai and King 2018). On DCE-MRI, high-vascularity pretreatment as inferred by parameters such as Ktrans and Kep is associated with a better treatment response (Dai and King 2018; Kim et al. 2010; Ng et al. 2016). Amide proton transfer imaging is a novel chemical exchange saturation MRI technique that provides indirect measurements regarding mobile proteins and peptides. Preliminary results in primary tumours in NPC suggest that early intra-­ treatment changes in amide levels may be a predictor of chemoradiotherapy response (Qamar et al. 2019). For PET CT, high-pretreatment FDG uptake in nodal metastases in HNSCC has been found to be a risk factor for distant recurrences (Kubicek et  al. 2010), whereas a low SUVMax (persistent low or decrease of FDG uptake) is predictive of locoregional control (Martens et al. 2019).

9.2

Post-treatment Assessment

Early post-treatment detection of residual or new nodal metastases is important for several reasons

424

K. S. S. Bhatia and A. D. King

Fig. 12  Axial DWI-MRI ADC map (left) and T1 W post-­ contrast MRI in a patient with a left tonsillar SCC and metastatic IIA node pretreatment (upper row) and 2 weeks intra-treatment with concomitant chemoradiotherapy (lower row). The ADC of the node increased from

1.1  ×  10−3  mm2/s pretreatment to 1.5  ×  10−3  mm2/s at 2  weeks intra-treatment. This indicates increased diffusion, which favours a treatment response. There has been no locoregional recurrence as of 3 years post-treatment

including improving the feasibility and outcomes of salvage surgery, especially as the latter is optimally performed no later than several months after chemoradiation, before neck fibrosis has become established. CT and MRI have high negative predictive values for residual disease (94– 97%) (Liauw et al. 2006; Ojiri et al. 2002; Yeung et al. 2008; Lin et al. 2007) if previously involved nodes are no longer visible or appear as tiny flattened foci measuring up to a few millimetres, termed scarred nodes (Fig. 13). In this scenario, further treatment of the neck can be avoided, even for patients who had locoregionally

advanced tumours including N3 disease before treatment (Hermann et al. 2013). The criteria for identifying new metastatic nodes after treatment are the same as used for pretreatment. Identification of viable residual tumour in nodal metastases after (chemo)radiotherapy is challenging because complete resolution of anatomic abnormalities on imaging in successfully treated metastatic nodes can take many months. Moreover, some nodes may never regress completely due to additional changes including fibrosis and hyalinisation, especially if nodes were large and necrotic/cystic before treatment, which

Neck Nodal Disease

425

Fig. 13  Axial T1  W post-contrast MRI showing right nodal metastasis pretreatment (left) and 6  months post-­ chemoradiotherapy (right). An enlarged right level II node (arrow) has reduced significantly in size following treat-

ment although there is a small amount of residual soft tissue remaining. This was believed to represent a post-treatment scar and there has been no regional recurrence as of 3 years post-treatment

is common in HNSCC.  The limited ability of imaging to differentiate post-treatment sterilised from viable nodal metastases is exemplified by suboptimal positive predictive values in the post-­ treatment setting (36–59%) (Liauw et  al. 2006; Ojiri et al. 2002; Inohara et al. 2009). Nodal size is the most widely used imaging criterion for identifying residual metastatic nodes, and uses a short-axis diameter cut-off on axial images ≥10 mm (same as for untreated metastatic nodes), although 15  mm is sometimes used as an alternative threshold. Nodal size thresholds of 10 mm and 15 mm are also used in chemoradiotherapy trials for the Response Evaluation Criteria in Solid Tumours (RECIST 1.1) to designate pathologically enlarged and measurable disease and pathologic non-target lesions. However, size alone has major limitations for identifying patients who require neck dissections after treatment as 17% of nodes 15  mm may be pathologically negative on neck dissection including those >50  mm in

size (Ojiri et al. 2002; Yao et al. 2004). Volumetric changes in nodes are also useful as some reports suggest that reductions over 75–90% may have predictive regional control (Labadie et al. 2000; Sanguineti et al. 1999). Necrotic or cystic areas, which can persist in nodes after treatment, can hamper the use of absolute or relative size ­criteria, although this can be circumvented by measuring only the solid enhancing components rather than the whole node (King et al. 2016). Other anatomical criteria are not validated although MRI features that are suspicious for residual disease are an intermediate T2 signal intensity in the solid component of a residual node in comparison to low T2 signal intensity in areas of scar tissue, and a thick irregular margin of a cystic node in comparison to a thin smooth rim (King et al. 2013). On DWI, nodes showing persistent restricted diffusion (high signal on high-b-value DWI images and low signal on ADC) are suspicious for residual viable metastatic disease (Mundada et al. 2018). Indeed, an incomplete nodal response may be detected on

426

K. S. S. Bhatia and A. D. King

DWI as early as 3–6  weeks post-treatment, (Vandecaveye and Hermans 2011), which is much earlier than anatomic MRI sequences (van der Hoorn et al. 2017). When evaluating DWI, it remains important to exclude areas of necrosis as these can undergo fibrotic organisation, which produces a fall in ADC mimicking residual cancer. Recurrences occurring at sites of previously treated nodes typically develop from small nests of viable tumour within apparent scar tissue rather than appearing as discrete enlarging nodes. Accordingly any nodal scar tissue should be examined carefully to identify suspicious tumour

foci, which are suggested by an intermediate T2 signal and restricted diffusion. FDG-PET CT is widely used for post-­ treatment assessment of HNSCC (Fig. 14), using mainly subjective assessments of FDG activity. When FDG-PET CT is performed at least 10–12  weeks after treatment, it is reported to have superior accuracy than CT alone or MRI (Wong 2008; Isles et  al. 2008; de Bree et  al. 2009). Results of meta-analyses show a similar trend to the other imaging modalities, with a high negative predictive value (94.5–96%) but low positive predictive value (49–52.1%) (Isles et al. 2008; Gupta et  al. 2011). Consequently, a

Fig. 14  Axial FDG-PET/CT at the oropharyngeal level in a patient with a right tonsillar SCC (short arrow) and a 1.8 cm metastatic right level II node (long arrow) before treatment (top row) and after completion of chemoradiation (bottom row). Pretreatment, both tumour sites show avid FDG uptake; tonsil SUVmax  =  6.1, level II node

SUVmax = 4.3. Post-treatment, neither site shows appreciable increased FDG uptake. The patient has had no recurrence as of 2 years post-treatment. Mild physiological FDG uptake is present in lymphoid tissue in the tongue base (arrowheads)

Neck Nodal Disease

427

complete anatomic and metabolic resolution of indeterminate residual abnormalities or at greater nodal metastases on PET CT reliably predicts risk of locoregional relapse should undergo reguregional control, but a positive PET CT is non- lar follow-up imaging. Most regional relapses/ specific and requires further investigation. recurrences occur within the first 3 years of treatReports from FDG-PET MRI show that com- ment, with the majority being diagnosed in the bining these two modalities improves diagnostic first 2 years. Imaging follow-up therefore is usuperformance, especially with respect to increas- ally performed every 4–6  months for 2–3 years ing specificity, resulting in a high accuracy for (Lell et al. 2000; Som et al. 1993) although it may detecting residual disease when PET, anatomical be more frequent in the earlier post-treatment MRI and DWI are all positive (Becker et al. 2018; period. FDG-PET CT can be used for the first Schouten et al. 2015). baseline assessment if available, although a MRI Ultrasound has limited utility for post-­ or CT is also recommended so that subsequent treatment assessment because, although the dis- surveillance studies, which are usually MRI or appearance of nodes indicates response CT, can be compared with the same modality. US (Furukawa and Furukawa 2010; Yusa et al. 2000), has selective roles post-treatment, including surabnormal vascularity, altered echogenicity, veillance of patients with a clinical N0 neck who absent hilum and border irregularity can be found are being managed by watchful waiting (van den in both metastatic and benign residual masses Brekel et al. 1999; Flach et al. 2013). A watchful (Yusa et  al. 2000). Furthermore, ultrasound-­ waiting strategy of regular US examinations guided biopsy of post-treatment nodes is prone to including every 8–10 weeks in the first year has sampling errors because residual nests of cancer been shown to detect nodal recurrences early and may be small and dispersed within scar tissue. optimise the chances of successful salvage (van den Brekel et al. 1999; Flach et al. 2013; de Bree et al. 2015).

9.3

Post-treatment Surveillance

After employing anatomical imaging, functional imaging and imaging-guided biopsy, it still may not be possible to distinguish between a post-­ treatment nodal mass that contains residual cancer and one that is composed only of benign post-treatment inflammation and scarring. Furthermore, some relapses arise from occult nodal metastases, and these may not manifest until much later in the post-treatment period. Consequently, there is still no substitute for close clinical and imaging surveillance. The timing of the first post-treatment baseline MRI or CT examination following chemoradiotherapy or surgery is controversial although it is usually performed early at around 6–8  weeks after completion of treatment or later at 10–12 weeks (Liauw et al. 2006; Lell et al. 2000; Som et al. 1993). An early assessment is preferable in those patients who had advanced disease and are at higher risk of relapse, whereas in those with a very low risk the baseline imaging may even be postponed to 4–6 months. Patients with

10

 rief Overview of Non-­ B HNSCC Lymphadenopathies

10.1

Lymphoma

Cervical nodes can be involved by lymphomas, which are broadly categorised into Hodgkin’s (HL) and non-Hodgkin’s lymphomas (NHL), but in fact comprise various subtypes and grades. Overall, cervical nodal involvement in NHL is more common, being present in around a half of head and neck NHLs, and the incidence is higher in diffuse large-cell lymphoma than natural killer/T-cell lymphomas (Chisin and Weber 1994; King et al. 2004b; Hanna et al. 1997). Neck nodes in NHL frequently coexist with extranodal sites of involvement and can involve nodes non-­ contiguously (Weber et al. 2003). HL classically involves single or multiple contiguous nodes, especially along the internal jugular chain, and frequently extends into the mediastinum (Weber et al. 2003).

K. S. S. Bhatia and A. D. King

428

The most common imaging appearance in nodal lymphoma is of single or multiple adjacent enlarged nodes, which can measure up to 10 cm in diameter, and which are homogeneous and round. On US nodes usually appear solid, markedly hypoechoic, and may contain internal reticulations although they can appear anechoic on older US machines, which has been described as “pseudocystic” (Ahuja et  al. 2008). Lymphomatous nodes commonly have preserved hilum and an enhanced hilar vascular pattern although the hilum can be displaced eccentrically (Fig.  15). These classical imaging appearances can overlap with chronic reactive nodes and other inflammatory or infectious lymphadenopathies. Although less common, nodal lymphomas can display more overtly suspicious imaging features, especially high-grade/aggressive subtypes, including absent hilum, necrosis, perinodal inflammation, matting and frank extranodal extension (King et  al. 2004b; Saito et  al. 2001; Wang et  al. 1999). Nodal calcification prior to therapy is rare and described mainly in mediastinal rather than cervical nodes (Apter et al. 2002). Similar to nodal metastases, size criteria for the diagnosis of lymphomatous nodes are arbitrary

although cut-offs between 1 cm and 1.5 cm are widely used for staging, and multiple smaller nodes are also considered suspicious (Cheson et  al. 1999). DWI has shown some promise for lymphoma detection as involved nodes have much lower ADCs than SCC and normal nodes (Abdel Razek et al. 2006; Sumi et al. 2003; King et al. 2007; Kwee et al. 2011). Both HL and NHL are FDG avid on FDG-PET CT, and aggressive lymphomas tend to have higher FDG uptake than indolent lymphomas (Hicks et  al. 2005). FDG-­ PET CT can identify disease in normal-sized lymph nodes, which may result in disease upstaging and alter radiotherapy planning. FDG-PET CT has superior accuracy for lymphoma compared to morphological imaging techniques, and is now the imaging gold standard for staging and post-therapeutic monitoring (Collins 2006). With respect to diagnosis, cytology is unable to diagnose or exclude lymphoma and although inclusion of flow cytometry can improve the accuracy, false negatives can still occur; besides flow cytometry is not widely available outside specialist centres (Herd et  al. 2012). Nowadays, US-CNB is used first line in suspected cases of lymphoma, obviating the need for excision biopsy in most cases, as US-CNB allows nodal architecture to be evaluated, which is required for accurate diagnosis including subtyping (Burke et al. 2011).

10.2

Fig. 15 Axial T1  W post-contrast MRI showing an appearance of nodal lymphoma. There is gross lymphadenopathy with multiple enlarged rounded nodes, most of which appear solid although a right level II node demonstrates necrosis (arrow). Hila can be preserved even in grossly enlarged lymphomatous nodes (arrowheads)

Thyroid Cancer

The frequency of nodal metastases in thyroid cancer depends on the histology. Papillary carcinoma, which accounts for over 90% of thyroid cancers, has the highest rate of nodal metastases (30–90%), followed by medullary carcinoma (50%), anaplastic carcinoma (40%) and follicular carcinoma (10%). Medullary carcinoma has a high propensity to spread bilaterally and to the mediastinum (Machens et  al. 2002). Table  1 shows routes of nodal spread and Table 2 shows nodal staging. Nodal disease is usually treated by surgical resection although small nodes may be treated by post-operative radioactive iodine ablation. Ultrasound-guided radiofrequency ablation

Neck Nodal Disease

Fig. 16  Greyscale US showing nodal metastases from thyroid papillary carcinoma. The nodes are hyperechoic, with one showing an anechoic cystic component (long arrows) and the other contains multiple tiny echogenic foci compatible with punctate microcalcifications (short arrows)

is an alternative treatment for recurrent nodal metastases (Na et al. 2012). Ultrasound is the method of choice for initial imaging of primary thyroid tumours and therefore is the most widely used method for assessing nodal metastases in the neck. In addition to the imaging features of metastatic nodes described for HNSCC, nodal metastases from papillary carcinoma frequently display increased echogenicity compared to muscle, micro and/or macrocalcifications and areas of cystic change (Fig.  16) (Ahuja et al. 1995; Rosario et al. 2005). CT also has a good diagnostic performance (Cho et  al. 2019), and can detect nodes not easily accessible by ultrasound such as those in the superior mediastinum, which are at increased risk in patients with low cervical nodal involvement or medullary carcinoma. CT also detects calcification and cystic change, and detection of small metastatic nodes is aided by early arterial enhancement on dynamic contrast-enhanced CT (Park et al. 2017). However, iodinated contrast media reduces the effectiveness of radioactive iodine ablation for a few months; thus contrast-enhanced CT should be avoided in patients with known or suspected differentiated thyroid carcinomas if subsequent radioactive iodine treatment is a possibility. MRI is not as sensitive for detecting calcifications whereas cystic nodes may be visualised, which occasionally may display a characteristic high T1-weighted signal intensity (Som et  al. 1994;

429

King 2008). The presence of unexplained nodes with calcifications on CT or high T1-weighted signal cystic regions on MRI should always prompt a search for a primary thyroid papillary carcinoma. Metastatic nodes located posteriorly in the central compartment, including the lower paratracheal and paraoesophageal groups, are more difficult to detect on US than in the lateral or anterior neck. Nonetheless, special effort should be made to assess these nodal groups as these are frequently responsible for so-called thyroid bed recurrences. Further challenges include the fact that nodal metastases are frequently smaller than 5 mm and US-FNAC of these may be technically difficult. Fortunately these small nodes usually respond to post-operative radioactive iodine ablation. Iodine-123 or iodine-131 scintigraphy is performed after surgery to identify and ablate residual nodal metastases or tumour deposits in the neck and elsewhere, but is ineffective for nodes that do not take up iodine. Indeed, tumour recurrences that were initially well differentiated and iodine avid may become dedifferentiated and non-iodine avid. Nevertheless, other techniques including FDG-PET CT may identify these nodes as a result of alterations in their metabolism (Marcus et al. 2014).

10.3

Salivary Gland Carcinoma

The majority of nodal metastases from salivary malignancies arise from primaries in the ­submandibular gland (42%), followed by parotid gland (25%) and oral cavity (9%) (Terhaard et al. 2004; Lloyd et al. 2010). Salivary malignancies are histologically diverse, and the frequency of nodal metastases is correlated with size, histology and grade. For example, one study reported a frequency of lymph node metastases of 59% and 6% in high-grade and low- or intermediate-grade parotid carcinomas, respectively, which equated to a frequency of 29% overall (Nishikado et  al. 2018). Routes of nodal spread from carcinomas of the major salivary glands are shown in Table 1. Nodal status is an independent prognostic factor for recurrence and survival (Klussmann et  al.

K. S. S. Bhatia and A. D. King

430

2008; Rodriguez-Cuevas et  al. 1995). Conventional imaging criteria for nodal metastases in salivary cancer are similar to those used for HNSCC.  FDG-PET CT may also be used for staging aggressive malignancies but has limited value in most cancers that are intermediate and low grade as a result of their lower metabolism (Park et  al. 2013). Nodal staging for salivary malignancies follows that of HNSCC although some evidence suggests that alternative criteria, including nodal number, may be more accurate for prognostication (Aro et al. 2018). In general, salivary cancers are relatively radioresistant; hence clinically evident nodal metastases are treated by therapeutic neck dissection (Stennert et  al. 2003). The reported incidence of occult nodal metastasis in salivary cancer ranges between 12% and 48% (Rodriguez-Cuevas et al. 1995; Stennert et al. 2003; Armstrong et al. 1992; Regis De Brito Santos et al. 2001), and the optimal management of clinical N0 disease is controversial. In general, most patients with a preoperative diagnosis of a salivary malignancy undergo a prophylactic selective neck dissection during primary tumour surgery as this incurs minimal additional morbidity and may reduce regional failures (Stennert et al. 2003; Regis De Brito Santos et  al. 2001; Frankenthaler et  al. 1993; Medina 1998; Korkmaz et  al. 2002). Intraoperative frozen section of lymph nodes may also be performed to guide surgery including whether or not to proceed to a comprehensive neck dissection (Vander Poorten et al. 2010).

10.4

Nasopharyngeal Carcinoma

Nasopharyngeal carcinoma commonly spreads to regional nodes, and cervical lymphadenopathy is a frequent presentation. Imaging of the primary tumour and nodes is optimally performed with MRI, and the criteria for nodal metastases are similar to those used for HNSCC although the staging system differs, as shown in Table  2. Lateral retropharyngeal nodes and upper internal jugular chain nodes including those posterior to the vein (especially level IIb) are frequently involved initially (Fig.  3a). Routes of nodal spread down the neck are shown in Table 1. In the

latest TNM staging system, N3 has been redefined for clarity to include nodes in the neck that are below the caudal margin of the cricoid cartilage, which previously were defined as in the supraclavicular fossa. Advanced neck nodal disease is one of the main risk factors for distant metastases and is an indication for a whole-body FDG-PET CT. ENE is not included in the staging system for NPC at present; this may need to be reviewed as a recent study found that ENE was an independent predictor of poorer outcomes (Ai et  al. 2019). In addition, unidimensional nodal measurements in NPC staging currently refer to largest diameter of a single node or conglomerate nodal mass, although measuring contiguous involved nodes has been found to have greater prognostic significance (Ai et al. 2018). Moreover, nodal volumes in NPC appear to be superior to unidimensional nodal measurements for predicting distant metastases (Ai et al. 2017). With respect to post-treatment surveillance in NPC, nodes that are usually spared from the radiation field in NPC, which includes submental, paratracheal and suprasternal nodes, frequently become enlarged chronically due to rerouting of lymphatic drainage to these sites. Thus the appearance of new enlarged nodes in these locations should be evaluated and not assumed automatically to be metastatic.

10.5

Skin Cancer

Skin cancer in the head and neck arises frequently in sun-exposed sites including the face, scalp and ear, and superficial nodal groups are more commonly involved when compared with other HNSCCs (Table  1). The appearances of metastatic nodes from cutaneous squamous cell carcinoma are similar to those of other HNSCC whereas nodal metastases from malignant melanoma tend to show less necrosis and less ENE (van den Brekel et  al. 1998b), and may have a high signal intensity on T1-weighted MR imaging due to the paramagnetic effects of melanin. For cutaneous SCC, approximately 5% of patients have nodal metastases at presentation, with the rate being higher in immunosuppressed individuals, high T-stage tumours including

Neck Nodal Disease

431

>2  cm diameter or ≥6  mm in depth (Brantsch et al. 2008), or recurrences. For cutaneous malignant melanoma, SNB has an established role in nodal staging, and the risk of nodal metastases is correlated with primary lesion thickness, equating to less than 5% if under 1 mm thickness than compared with around 35% if over 4 mm thickness (McMasters and Swetter 2003).

chapter although tuberculous lymphadenitis merits special attention as it is one of the commonest causes of persistent cervical lymphadenopathy in the developing world and parts of the developed world. Tuberculous nodes may be discrete, matted or confluent masses, which usually develop insidiously and more frequently arise in the posterior triangle than HNSCC nodes. They may be initially solid, homogeneous and smoothly enlarged, resembling inflammatory adenopathies 11 Squamous Cell Carcinoma and lymphoma. However, focal abnormalities are often seen even in small nodes due to coagulative of Unknown Primary or liquefactive necrosis, which may progress to In a small percentage of patients presenting with frank abscesses with eventual nodal capsular rupneck nodal metastases from an SCC as deter- ture (Ahuja and Ying 2003; King et  al. 1999; mined by cytology or histology, a primary tumour Moon et al. 1997). Perinodal inflammation (peri-­ cannot be identified after detailed search on adenitis) is common, and nodes show increased cross-sectional imaging and FDG-PET CT, clini- FDG uptake. Unsurprisingly, tuberculous lymphcal assessment and surgical biopsies of all high-­ adenitis can overlap completely with nodal risk mucosal sites including the nasopharynx, metastases on imaging. More advanced cases can faucial tonsils, tongue base and hypopharynx present as complex necrotic masses with no evi(Guntinas-Lichius et al. 2006). This is termed a dence of the original nodal architecture and intercarcinoma of unknown primary (CUP) and the connected abscesses, and may have sinuses staging of such nodes has been revised in the lat- tracking to the skin, which are characteristic. est TNM classification to take into account of There is often a paucity of acute inflammatory their most likely sites of origin based on their features, which is also suggestive of this condiEBV and HPV status on pathological analysis. tion and has led to the description “cold abscess”. Accordingly, nodes that are HPV/p16 positive are Several non-infective cervical lymphadenopastaged as for HPV-positive oropharyngeal SCC; thies may overlap with malignancy on nodes that are EBV positive are staged as for imaging including Castleman disease (Fig.  17), nasopharyngeal cancer; and nodes that are HPV/ p16 and EBV negative are staged as for HPV-­ negative oropharyngeal SCC (Lydiatt et al. 2017).

12

Non-malignant Lymphadenopathy

Conventional imaging criteria using size, morphology as well as FDG uptake to predict metastatic lymphadenopathy are based on studies that predominantly compare metastatic nodes with normal and reactive lymph nodes. However, these same imaging criteria are much less accurate for differentiating metastatic nodes from other nodal pathologies. As a consequence, tissue diagnosis is almost invariably required for diagnosis. A detailed account of benign cervical lymphadenopathies is beyond the scope of this

Fig. 17 Axial CT post-contrast showing an enlarged homogeneous left level II lymph node. This was diagnosed as Castleman disease

K. S. S. Bhatia and A. D. King

432 Table 3  Summary of main points in the evaluation of cervical nodal metastases Diagnosis

Staging

Impact on management

Post-­treatment assessment

HNSCC: enlarged (size based on minimum axial diameter), round shape, absent hilum, peripheral vascularity, heterogeneous parenchyma (necrosis), ENE, restricted diffusion (low ADC) and avid FDG uptake (lymphoma and some non-malignant lymphadenopathies can have a similar appearance) Papillary thyroid carcinoma: in addition to the above, one may find calcification, cysts of high T1 signal intensity and increased echogenicity Melanoma: high T1 signal intensity may be seen HNSCC: single ipsilateral node 6 cm or with ENE indicates N3; any other combination indicates N2 disease (size based on maximum diameter) p16-positive oropharyngeal, nasopharyngeal and thyroid carcinomas have separate staging systems Distinguish between patients with an N0 and an N+ neck (a single metastatic node in an otherwise normal neck has the potential to have a major impact on treatment and so indeterminate nodes may require further evaluation) Assess nodes outside the usual treatment fields (contralateral neck; retropharyngeal, parotid, post-­auricular, suboccipital, buccal, facial and level VI nodes; also assess nodes adjacent to sites such as the salivary glands which may be preferentially spared during radiotherapy) Extranodal tumour spread: indicator of poor prognosis (unresectable disease may be present if there is nodal invasion of the common or internal carotid artery or pre- or paravertebral muscles; internal jugular vein, sternocleidomastoid muscle, submandibular gland and skin involvement should be documented although these structures can be resected) Advanced nodal disease is a risk factor for distant metastases (multiple, bulky and low-lying cervical nodes) Imaging techniques tend to perform well for excluding residual disease (high NPV) but poorly in differentiating residual abnormalities (low PPV). All modalities have advantages and disadvantages and the best performance is probably from multiparametric imaging using a combination of FDG-PET CT, conventional anatomical MRI sequences and DWI Surveillance: most nodal relapse/recurrence occurs within the first 3 years of treatment, with the majority of these cases being diagnosed in the first 2 years. Imaging follow-up is therefore usually performed every 4–6 months for 2–3 years, and is more frequent in the earlier post-treatment period

sarcoidosis, Rosai-Dorfman disease and Kikuchi-­ References Fujimoto disease (Koyama et al. 2004; La Barge 2008; Fulcher 1993; Na et al. 1997b). Abdel Razek AA, Soliman NY, Elkhamary S, Alsharaway

13

Conclusion

Imaging of cervical nodal metastases has an important role in nodal detection, treatment planning and post-treatment assessment. Imaging often requires a multimodality approach for the best results and a summary of the main points to be remembered when assessing nodal metastases is listed in Table 3.

MK, Tawfik A (2006) Role of diffusion-weighted MR imaging in cervical lymphadenopathy. Eur Radiol 16:1468–1477 Agrawal A, Civantos FJ, Brumund KT et al (2015) [(99m) Tc]Tilmanocept accurately detects sentinel lymph nodes and predicts node pathology status in patients with oral squamous cell carcinoma of the head and neck: results of a phase III multi-institutional trial. Ann Surg Oncol 22:3708–3715 Ahuja A, Ying M (2002) An overview of neck node sonography. Investig Radiol 37:333–342 Ahuja A, Ying M (2003) Sonography of neck lymph nodes. Part II: abnormal lymph nodes. Clin Radiol 58:359–366

Neck Nodal Disease Ahuja AT, Chow L, Chick W, King W, Metreweli C (1995) Metastatic cervical nodes in papillary carcinoma of the thyroid: ultrasound and histological correlation. Clin Radiol 50:229–231 Ahuja A, Ying M, King A, Yuen HY (2001a) Lymph node hilus: gray scale and power Doppler sonography of cervical nodes. J Ultrasound Med 20:987–992. quiz 994 Ahuja AT, Ying M, Ho SS, Metreweli C (2001b) Distribution of intranodal vessels in differentiating benign from metastatic neck nodes. Clin Radiol 56:197–201 Ahuja AT, Ying M, Ho SY et  al (2008) Ultrasound of malignant cervical lymph nodes. Cancer Imaging 8:48–56 Ai QY, King AD, Mo FKF et  al (2017) Prediction of distant metastases from nasopharyngeal carcinoma: improved diagnostic performance of MRI using nodal volume in N1 and N2 stage disease. Oral Oncol 69: 74–79 Ai QY, King AD, Mo FKF et  al (2018) Staging nodal metastases in nasopharyngeal carcinoma: which method should be used to measure nodal dimension on MRI? Clin Radiol 73:640–646 Ai QY, King AD, Poon DMC et  al (2019) Extranodal extension is a criterion for poor outcome in patients with metastatic nodes from cancer of the nasopharynx. Oral Oncol 88:124–130 Alkureishi LW, Ross GL, Shoaib T et al (2010) Sentinel node biopsy in head and neck squamous cell cancer: 5-year follow-up of a European multicenter trial. Ann Surg Oncol 17:2459–2464 Alvarez Amezaga J, Barbier Herrero L, Pijoan del Barrio JI et  al (2007) Diagnostic efficacy of sentinel node biopsy in oral squamous cell carcinoma. Cohort study and meta-analysis. Med Oral Patol Oral Cir Bucal 12:E235–E243 Alvi A, Johnson JT (1997) Development of distant metastasis after treatment of advanced-stage head and neck cancer. Head Neck 19:500–505 Amin MB, Edge SB, Greene FL et al (2017) AJCC cancer staging manual, 8th edn. Springer International Publishing, New York Andersen PE, Warren F, Spiro J et  al (2002) Results of selective neck dissection in management of the node-­ positive neck. Arch Otolaryngol Head Neck Surg 128:1180–1184 Apter S, Avigdor A, Gayer G, Portnoy O, Zissin R, Hertz M (2002) Calcification in lymphoma occurring before therapy: CT features and clinical correlation. AJR Am J Roentgenol 178:935–938 Ariji Y, Kimura Y, Hayashi N et al (1998) Power Doppler sonography of cervical lymph nodes in patients with head and neck cancer. AJNR Am J Neuroradiol 19:303–307 Armstrong JG, Harrison LB, Thaler HT et al (1992) The indications for elective treatment of the neck in cancer of the major salivary glands. Cancer 69:615–619

433 Aro K, Ho AS, Luu M et  al (2018) Development of a novel salivary gland cancer lymph node staging system. Cancer 124:3171–3180 Atula TS, Varpula MJ, Kurki TJ, Klemi PJ, Grenman R (1997) Assessment of cervical lymph node status in head and neck cancer patients: palpation, computed tomography and low field magnetic resonance imaging compared with ultrasound-guided fine-needle aspiration cytology. Eur J Radiol 25:152–161 Barchetti F, Pranno N, Giraldi G et al (2014) The role of 3 Tesla diffusion-weighted imaging in the differential diagnosis of benign versus malignant cervical lymph nodes in patients with head and neck squamous cell carcinoma. Biomed Res Int 2014:532095 Becker M, Varoquaux AD, Combescure C et  al (2018) Local recurrence of squamous cell carcinoma of the head and neck after radio(chemo)therapy: diagnostic performance of FDG-PET/MRI with diffusion-­ weighted sequences. Eur Radiol 28:651–663 Bhatia KS, Lee YY, Yuen EH, Ahuja AT (2013) Ultrasound elastography in the head and neck. Part II. Accuracy for malignancy. Cancer Imaging 13:260–276 Biau J, Lapeyre M, Troussier I et al (2019) Selection of lymph node target volumes for definitive head and neck radiation therapy: a 2019 update. Radiother Oncol 134:1–9 Bluemel C, Herrmann K, Kubler A et  al (2014) Intraoperative 3-D imaging improves sentinel lymph node biopsy in oral cancer. Eur J Nucl Med Mol Imaging 41:2257–2264 Bluemel C, Herrmann K, Giammarile F et  al (2015) EANM practice guidelines for lymphoscintigraphy and sentinel lymph node biopsy in melanoma. Eur J Nucl Med Mol Imaging 42:1750–1766 de Bondt RB, Nelemans PJ, Hofman PA et  al (2007) Detection of lymph node metastases in head and neck cancer: a meta-analysis comparing US, USgFNAC, CT and MR imaging. Eur J Radiol 64:266–272 de Bondt RB, Nelemans PJ, Bakers F et  al (2009a) Morphological MRI criteria improve the detection of lymph node metastases in head and neck squamous cell carcinoma: multivariate logistic regression analysis of MRI features of cervical lymph nodes. Eur Radiol 19:626–633 de Bondt RB, Hoeberigs MC, Nelemans PJ et al (2009b) Diagnostic accuracy and additional value of diffusion-­ weighted imaging for discrimination of malignant cervical lymph nodes in head and neck squamous cell carcinoma. Neuroradiology 51:183–192 Brantsch KD, Meisner C, Schonfisch B et  al (2008) Analysis of risk factors determining prognosis of cutaneous squamous-cell carcinoma: a prospective study. Lancet Oncol 9:713–720 de Bree R, Deurloo EE, Snow GB, Leemans CR (2000) Screening for distant metastases in patients with head and neck cancer. Laryngoscope 110:397–401 de Bree R, van der Putten L, Brouwer J, Castelijns JA, Hoekstra OS, Leemans CR (2009) Detection of

434 locoregional recurrent head and neck cancer after (chemo)radiotherapy using modern imaging. Oral Oncol 45:386–393 de Bree R, Takes RP, Castelijns JA et al (2015) Advances in diagnostic modalities to detect occult lymph node metastases in head and neck squamous cell carcinoma. Head Neck 37:1829–1839 van den Brekel MW, Stel HV, Castelijns JA et al (1990) Cervical lymph node metastasis: assessment of radiologic criteria. Radiology 177:379–384 van den Brekel MW, Stel HV, van der Valk P, van der Waal I, Meyer CJ, Snow GB (1992) Micrometastases from squamous cell carcinoma in neck dissection specimens. Eur Arch Otorhinolaryngol 249:349–353 van den Brekel MW, Castelijns JA, Stel HV, Golding RP, Meyer CJ, Snow GB (1993) Modern imaging techniques and ultrasound-guided aspiration cytology for the assessment of neck node metastases: a prospective comparative study. Eur Arch Otorhinolaryngol 250:11–17 van den Brekel MW, van der Waal I, Meijer CJ, Freeman JL, Castelijns JA, Snow GB (1996) The incidence of micrometastases in neck dissection specimens obtained from elective neck dissections. Laryngoscope 106:987–991 van den Brekel MW, Castelijns JA, Snow GB (1998a) The size of lymph nodes in the neck on sonograms as a radiologic criterion for metastasis: how reliable is it? AJNR Am J Neuroradiol 19:695–700 van den Brekel MW, Pameijer FA, Koops W, Hilgers FJ, Kroon BB, Balm AJ (1998b) Computed tomography for the detection of neck node metastases in melanoma patients. Eur J Surg Oncol 24:51–54 van den Brekel MW, Castelijns JA, Reitsma LC, Leemans CR, van der Waal I, Snow GB (1999) Outcome of observing the N0 neck using ultrasonographic-guided cytology for follow-up. Arch Otolaryngol Head Neck Surg 125:153–156 Brierley JD, Gospodarowicz MK, Wittekind C (2017) TNM classification of malignant tumours, 8th edn. Wiley-Blackwell, Navarra, Spain Brouwer J, de Bree R, Comans EF, Castelijns JA, Hoekstra OS, Leemans CR (2004) Positron emission tomography using [18F]fluorodeoxyglucose (FDG-PET) in the clinically negative neck: is it likely to be superior? Eur Arch Otorhinolaryngol 261:479–483 Brouwer J, de Bree R, Hoekstra OS et al (2005) Screening for distant metastases in patients with head and neck cancer: is chest computed tomography sufficient? Laryngoscope 115:1813–1817 Brouwer J, Senft A, de Bree R et  al (2006) Screening for distant metastases in patients with head and neck cancer: is there a role for (18)FDG-PET? Oral Oncol 42:275–280 Burke C, Thomas R, Inglis C et  al (2011) Ultrasound-­ guided core biopsy in the diagnosis of lymphoma of the head and neck. A 9 year experience. Br J Radiol 84:727–732 Bussels B, Hermans R, Reijnders A, Dirix P, Nuyts S, Van den Bogaert W (2006) Retropharyngeal nodes in

K. S. S. Bhatia and A. D. King squamous cell carcinoma of oropharynx: incidence, localization, and implications for target volume. Int J Radiat Oncol Biol Phys 65:733–738 Cantrell SC, Peck BW, Li G, Wei Q, Sturgis EM, Ginsberg LE (2013) Differences in imaging characteristics of HPV-positive and HPV-negative oropharyngeal cancers: a blinded matched-pair analysis. AJNR Am J Neuroradiol 34:2005–2009 Castelijns JA, van den Brekel MW (2002) Imaging of lymphadenopathy in the neck. Eur Radiol 12: 727–738 Chen Y, Ren W, Zheng D et al (2015) Diffusion kurtosis imaging predicts neoadjuvant chemotherapy responses within 4 days in advanced nasopharyngeal carcinoma patients. J Magn Reson Imaging 42:1354–1361 Cheson BD, Horning SJ, Coiffier B et  al (1999) Report of an international workshop to standardize response criteria for non-Hodgkin’s lymphomas. NCI sponsored international working group. J Clin Oncol 17:1244 Chisin R, Weber AL (1994) Imaging of lymphoma manifestations in the extracranial head and neck region. Leuk Lymphoma 12:177–189 Cho SJ, Suh CH, Baek JH, Chung SR, Choi YJ, Lee JH (2019) Diagnostic performance of CT in detection of metastatic cervical lymph nodes in patients with thyroid cancer: a systematic review and meta-analysis. Eur Radiol 29:4635–4647 Choi YJ, Lee JH, Baek JH (2015) Ultrasound elastography for evaluation of cervical lymph nodes. Ultrasonography 34:157–164 Cintra MB, Ricz H, Mafee MF, Dos Santos AC (2018) Magnetic resonance imaging: dynamic contrast enhancement and diffusion-weighted imaging to identify malignant cervical lymph nodes. Radiol Bras 51:71–75 Civantos FJ, Zitsch RP, Schuller DE et al (2010) Sentinel lymph node biopsy accurately stages the regional lymph nodes for T1-T2 oral squamous cell carcinomas: results of a prospective multi-institutional trial. J Clin Oncol 28:1395–1400 Collins SL (1987) Controversies in the management of cancer of the neck. In: Thawley SE, Panje WR (eds) Comprehensive management of head and neck tumours. Saunders, Philadelphia, PA, pp 1386–1443 Collins CD (2006) PET in lymphoma. Cancer Imaging 6:S63–S70 Curtin HD, Ishwaran H, Mancuso AA, Dalley RW, Caudry DJ, McNeil BJ (1998) Comparison of CT and MR imaging in staging of neck metastases. Radiology 207:123–130 Dai YL, King AD (2018) State of the art MRI in head and neck cancer. Clin Radiol 73:45–59 D'Cruz AK, Vaish R, Kapre N et al (2015) Elective versus therapeutic neck dissection in node-negative oral cancer. N Engl J Med 373:521–529 Dirix P, Nuyts S, Bussels B, Hermans R, Van den Bogaert W (2006) Prognostic influence of retropharyngeal lymph node metastasis in squamous cell carcinoma

Neck Nodal Disease of the oropharynx. Int J Radiat Oncol Biol Phys 65:739–744 Don DM, Anzai Y, Lufkin RB, Fu YS, Calcaterra TC (1995) Evaluation of cervical lymph node metastases in squamous cell carcinoma of the head and neck. Laryngoscope 105:669–674 Dragoni F, Cartoni C, Pescarmona E et al (1999) The role of high resolution pulsed and color Doppler ultrasound in the differential diagnosis of benign and malignant lymphadenopathy: results of multivariate analysis. Cancer 85:2485–2490 Eisenkraft BL, Som PM (1999) The spectrum of benign and malignant etiologies of cervical node calcification. AJR Am J Roentgenol 172:1433–1437 Evangelista L, Cervino AR, Chondrogiannis S et al (2014) Comparison between anatomical cross-sectional imaging and 18F-FDG PET/CT in the staging, restaging, treatment response, and long-term surveillance of squamous cell head and neck cancer: a systematic literature overview. Nucl Med Commun 35:123–134 Ferlito A, Rinaldo A, Devaney KO, Nakashiro K, Hamakawa H (2008) Detection of lymph node micrometastases in patients with squamous carcinoma of the head and neck. Eur Arch Otorhinolaryngol 265:1147–1153 Ferris RL, Stefanika P, Xi L, Gooding W, Seethala RR, Godfrey TE (2012) Rapid molecular detection of metastatic head and neck squamous cell carcinoma as an intraoperative adjunct to sentinel lymph node biopsy. Laryngoscope 122:1020–1030 Fischbein NJ, Noworolski SM, Henry RG, Kaplan MJ, Dillon WP, Nelson SJ (2003) Assessment of metastatic cervical adenopathy using dynamic contrast-enhanced MR imaging. AJNR Am J Neuroradiol 24:301–311 Flach GB, Tenhagen M, de Bree R et al (2013) Outcome of patients with early stage oral cancer managed by an observation strategy towards the N0 neck using ultrasound guided fine needle aspiration cytology: no survival difference as compared to elective neck dissection. Oral Oncol 49:157–164 Foust AM, Ali RM, Nguyen XV et al (2018) Dual-energy CT-derived iodine content and spectral attenuation analysis of metastatic versus nonmetastatic lymph nodes in squamous cell carcinoma of the oropharynx. Tomography 4:66–71 Frankenthaler RA, Byers RM, Luna MA, Callender DL, Wolf P, Goepfert H (1993) Predicting occult lymph node metastasis in parotid cancer. Arch Otolaryngol Head Neck Surg 119:517–520 Freeman SB (2005) Advanced cervical metastasis involving the carotid artery. Curr Opin Otolaryngol Head Neck Surg 13:107–111 Friedman M, Roberts N, Kirshenbaum GL, Colombo J (1993) Nodal size of metastatic squamous cell carcinoma of the neck. Laryngoscope 103:854–856 Fulcher AS (1993) Cervical lymphadenopathy due to Kikuchi disease: US and CT appearance. J Comput Assist Tomogr 17:131–133 Furukawa MK, Furukawa M (2010) Diagnosis of lymph node metastases of head and neck cancer and evalua-

435 tion of effects of chemoradiotherapy using ultrasonography. Int J Clin Oncol 15:23–32 Giammarile F, Alazraki N, Aarsvold JN et al (2013) The EANM and SNMMI practice guideline for lymphoscintigraphy and sentinel node localization in breast cancer. Eur J Nucl Med Mol Imaging 40:1932–1947 Giammarile F, Schilling C, Gnanasegaran G et al (2019) The EANM practical guidelines for sentinel lymph node localisation in oral cavity squamous cell carcinoma. Eur J Nucl Med Mol Imaging 46:623–637 Gormly K, Glastonbury CM (2004) Calcified nodal metastasis from squamous cell carcinoma of the head and neck. Australas Radiol 48:240–242 Goyal N, Zacharia TT, Goldenberg D (2012) Differentiation of branchial cleft cysts and malignant cystic adenopathy of pharyngeal origin. AJR Am J Roentgenol 199:W216–W221 Grabenbauer GG, Steininger H, Meyer M et  al (1998) Nodal CT density and total tumor volume as prognostic factors after radiation therapy of stage III/IV head and neck cancer. Radiother Oncol 47:175–183 Gregoire V, Ang K, Budach W et al (2014) Delineation of the neck node levels for head and neck tumors: a 2013 update. DAHANCA, EORTC, HKNPCSG, NCIC CTG, NCRI, RTOG, TROG consensus guidelines. Radiother Oncol 110:172–181 Gregoire V, Grau C, Lapeyre M, Maingon P (2018) Target volume selection and delineation (T and N) for primary radiation treatment of oral cavity, oropharyngeal, hypopharyngeal and laryngeal squamous cell carcinoma. Oral Oncol 87:131–137 Gritzmann N, Grasl MC, Helmer M, Steiner E (1990) Invasion of the carotid artery and jugular vein by lymph node metastases: detection with sonography. AJR Am J Roentgenol 154:411–414 Guntinas-Lichius O, Peter Klussmann J, Dinh S et  al (2006) Diagnostic work-up and outcome of cervical metastases from an unknown primary. Acta Otolaryngol 126:536–544 Gupta T, Master Z, Kannan S et  al (2011) Diagnostic performance of post-treatment FDG PET or FDG PET/CT imaging in head and neck cancer: a systematic review and meta-analysis. Eur J Nucl Med Mol Imaging 38:2083–2095 Haerle SK, Strobel K, Ahmad N, Soltermann A, Schmid DT, Stoeckli SJ (2011) Contrast-enhanced (1)(8) F-FDG-PET/CT for the assessment of necrotic lymph node metastases. Head Neck 33:324–329 Hanna E, Wanamaker J, Adelstein D, Tubbs R, Lavertu P (1997) Extranodal lymphomas of the head and neck. A 20-year experience. Arch Otolaryngol Head Neck Surg 123:1318–1323 Hauser T, Essig M, Jensen A et  al (2014) Prediction of treatment response in head and neck carcinomas using IVIM-DWI: evaluation of lymph node metastasis. Eur J Radiol 83:783–787 Hegde P, Roy S, Shetty T, Prasad BR, Shetty U (2017) Tumor infiltration depth as a prognostic parameter for nodal metastasis in oral squamous cell carcinoma. Int J Appl Basic Med Res 7:252–257

436 Herd MK, Woods M, Anand R, Habib A, Brennan PA (2012) Lymphoma presenting in the neck: current concepts in diagnosis. Br J Oral Maxillofac Surg 50:309–313 Hermann RM, Christiansen H, Rodel RM (2013) Lymph node positive head and neck carcinoma after curative radiochemotherapy: a long lasting debate on elective post-therapeutic neck dissections comes to a conclusion. Cancer Radiother 17:323–331 Heusch P, Sproll C, Buchbender C et al (2014) Diagnostic accuracy of ultrasound, (1)(8)F-FDG-PET/CT, and fused (1)(8)F-FDG-PET-MR images with DWI for the detection of cervical lymph node metastases of HNSCC. Clin Oral Investig 18:969–978 Hicks RJ, Mac Manus MP, Seymour JF (2005) Initial staging of lymphoma with positron emission tomography and computed tomography. Semin Nucl Med 35:165–175 Holzapfel K, Duetsch S, Fauser C, Eiber M, Rummeny EJ, Gaa J (2009) Value of diffusion-weighted MR imaging in the differentiation between benign and malignant cervical lymph nodes. Eur J Radiol 72:381–387 van der Hoorn A, van Laar PJ, Holtman GA, Westerlaan HE (2017) Diagnostic accuracy of magnetic resonance imaging techniques for treatment response evaluation in patients with head and neck tumors, a systematic review and meta-analysis. PLoS One 12:e0177986 Houghton DJ, Hughes ML, Garvey C et  al (1998) Role of chest CT scanning in the management of patients presenting with head and neck cancer. Head Neck 20:614–618 Hyde NC, Prvulovich E, Newman L, Waddington WA, Visvikis D, Ell P (2003) A new approach to pre-­ treatment assessment of the N0 neck in oral squamous cell carcinoma: the role of sentinel node biopsy and positron emission tomography. Oral Oncol 39:350–360 Inohara H, Enomoto K, Tomiyama Y et al (2009) The role of CT and (18)F-FDG PET in managing the neck in node-positive head and neck cancer after chemoradiotherapy. Acta Otolaryngol 129:893–899 Isles MG, McConkey C, Mehanna HM (2008) A systematic review and meta-analysis of the role of positron emission tomography in the follow up of head and neck squamous cell carcinoma following radiotherapy or chemoradiotherapy. Clin Otolaryngol 33:210–222 Johnson JT (1990) A surgeon looks at cervical lymph nodes. Radiology 175:607–610 de Juan J, Garcia J, Lopez M et  al (2013) Inclusion of extracapsular spread in the pTNM classification system: a proposal for patients with head and neck carcinoma. JAMA Otolaryngol Head Neck Surg 139:483–488 Katayama I, Sasaki M, Kimura Y et al (2012) Comparison between ultrasonography and MR imaging for discriminating squamous cell carcinoma nodes with extranodal spread in the neck. Eur J Radiol 81:3326–3331 Kim S, Loevner L, Quon H et  al (2009) Diffusion-­ weighted magnetic resonance imaging for predicting and detecting early response to chemoradiation ther-

K. S. S. Bhatia and A. D. King apy of squamous cell carcinomas of the head and neck. Clin Cancer Res 15:986–994 Kim S, Loevner LA, Quon H et  al (2010) Prediction of response to chemoradiation therapy in squamous cell carcinomas of the head and neck using dynamic contrast-­enhanced MR imaging. AJNR Am J Neuroradiol 31:262–268 Kimura Y, Sumi M, Sakihama N, Tanaka F, Takahashi H, Nakamura T (2008) MR imaging criteria for the prediction of extranodal spread of metastatic cancer in the neck. AJNR Am J Neuroradiol 29:1355–1359 King AD (2008) Imaging for staging and management of thyroid cancer. Cancer Imaging 8:57–69 King AD, Ahuja AT, Metreweli C (1999) MRI of tuberculous cervical lymphadenopathy. J Comput Assist Tomogr 23:244–247 King AD, Tse GM, Ahuja AT et  al (2004a) Necrosis in metastatic neck nodes: diagnostic accuracy of CT, MR imaging, and US. Radiology 230:720–726 King AD, Lei KI, Ahuja AT (2004b) MRI of neck nodes in non-Hodgkin’s lymphoma of the head and neck. Br J Radiol 77:111–115 King AD, Ahuja AT, Yeung DK et  al (2007) Malignant cervical lymphadenopathy: diagnostic accuracy of diffusion-­ weighted MR imaging. Radiology 245:806–813 King AD, Ma BB, Yau YY et al (2008) The impact of 18F-­ FDG PET/CT on assessment of nasopharyngeal carcinoma at diagnosis. Br J Radiol 81:291–298 King AD, Keung CK, Yu KH et  al (2013) T2-weighted MR imaging early after chemoradiotherapy to evaluate treatment response in head and neck squamous cell carcinoma. AJNR Am J Neuroradiol 34: 1237–1241 King AD, Yu KH, Mo FK et  al (2016) Cervical nodal metastases from head and neck squamous cell carcinoma: MRI criteria for treatment assessment. Head Neck 38(Suppl 1):E1598–E1604 Kligerman J, Lima RA, Soares JR et  al (1994) Supraomohyoid neck dissection in the treatment of T1/T2 squamous cell carcinoma of oral cavity. Am J Surg 168:391–394 Klussmann JP, Ponert T, Mueller RP, Dienes HP, Guntinas-­ Lichius O (2008) Patterns of lymph node spread and its influence on outcome in resectable parotid cancer. Eur J Surg Oncol 34:932–937 Koc O, Paksoy Y, Erayman I, Kivrak AS, Arbag H (2007) Role of diffusion weighted MR in the discrimination diagnosis of the cystic and/or necrotic head and neck lesions. Eur J Radiol 62:205–213 Korkmaz H, Yoo GH, Du W et  al (2002) Predictors of nodal metastasis in salivary gland cancer. J Surg Oncol 80:186–189 Koyama T, Ueda H, Togashi K, Umeoka S, Kataoka M, Nagai S (2004) Radiologic manifestations of sarcoidosis in various organs. Radiographics 24:87–104 Krabbe CA, Dijkstra PU, Pruim J et al (2008) FDG PET in oral and oropharyngeal cancer. Value for confirmation of N0 neck and detection of occult metastases. Oral Oncol 44:31–36

Neck Nodal Disease Kubicek GJ, Champ C, Fogh S et  al (2010) FDG-PET staging and importance of lymph node SUV in head and neck cancer. Head Neck Oncol 2:19 Kwee TC, Ludwig I, Uiterwaal CS et  al (2011) ADC measurements in the evaluation of lymph nodes in patients with non-Hodgkin lymphoma: feasibility study. MAGMA 24(1):1–8. https://doi.org/10.1007/ s10334-010-0226-7 La Barge DV 3rd, Salzman KL, Harnsberger HR et  al (2008) Sinus histiocytosis with massive lymphadenopathy (Rosai-Dorfman disease): imaging manifestations in the head and neck. AJR Am J Roentgenol 191:W299–W306 Labadie RF, Yarbrough WG, Weissler MC, Pillsbury HC, Mukherji SK (2000) Nodal volume reduction after concurrent chemo- and radiotherapy: correlation between initial CT and histopathologic findings. AJNR Am J Neuroradiol 21:310–314 Larsen SR, Johansen J, Sorensen JA, Krogdahl A (2009) The prognostic significance of histological features in oral squamous cell carcinoma. J Oral Pathol Med 38:657–662 Layland MK, Sessions DG, Lenox J (2005) The influence of lymph node metastasis in the treatment of squamous cell carcinoma of the oral cavity, oropharynx, larynx, and hypopharynx: N0 versus N+. Laryngoscope 115:629–639 Leemans CR, Tiwari R, Nauta JJ, van der Waal I, Snow GB (1993) Regional lymph node involvement and its significance in the development of distant metastases in head and neck carcinoma. Cancer 71:452–456 Lell M, Baum U, Greess H et  al (2000) Head and neck tumors: imaging recurrent tumor and post-therapeutic changes with CT and MRI. Eur J Radiol 33:239–247 Liang L, Luo X, Lian Z et al (2017) Lymph node metastasis in head and neck squamous carcinoma: efficacy of intravoxel incoherent motion magnetic resonance imaging for the differential diagnosis. Eur J Radiol 90:159–165 Liauw SL, Mancuso AA, Amdur RJ et  al (2006) Postradiotherapy neck dissection for lymph node-­ positive head and neck cancer: the use of computed tomography to manage the neck. J Clin Oncol 24:1421–1427 Lim HK, Lee JH, Baek HJ et  al (2014) Is diffusion-­ weighted MRI useful for differentiation of small non-­necrotic cervical lymph nodes in patients with head and neck malignancies? Korean J Radiol 15: 810–816 Lin D, Glastonbury CM, Rafaelian O, Eisele DW, Wang SJ (2007) Management of advanced nodal disease following chemoradiation for head and neck squamous cell carcinoma: role of magnetic resonance imaging. J Otolaryngol 36:350–356 Liu X, Ouyang D, Li H et al (2015) Papillary thyroid cancer: dual-energy spectral CT quantitative parameters for preoperative diagnosis of metastasis to the cervical lymph nodes. Radiology 275:167–176 Lloyd S, Yu JB, Ross DA, Wilson LD, Decker RH (2010) A prognostic index for predicting lymph node

437 metastasis in minor salivary gland cancer. Int J Radiat Oncol Biol Phys 76:169–175 Lu Y, Jansen JF, Stambuk HE et al (2013) Comparing primary tumors and metastatic nodes in head and neck cancer using intravoxel incoherent motion imaging: a preliminary experience. J Comput Assist Tomogr 37:346–352 Lydiatt WM, Patel SG, O’Sullivan B et  al (2017) Head and neck cancers—major changes in the American Joint Committee on Cancer eighth edition cancer staging manual. CA Cancer J Clin 67:122–137 Machens A, Hinze R, Thomusch O, Dralle H (2002) Pattern of nodal metastasis for primary and reoperative thyroid cancer. World J Surg 26:22–28 Mann WJ, Beck A, Schreiber J, Maurer J, Amedee RG, Gluckmann JL (1994) Ultrasonography for evaluation of the carotid artery in head and neck cancer. Laryngoscope 104:885–888 Marcus C, Whitworth PW, Surasi DS, Pai SI, Subramaniam RM (2014) PET/CT in the management of thyroid cancers. AJR Am J Roentgenol 202:1316–1329 Martens RM, Noij DP, Ali M et al (2019) Functional imaging early during (chemo)radiotherapy for response prediction in head and neck squamous cell carcinoma; a systematic review. Oral Oncol 88:75–83 Martinez-Gimeno C, Rodriguez EM, Vila CN, Varela CL (1995) Squamous cell carcinoma of the oral cavity: a clinicopathologic scoring system for evaluating risk of cervical lymph node metastasis. Laryngoscope 105:728–733 McMasters KM, Swetter SM (2003) Current management of melanoma: benefits of surgical staging and adjuvant therapy. J Surg Oncol 82:209–216 Medina JE (1998) Neck dissection in the treatment of cancer of major salivary glands. Otolaryngol Clin N Am 31:815–822 Merritt RM, Williams MF, James TH, Porubsky ES (1997) Detection of cervical metastasis. A meta-­ analysis comparing computed tomography with physical examination. Arch Otolaryngol Head Neck Surg 123: 149–152 Mizowaki T, Nishimura Y, Shimada Y et  al (1996) Optimal size criteria of malignant lymph nodes in the treatment planning of radiotherapy for esophageal cancer: evaluation by computed tomography and magnetic resonance imaging. Int J Radiat Oncol Biol Phys 36:1091–1098 Moon WK, Han MH, Chang KH et al (1997) CT and MR imaging of head and neck tuberculosis. Radiographics 17:391–402 Munck JN, Cvitkovic E, Piekarski JD et  al (1991) Computed tomographic density of metastatic lymph nodes as a treatment-related prognostic factor in advanced head and neck cancer. J Natl Cancer Inst 83:569–575 Mundada P, Varoquaux AD, Lenoir V et al (2018) Utility of MRI with morphologic and diffusion weighted imaging in the detection of post-treatment nodal disease in head and neck squamous cell carcinoma. Eur J Radiol 101:162–169

438 Myers JN, Greenberg JS, Mo V, Roberts D (2001) Extracapsular spread. A significant predictor of treatment failure in patients with squamous cell carcinoma of the tongue. Cancer 92:3030–3036 Na DG, Lim HK, Byun HS, Kim HD, Ko YH, Baek JH (1997a) Differential diagnosis of cervical lymphadenopathy: usefulness of color Doppler sonography. AJR Am J Roentgenol 168:1311–1316 Na DG, Chung TS, Byun HS, Kim HD, Ko YH, Yoon JH (1997b) Kikuchi disease: CT and MR findings. AJNR Am J Neuroradiol 18:1729–1732 Na DG, Lee JH, Jung SL et  al (2012) Radiofrequency ablation of benign thyroid nodules and recurrent thyroid cancers: consensus statement and recommendations. Korean J Radiol 13:117–125 Ng SH, Liao CT, Lin CY et al (2016) Dynamic contrast-­ enhanced MRI, diffusion-weighted MRI and (18) F-FDG PET/CT for the prediction of survival in oropharyngeal or hypopharyngeal squamous cell carcinoma treated with chemoradiation. Eur Radiol 26:4162–4172 Nishikado A, Kawata R, Haginomori SI et  al (2018) A clinicopathological study of parotid carcinoma: 18-year review of 171 patients at a single institution. Int J Clin Oncol 23:615–624 Ojiri H, Mendenhall WM, Stringer SP, Johnson PL, Mancuso AA (2002) Post-RT CT results as a predictive model for the necessity of planned post-RT neck dissection in patients with cervical metastatic disease from squamous cell carcinoma. Int J Radiat Oncol Biol Phys 52:420–428 Park HL, Yoo Ie R, Lee N et al (2013) The value of F-18 FDG PET for planning treatment and detecting recurrence in malignant salivary gland tumors: comparison with conventional imaging studies. Nucl Med Mol Imaging 47:242–248 Park JE, Lee JH, Ryu KH et al (2017) Improved diagnostic accuracy using arterial phase CT for lateral cervical lymph node metastasis from papillary thyroid Cancer. AJNR Am J Neuroradiol 38:782–788 Partovi S, Kohan A, Vercher-Conejero JL et  al (2014) Qualitative and quantitative performance of (1)(8) F-FDG-PET/MRI versus (1)(8)F-FDG-PET/CT in patients with head and neck cancer. AJNR Am J Neuroradiol 35:1970–1975 Pentenero M, Gandolfo S, Carrozzo M (2005) Importance of tumor thickness and depth of invasion in nodal involvement and prognosis of oral squamous cell carcinoma: a review of the literature. Head Neck 27:1080–1091 Pfister DG, Ang K, Brockstein B et al (2000) NCCN practice guidelines for head and neck cancers. Oncology (Williston Park) 14:163–194 Pillsbury HC 3rd, Clark M (1997) A rationale for therapy of the N0 neck. Laryngoscope 107:1294–1315 Qamar S, King AD, Ai QY et  al (2019) Amide proton transfer MRI detects early changes in nasopharyngeal carcinoma: providing a potential imaging marker for treatment response. Eur Arch Otorhinolaryngol 276:505–512

K. S. S. Bhatia and A. D. King Regis De Brito Santos I, Kowalski LP, Cavalcante De Araujo V, Flavia Logullo A, Magrin J (2001) Multivariate analysis of risk factors for neck metastases in surgically treated parotid carcinomas. Arch Otolaryngol Head Neck Surg 127:56–60 Robbins KT, Clayman G, Levine PA et  al (2002) Neck dissection classification update: revisions proposed by the American Head and Neck Society and the American Academy of Otolaryngology-Head and Neck Surgery. Arch Otolaryngol Head Neck Surg 128:751–758 Rodriguez-Cuevas S, Labastida S, Baena L, Gallegos F (1995) Risk of nodal metastases from malignant salivary gland tumors related to tumor size and grade of malignancy. Eur Arch Otorhinolaryngol 252: 139–142 Roh JL, Yeo NK, Kim JS et al (2007) Utility of 2-[18F] fluoro-2-deoxy-D-glucose positron emission tomography and positron emission tomography/computed tomography imaging in the preoperative staging of head and neck squamous cell carcinoma. Oral Oncol 43:887–893 Rosario PW, de Faria S, Bicalho L et  al (2005) Ultrasonographic differentiation between metastatic and benign lymph nodes in patients with papillary thyroid carcinoma. J Ultrasound Med 24:1385–1389 Rosenkrantz AB, Padhani AR, Chenevert TL et al (2015) Body diffusion kurtosis imaging: basic principles, applications, and considerations for clinical practice. J Magn Reson Imaging 42:1190–1202 Saito A, Takashima S, Takayama F, Kawakami S, Momose M, Matsushita T (2001) Spontaneous extensive necrosis in non-Hodgkin lymphoma: prevalence and clinical significance. J Comput Assist Tomogr 25:482–486 Sanguineti G, Corvo R, Benasso M et  al (1999) Management of the neck after alternating chemoradiotherapy for advanced head and neck cancer. Head Neck 21:223–228 Schilling C, Stoeckli SJ, Haerle SK et al (2015) Sentinel European node trial (SENT): 3-year results of sentinel node biopsy in oral cancer. Eur J Cancer 51:2777–2784 Schmitz S, Machiels JP, Weynand B, Gregoire V, Hamoir M (2009) Results of selective neck dissection in the primary management of head and neck squamous cell carcinoma. Eur Arch Otorhinolaryngol 266: 437–443 Schouten CS, de Graaf P, Alberts FM et  al (2015) Response evaluation after chemoradiotherapy for advanced nodal disease in head and neck cancer using diffusion-weighted MRI and 18F-FDG-PET-CT. Oral Oncol 51:541–547 Seidler M, Forghani B, Reinhold C et  al (2019) Dual-­ energy CT texture analysis with machine learning for the evaluation and characterization of cervical lymphadenopathy. Comput Struct Biotechnol J 17:1009–1015 Sharma D, Koshy G, Grover S, Sharma B (2017) Sentinel lymph node biopsy: a new approach in the management of head and neck cancers. Sultan Qaboos Univ Med J 17:e3–e10

Neck Nodal Disease Shear M, Hawkins DM, Farr HW (1976) The prediction of lymph node metastases from oral squamous carcinoma. Cancer 37:1901–1907 Shozushima M, Suzuki M, Nakasima T, Yanagisawa Y, Sakamaki K, Takeda Y (1990) Ultrasound diagnosis of lymph node metastasis in head and neck cancer. Dentomaxillofac Radiol 19:165–170 Snow GB, Annyas AA, van Slooten EA, Bartelink H, Hart AA (1982) Prognostic factors of neck node metastasis. Clin Otolaryngol Allied Sci 7:185–192 Som PM, Urken ML, Biller H, Lidov M (1993) Imaging the postoperative neck. Radiology 187:593–603 Som PM, Brandwein M, Lidov M, Lawson W, Biller HF (1994) The varied presentations of papillary thyroid carcinoma cervical nodal disease: CT and MR findings. AJNR Am J Neuroradiol 15:1123–1128 Som PM, Curtin HD, Mancuso AA (1999) An imaging-­ based classification for the cervical nodes designed as an adjunct to recent clinically based nodal classifications. Arch Otolaryngol Head Neck Surg 125: 388–396 Souter MA, Allison RS, Clarkson JH, Cowan IA, Coates MH, Wells JE (2009) Sensitivity and specificity of computed tomography for detection of extranodal spread from metastatic head and neck squamous cell carcinoma. J Laryngol Otol 123:778–782 Steinkamp HJ, Maurer J, Cornehl M, Knobber D, Hettwer H, Felix R (1994) Recurrent cervical lymphadenopathy: differential diagnosis with color-duplex sonography. Eur Arch Otorhinolaryngol 251:404–409 Steinkamp HJ, Cornehl M, Hosten N, Pegios W, Vogl T, Felix R (1995) Cervical lymphadenopathy: ratio of long- to short-axis diameter as a predictor of malignancy. Br J Radiol 68:266–270 Steinkamp HJ, van der Hoeck E, Bock JC, Felix R (1999) The extracapsular spread of cervical lymph node metastases: the diagnostic value of computed tomography. Rofo 170:457–462 Steinkamp HJ, Beck A, Werk M, Felix R (2002) Extracapsular spread of cervical lymph node metastases: diagnostic value of magnetic resonance imaging. Rofo 174:50–55 Stennert E, Kisner D, Jungehuelsing M et al (2003) High incidence of lymph node metastasis in major salivary gland cancer. Arch Otolaryngol Head Neck Surg 129:720–723 Stuckensen T, Kovacs AF, Adams S, Baum RP (2000) Staging of the neck in patients with oral cavity squamous cell carcinomas: a prospective comparison of PET, ultrasound, CT and MRI.  J Craniomaxillofac Surg 28:319–324 Su Z, Duan Z, Pan W et al (2016) Predicting extracapsular spread of head and neck cancers using different imaging techniques: a systematic review and meta-analysis. Int J Oral Maxillofac Surg 45:413–421 Suh CH, Choi YJ, Baek JH, Lee JH (2018) The diagnostic value of diffusion-weighted imaging in differentiating metastatic lymph nodes of head and neck squamous cell carcinoma: a systematic review and meta-­analysis. AJNR Am J Neuroradiol 39:1889–1895

439 Sumi M, Sakihama N, Sumi T et al (2003) Discrimination of metastatic cervical lymph nodes with diffusion-­ weighted MR imaging in patients with head and neck cancer. AJNR Am J Neuroradiol 24:1627–1634 Sun R, Tang X, Yang Y, Zhang C (2015) (18)FDG-PET/ CT for the detection of regional nodal metastasis in patients with head and neck cancer: a meta-analysis. Oral Oncol 51:314–320 Tankere F, Camproux A, Barry B, Guedon C, Depondt J, Gehanno P (2000) Prognostic value of lymph node involvement in oral cancers: a study of 137 cases. Laryngoscope 110:2061–2065 Tawfik AM, Razek AA, Kerl JM, Nour-Eldin NE, Bauer R, Vogl TJ (2014) Comparison of dual-energy CT-derived iodine content and iodine overlay of normal, inflammatory and metastatic squamous cell carcinoma cervical lymph nodes. Eur Radiol 24:574–580 Terhaard CH, Lubsen H, Van der Tweel I et  al (2004) Salivary gland carcinoma: independent prognostic factors for locoregional control, distant metastases, and overall survival: results of the Dutch head and neck oncology cooperative group. Head Neck 26:681–692; discussion 692–683 Traynor SJ, Cohen JI, Gray J, Andersen PE, Everts EC (1996) Selective neck dissection and the management of the node-positive neck. Am J Surg 172:654–657 Vandecaveye V, Hermans R (2011) Diffusion weighted magnetic resonance imaging early after (chemo)radiotherapy to monitor treatment response in head and neck squamous cell carcinoma. Int J Rad Oncol Biol Phys 83(2):1098–1107 Vandecaveye V, De Keyzer F, Vander Poorten V et  al (2009) Head and neck squamous cell carcinoma: value of diffusion-weighted MR imaging for nodal staging. Radiology 251:134–146 Vandecaveye V, Dirix P, De Keyzer F et  al (2010) Predictive value of diffusion-weighted magnetic resonance imaging during chemoradiotherapy for head and neck squamous cell carcinoma. Eur Radiol 20:1703–1714 Vander Poorten VL, Marchal F, Nuyts S, Clement PM (2010) Parotid carcinoma: current diagnostic workup and treatment. Indian J Surg Oncol 1:96–111 Vassallo P, Wernecke K, Roos N, Peters PE (1992) Differentiation of benign from malignant superficial lymphadenopathy: the role of high-resolution US. Radiology 183:215–220 Wang HM, Ng SH, Wang CH, Liaw CC, Tsai MH, Lai GM (1996) Correlation between computed tomographic density of lymph node metastases and response to cisplatin-based chemotherapy in patients with head and neck squamous cell carcinoma in an area in which betel quid chewing is prevalent. Cancer 78: 1972–1979 Wang P, Yu Q, Shi H (1999) CT findings of non-Hodgkin lymphoma in the head and neck. Zhonghua Kou Qiang Yi Xue Za Zhi 34:208–210 Wang YJ, Xu XQ, Hu H et al (2018) Histogram analysis of apparent diffusion coefficient maps for the differentiation between lymphoma and metastatic lymph

440 nodes of squamous cell carcinoma in head and neck region. Acta Radiol 59:672–680 Weber AL, Rahemtullah A, Ferry JA (2003) Hodgkin and non-Hodgkin lymphoma of the head and neck: clinical, pathologic, and imaging evaluation. Neuroimaging Clin N Am 13:371–392 Weiss MH, Harrison LB, Isaacs RS (1994) Use of decision analysis in planning a management strategy for the stage N0 neck. Arch Otolaryngol Head Neck Surg 120:699–702 Wong RJ (2008) Current status of FDG-PET for head and neck cancer. J Surg Oncol 97:649–652 Wreesmann VB, Katabi N, Palmer FL et  al (2016) Influence of extracapsular nodal spread extent on prognosis of oral squamous cell carcinoma. Head Neck 38(Suppl 1):E1192–E1199 Wu CH, Chang YL, Hsu WC, Ko JY, Sheen TS, Hsieh FJ (1998) Usefulness of Doppler spectral analysis and power Doppler sonography in the differentiation of cervical lymphadenopathies. AJR Am J Roentgenol 171:503–509 Yan S, Wang Z, Li L et  al (2016) Characterization of cervical lymph nodes using DCE-MRI: differentiation between metastases from SCC of head and neck and benign lymph nodes. Clin Hemorheol Microcirc 64:213–222 Yao M, Graham MM, Smith RB et al (2004) Value of FDG PET in assessment of treatment response and surveil-

K. S. S. Bhatia and A. D. King lance in head-and-neck cancer patients after intensity modulated radiation treatment: a preliminary report. Int J Radiat Oncol Biol Phys 60:1410–1418 Yeung AR, Liauw SL, Amdur RJ et  al (2008) Lymph node-positive head and neck cancer treated with definitive radiotherapy: can treatment response determine the extent of neck dissection? Cancer 112:1076–1082 Ying M, Ahuja A, Metreweli C (1998) Diagnostic accuracy of sonographic criteria for evaluation of cervical lymphadenopathy. J Ultrasound Med 17:437–445 Ying M, Ahuja A, Brook F, Metreweli C (2001) Vascularity and grey-scale sonographic features of normal cervical lymph nodes: variations with nodal size. Clin Radiol 56:416–419 Ying M, Ahuja A, Brook F (2002) Sonographic appearances of cervical lymph nodes: variations by age and sex. J Clin Ultrasound 30:1–11 Yoo GH, Hocwald E, Korkmaz H et al (2000) Assessment of carotid artery invasion in patients with head and neck cancer. Laryngoscope 110:386–390 Yousem DM, Hatabu H, Hurst RW et  al (1995) Carotid artery invasion by head and neck masses: prediction with MR imaging. Radiology 195:715–720 Yusa H, Yoshida H, Iwasa S et al (2000) Ultrasonographic assessment for response to radiochemotherapy of metastatic cervical lymph nodes in head and neck cancer: usefulness of grey-scale and color Doppler sonography. Ultrasound Med Biol 26:1081–1087

Neck Lymphoma Frank A. Pameijer and Rick L. M. Haas

Contents 1 1.1  1.2  1.3 

Introduction Epidemiology Etiology Pathology and Classifications

 442  442  442  442

2

Hodgkin’s Lymphoma

 443

 on-Hodgkin’s Lymphomas (NHL) N and Specific Entities 3.1  B Cell Neoplasms 3.2  T Cell and Natural Killer (NK)-Cell Neoplasms 3.3  Hodgkin’s Lymphoma (Hodgkin’s Disease) 3

 444  444

Workup Diagnosis Initial Imaging Staging

 445  445  445  447

5

Treatment

 449

6

Response Assessment

 449

8

Extranodal Disease

9

Conclusion

References

 452  455  457  460  461  462  462  463  463

 444  445

4 4.1  4.2  4.3 

7 N  odal Disease 7.1  T  he Common Sites 7.2  The Uncommon Sites

8.1  W  aldeyer’s Ring and the Upper Aerodigestive Tract 8.2  Orbit 8.3  Salivary Glands 8.4  Sinonasal Cavities 8.5  Thyroid 8.6  Bone 8.7  Skin

 450  450  451  451

F. A. Pameijer (*) Department of Radiology, University Medical Center Utrecht, Utrecht, The Netherlands e-mail: [email protected] R. L. M. Haas Department of Radiotherapy, The Netherlands Cancer Institute—Antoni van Leeuwenhoek Hospital, Amsterdam, The Netherlands Department of Radiotherapy, The Leiden University Medical Center, Leiden, The Netherlands

Abstract

Lymphomas represent about 5% of all malignant neoplasms of the head and neck area. Frequently, lymphoma is not limited to the head and neck region, but also involves other parts of the body. The workup of a patient with lymphoma requires a multidisciplinary approach. This chapter approaches lymphoma as a systemic disease describing the imaging workup of patients with newly diagnosed head and neck lymphoma. This includes discussion of commonly applied imaging techniques such as chest X-ray, computed tomography (CT), ultrasound, magnetic resonance imaging, and metabolic imaging (PET-CT). Both nodal disease and non-nodal lymphoma manifestations (e.g., in Waldeyer’s ring, larynx, orbit, lacrimal gland, parotid gland, sinonasal

Med Radiol Diagn Imaging (2020) https://doi.org/10.1007/174_2020_232, © Springer Nature Switzerland AG Published Online: 23 April 2020

441

F. A. Pameijer and R. L. M. Haas

442

cavities, thyroid, and bone) are discussed and illustrated with examples from daily practice. Frequently, imaging findings are nonspecific. However, some imaging patterns can suggest the diagnosis of lymphoma.

1  Introduction The most frequent group of neoplasms in the neck is the carcinomas, followed by the lymphomas. Only 5% of all neoplasms in the neck are malignant lymphomas. Lymphomas are neoplasms of the lymphoreticular system. They arise from lymphocytes and their derivates. Hodgkin’s lymphoma (HL) and non-Hodgkin’s lymphoma (NHL) are the most common malignancies of the hematopoietic system observed in the head and neck. Frequently, lymphoma is not limited to the head and neck region but also involves other parts of the body. This chapter approaches lymphoma as a systemic disease that can manifest itself in many forms in the head and neck. In many instances, the imaging findings are nonspecific and tissue sampling remains the mainstay of making the diagnosis. However, some imaging patterns can strongly suggest the diagnosis of lymphoma.

1.1  Epidemiology Lymphomas are the fifth most frequently occurring type of cancer in The United States comprising 5–6% of all malignancies. Hodgkin’s lymphoma (HL) accounts for about 10% of all lymphomas and the remaining 90% are referred to as non-Hodgkin’s lymphoma (NHL). Around 65,540 new cases of NHL were diagnosed in the United States in 2007 (Shankland et  al. 2012). The annual incidence rate of lymphomas in 1995–1999 was 19 cases per 100,000 persons but seems to rise gradually (Walter et al. 2015). Rates of lymphomas historically have been about 40% higher in urban counties than in rural counties. The incidence rates of lymphomas in the United States tend to exceed those of most other countries. The rapid increase (1–5% annually) in

lymphoma incidence in the 1970 and 1980s has been exceeded only by the increase in lung cancer in women and malignant melanomas in both sexes. The increase has been seen for males and females, for whites and blacks, and in all age groups (especially over 65 years of age), except for children. Malignant lymphomas in the head and neck usually presents as a neck lump, caused by lymphadenopathy (Cartwright et  al. 1999; Vose et al. 2002).

1.2  Etiology The majority of lymphomas arise in lymph nodes, but primary extranodal disease now accounts for 20–30% of all cases. The most frequent primary extranodal sites are the stomach, small intestine, skin, and brain. The incidence of extranodal disease has increased more rapidly than nodal disease. Immunodeficiency, including both congenital and acquired conditions, especially in patients on immunosuppressive drugs after organ transplantation, is strongly associated with an increased lymphoma risk (Pickhardt and Wippold 2nd 1999). Viruses like the Epstein– Barr virus appear to be important cofactors. A history of lymphatic malignancies in close relatives has been repeatedly shown to increase the risk of NHL by two- to threefold. Lymphomas may also cluster within families, not by an inherited genetic susceptibility, but because of shared environmental determinants (Cartwright et  al. 1999; Vose et al. 2002).

1.3  Pathology and Classifications In 1832, Thomas Hodgkin described a disease “lymphogranulomatosis maligna” that nowadays bears his name. From that moment on new lymphoma subtypes were recognized that closely resembled but were not exactly the disease discovered by Thomas Hodgkin, and were therefore epiphrased as NHL. Since the early 1960s in the previous century, several pathologists have attempted to produce classification systems in order to clarify the growing group of NHLs. Each

Neck Lymphoma

443

Table 1  Nomenclature and prevalence of the four most important lymphoma entities Official name from the WHO Classification (abbreviation) Diffuse large B cell lymphoma (DLBCL) Follicular lymphoma (FL)

Marginal zone lymphoma of MALT-type (MZL) Mantle cell lymphoma (MCL)

Synonyms DLCL Follicle center cell lymphoma (FCC), follicular centricity (cc) follicular centrocytic/ centroblastic (cc/cb) MALT lymphoma, maltoma (Relatively newly recognized entity)

new system was a revision of its predecessor adding new concepts and providing new names to the same entities. In 1982 a consensus system called the “Working Formulation” (WF) was produced combining several previous systems: the British National Lymphoma Investigation’s Classification, the Rappaport Classification, the Lukes and Collins Classification, the Kiel Classification, the Dorfman Classification, and the first World Health Organization (WHO) Classification. The WF lumped all diseases into three groups depending on their clinical behavior; the low-, intermediate-, and high-grade lymphomas abbreviated as LG-NHL, IG-NHL, and HG-NHL. The combination of several subtypes in just three groups was intended to help the clinician to distinguish lymphomas with more or less the same clinical course and comparable patient management. Soon, this lumping of different diseases in three groups without any biological basis was found to be artificial. In 1994, the WF was replaced by the Revised European-American Classification of Lymphoid Neoplasms, abbreviated as the REAL Classification (Harris et  al. 1994). The basis of this system was a better understanding of the putative normal counterparts and their functions within specific sites of a normal lymph node, like the follicles, the germinal centers, and the mantle- and marginal zones. The REAL Classification is based on a subdivision in three groups: B cell neoplasms, T cell/ natural killer (NK) cell neoplasms, and Hodgkin’s disease. The subsequent WHO Classifications (e.g., the 2001 and 2016 editions) are a further refinement of the REAL system.

Malignancy grade Intermediate-­grade NHL Low-grade NHL

Prevalence among the lymphomas (%) 30–40 15–25

Low-grade NHL

~10

Intermediate-­grade NHL

~5

Clinicians still tend to communicate in terminology, which is a mixture derived from the WF, the REAL, and the WHO systems. Table 1 simplifies the four most important lymphoma entities, their nomenclature, and their prevalence among the lymphomas (Harris et al. 2000; WHO 2001; Swerdlow et al. 2016). The indolent lymphomas are the former LG-NHL, the aggressive lymphomas used to be called IG-NHL or HG-NHL.

2  Hodgkin’s Lymphoma Originally, Thomas Hodgkin described in 1832 a disease “lymphogranulomatosis maligna” that afterward was called Hodgkin’s disease. Since the WHO Classification of 2001, this entity is called HL. Historically, HL is further divided into four subtypes (see also Sect. 3; WHO classification Sect. 3.3, 1(a)–(d)): 1 . Nodular sclerosis (NS-HL; 40–60% of all HL) 2. Lymphocyte predominant (LP-HL; ~2% of all HL) 3. Mixed cellularity (MC-HL; 30–40% of all HL) 4. Lymphocyte depleted (LD-HL; ~5% of all HL) Most HL patients present in early stages predominantly with lymphadenopathy in the neck and upper mediastinum (see also Sect. 7). Usually, these patients complain about a neck lump with or without systemic symptoms such as night sweats, fever, and weight loss. Extranodal spread in HL is very rare, splenic involvement being the most prevalent. Usually, only patients in advanced stage disease show organ

444

involvement by HL. In those cases bone marrow infiltration, intrapulmonary spread (either as deposits or as a direct extension from hilar nodes into the lung parenchyma), liver and other organ involvement can be seen. From an imaging point of view, the following aspects should be emphasized: • Neck nodes involved by HL usually have the same appearance as other lymphadenopathies. Sometimes, the imaging presentation is more characteristic and the diagnosis of lymphoma can be suggested by the radiologist (Table 3). • NS-HL poses a problem in response assessment after therapy. Histologically, the malignant component of HL is embedded in a large amount of fibrous tissues (hence the name “sclerosis”). Frequently, residual masses are seen on ultrasound, CT, or MRI investigations of the neck and mediastinum after successful treatment of NS-HL.  These masses, though diagnosed as “partial response” on imaging do not necessarily contain viable lymphoma.

3  Non-Hodgkin’s Lymphomas (NHL) and Specific Entities Worldwide, the incidence of NHL is about 5–10 times higher than the incidence of HL, largely dependent on regional differences. Of all cases of NHL, 80% is of B cell origin and 20% is T cell derived (Cartwright et al. 1999; Vose et al. 2002). The WHO Classification (Swerdlow et  al. 2016) of lymphoid neoplasms describes the following diseases:

3.1  B Cell Neoplasms 1. Precursor B cell neoplasm (a) Precursor B lymphoblastic leukemia/ lymphoma (precursor B cell acute lymphoblastic leukemia) 2. Mature (peripheral) B cell neoplasms (a) B cell chronic lymphocytic leukemia/ small lymphocytic lymphoma (b) B cell prolymphocytic leukemia

F. A. Pameijer and R. L. M. Haas

(c) Lymphoplasmacytic lymphoma (d) Splenic marginal zone B cell lymphoma (±villous lymphocytes) (e) Hairy cell leukemia (f) Plasma cell myeloma/plasmacytoma (g) Extranodal marginal zone B cell lym phoma of mucosa-associated lymphoid tissues (MALT) type (h) Nodal marginal zone B cell lymphoma (±monocytoid B cells) (i) Follicular lymphoma (j) Mantle-cell lymphoma (k) Diffuse large B cell lymphoma (l) Mediastinal large B cell lymphoma (m) Primary effusion lymphoma (n) Burkitt’s lymphoma/Burkitt cell leukemia

3.2  T  Cell and Natural Killer (NK)Cell Neoplasms 1. Precursor T cell neoplasm (a) Precursor T lymphoblastic lymphoma/ leukemia (precursor T cell acute lymphoblastic leukemia) 2. Mature (peripheral) T cell neoplasms (a) T cell prolymphocytic leukemia (b) T cell granular lymphocytic leukemia (c) Aggressive NK-cell leukemia (d) Adult T cell lymphoma/leukemia (HTLV1+) (e) Extranodal NK/T cell lymphoma, nasal type (f) Enteropathy-type T cell lymphoma (g) Hepatosplenic gamma–delta T cell lymphoma (h) Subcutaneous panniculitis-like T cell lymphoma (i) Mycosis fungoides/Sezary syndrome (j) Anaplastic large-cell lymphoma, T/null cell, primary cutaneous type (k) Peripheral T cell lymphoma, not otherwise characterized (l) Angioimmunoblastic T cell lymphoma (m) Anaplastic large-cell lymphoma, T/null cell, primary systemic type

Neck Lymphoma

445

Table 2  Ann Arbor Staging system for Hodgkin’s lymphoma Stage I II

III

IV

Suffices

Description Involvement of a single lymph node region II = involvement of more than one lymphatic region on only one side of the diaphragm, IIE = localized involvement of one extralymphatic organ or site and its regional lymph nodes with or without other nodes on the same side of the diaphragm, IIS = involvement of more than one lymphatic region on only one side of the diaphragm plus involvement of the spleen, IIES = both III = involvement of lymph node regions on both sides of the diaphragm, IIIE = involvement of lymph node regions on both sides of the diaphragm plus localized involvement of an extralymphatic organ or site, IIIS = involvement of lymph node regions on both sides of the diaphragm plus involvement of the spleen, IIIES = both Diffuse or disseminated involvement of one or more extralymphatic organs or tissues with or without associated lymph node enlargement. Organs considered distant include liver, bone, bone marrow, lung and/or pleura, and kidney A Without symptoms B With symptoms Night sweats Unexplained fever >38 °C Unexplained weight loss >10%, within last 6 months

3.3  Hodgkin’s Lymphoma (Hodgkin’s Disease) 1. Nodular lymphocyte-predominant Hodgkin’s lymphoma 2. Classical Hodgkin’s lymphoma (a) Nodular sclerosis Hodgkin’s lymphoma (grades 1 and 2) (b) Lymphocyte-rich classical Hodgkin’s lymphoma (c) Mixed cellularity Hodgkin’s lymphoma (d) Lymphocyte depletion Hodgkin’s lymphoma Note: The follicular (Sect. 3.1, 2(i)) and diffuse large B cell lymphomas (Sect. 3.1, 2(k)) make up more than half of all lymphomas diagnosed (see also Table 1).

4  Workup The workup of a patient with lymphoma requires a multidisciplinary approach involving close cooperation among the medical oncologist or hematologist–oncologist, the head and neck surgeon, the radiation oncologist, the pathologist, and the radiologist.

4.1  Diagnosis The diagnostic process to fully document new lymphoma patients, whether or not localized in the neck, comprises the following minimum requirements: • Full history (including presence of fever, night sweats and weight-loss; the so-called B symptoms) and physical examination with emphasis on all peripheral lymph node regions (neck, axilla and groin), examination of Waldeyer’s ring, liver, spleen, and skin. • Full blood count including a differential count of the leukocytes. • Representative excision biopsy of an entire enlarged lymph node (in nodal disease). • Bone marrow biopsy. • Imaging of all lymphatic regions.

4.2  Initial Imaging Standard radiological workup of a newly diagnosed patient with head and neck lymphoma should include: • Posteroanterior and lateral chest X-ray.

F. A. Pameijer and R. L. M. Haas

446

• Contrast-enhanced computed tomography (CECT) of the neck, chest, abdomen, and pelvis. • Ultrasound of the neck may be useful as an adjunct study. • Magnetic resonance imaging (MRI) may substitute CECT of the neck, especially if Waldeyer’s ring is involved or extranodal head and neck disease is present.

a

b

This imaging strategy is usually sufficient to stage a lymphoma patient. CECT and MRI have completely replaced lymphangiography as a diagnostic tool in lymphoma patients. In 2014, fluorodeoxyglucose (FDG) positron emission tomography (PET) computed tomography (CT) was formally accepted into standard staging for (FDG-avid) lymphomas (Cheson et  al. 2014). FDG PET-CT is also used for response

c

d

Fig. 1  Patient with Hodgkin’s lymphoma (HL). Initial staging with MRI of the neck shows extensive lymphadenopathy. (a) Enlarged left retropharyngeal node. (b) Bilateral neck and left axillary adenopathy (presenting symptom). PET study confirms the MR-findings in the

neck (c, d). (e) Periaortic lymph nodes (slim arrows) and lymphadenopathy in the liver hilus (thick arrow) also seen on abdominal CT (not shown), indicating Stage 3 disease HL. Note: Normal activity in the heart, right ureter, and bladder

Neck Lymphoma

e

Fig. 1 (continued)

assessment of lymphomas (Cheson et  al. 2007; Meignan et  al. 2009). The use of PET-CT is strongly recommended before treatment for patients with FDG-avid lymphomas (e.g., diffuse large B cell lymphomas and Hodgkin’s lymphoma) to better delineate and stage the disease (Fig. 1). Recent improvements in MRI technology have resulted in the availability of sufficiently fast and diagnostic sequences to perform “whole-­ body magnetic resonance imaging” (WB-MRI). This MR technique lacks the use of ionizing radiation which is an obvious advantage when compared to CT and PET-CT.  Especially when combined with recently developed functional WB-MRI techniques such as diffusion-weighted

447

imaging (DWI) (Kwee et al. 2008). Earlier data showed that initial staging of malignant lymphoma using WB-MRI (without DWI and with DWI) produced equal results to staging using CT (Fig. 2) in the majority of patients (Kwee et al. 2009). Especially, in young patients with lymphoma it is important to minimize radiation exposure. Recent reports show that WB-MRI could potentially serve as a good radiation-free alternative to FDG-PET/CT (Littooij et al. 2014; Petralia and Padhani 2018). However, worldwide, CT remains the commonly used “workhorse” imaging modality for staging malignant lymphoma because of its widespread availability and relatively low cost (Kwee et al. 2008). In MR imaging of the head and neck, diffusion-­ weighted imaging (DWI) has evolved as an important functional technique with a variety of potential applications. DWI can be easily incorporated in diagnostic MR protocols since DWI sequences typically only take a few minutes to obtain. In addition, DWI can be performed on standard 1.5and 3-T clinical MR systems (Thoeny et al. 2012). In (head and neck) lymphomas an interesting application of DWI is discrimination between squamous cell carcinoma (SCC) and lymphoma based on differences in ADC values. ADCs reported for both nodal and extranodal locations were significantly lower in lymphoma than in SCC (King et al. 2007; Sumi et al. 2007). This difference can be explained by the greater cellularity of lymphomas which leads to increased diffusion restriction and hence a lower ADC (Thoeny et al. 2012). Abdel Razek et al. determined ADC values in patients with (head and neck) non-Hodgkin lymphoma (NHL). The mean ADC value was lower in patients with NHL compared to patients with HL (Abdel Razek et al. 2006).

4.3  Staging Designed for Hodgkin’s lymphoma the Ann-­ Arbor system is now used for staging all ­lymphomas (Table  2). The system was slightly modified after the Cotswold Convention (Lister et al. 1989; Smithers 1971).

F. A. Pameijer and R. L. M. Haas

448

a

b

c

d

Fig. 2 Pediatric patient with Hodgkin’s lymphoma. Initial staging and response assessment 3 months post chemotherapy with PET-CT. (a) Fused coronal PET-CT image reveals marked FDG uptake in bilateral neck nodes, right axillary nodes (presenting symptom), and bilateral mediastinal nodes (arrows), indicating Stage 2 disease. Note: Normal FDG uptake in the heart and right kidney. (b) Coronal T2 short tau inversion recovery (STIR) image

showing good correlation with PET-CT (arrows). Note: Left neck and mediastinal nodes not shown on this section. (c) Post chemotherapy fused coronal PET-CT image shows complete nodal regression and absence of abnormal FDG uptake. (d) Coronal STIR WB-MRI image with good correlation with PET-CT demonstrating complete nodal regression and no areas of abnormal increased signal intensity indicative of complete response

Neck Lymphoma

449

Table 3  Lymph node imaging features on US, CT, and MRI for Hodgkin’s and non-Hodgkin’s lymphoma. For comparison, imaging features for squamous cell carcinoma nodes have been added Imaging features Hodgkin’s lymphoma US Well-defined enlarged round nodes with homogeneous appearance CT Lobulated round nodal masses with variable enhancement. Central necrosis may be present. Calcification (pre-­ treatment) uncommon MRI T1WI: enlarged round iso- to hypointense nodes. T2WI: nodes are hyperintense to muscle. T1 C+: variable: see CT DWI: low ADC value

Non-Hodgkin’s lymphoma Diffuse homogeneous decreased echogenicity characteristic

Squamous cell carcinoma Round nodes with loss of hilar echogenicity

Multiple bilateral ovoid masses in multiple nodal chains. Variable enhancement ranging from isodense to muscle to strong enhancement. Central necrosis may be present. Enhancement pattern may vary within same patient T1WI: enlarged nodes isointense to muscle. T2WI: nodes are hyperintense to muscle. T1 C+: variable: see CT. DWI (marked) low ADC value

Diffuse or rim-enhancement of nodes. Central necrosis is typical

T1WI: isointense to muscle; necrosis seen as hypointense focus. T2WI: hyperintense; necrosis shows focal marked hyperintensity. T1 C+: inhomogeneous

C+ contrast enhancement. (Adapted from Harnsberger (2004))

5  Treatment Management of lymphomas is very diverse and depends fully on the specific lymphoma subtype, the stage in which the patient is diagnosed, comorbidity, the performance status, and biological age of the patient. As a concept, lymphoma patients can be treated by chemotherapy, radiotherapy, and immunotherapy, and usually in a combined modality approach. Some frequently applied regimens are the following:

•

• Chemotherapy, single agent; Chlorambucil (oral), Cyclophosphamide (oral), Fludarabin (oral and iv.): in case of indolent lymphomas, predominantly follicular lymphoma. • Chemotherapy, multi-agent; CVP (=COP), Chlorambucil  +  Vincristine (=Oncovin®) + prednisone in case of indolent lymphomas, R-CHOP (Rituxan®; CHOP = COP + Adriamycin), DHAP (Dexamethasone + high dose Ara-C + Cisplatin), VIM (VP-16 (=Etop oside)  +  Ifosphamide  +  Mesna) in case of aggressive lymphomas, predominantly DLCL, BEAM (BCNU + Etoposide + Ara-C + Melp halan) as a conditioning regimen before stem cell transplantations for aggressive lymphomas, ABVD (Adriamycin + Bleomycin + Vin blastin + Dacarbazine), MOPP (Mitoxin + On

•

•

•

•

covin  +  Procarbazin  +  Prednisone), MOPPABV for Hodgkin’s lymphoma. IF-RT (=involved field radiotherapy); all lymphomas; irradiation on the affected site only, no prophylactic irradiation to adjacent areas. IN-RT (=involved node radiotherapy); irradiation to the affected individual lymph nodes only, refraining from irradiation of the entire nodal basin. EF-RT (=extended field radiotherapy); all lymphomas; irradiation on the affected site plus prophylactic irradiation to adjacent areas. [S]TNI  =  [sub]total nodal irradiation; Hodgkin’s lymphoma; this regimen of irradiation is now considered obsolete. Combined modality regimens; 3–4 × CHOP + IF-RT: aggressive lymphomas, predominantly DLCL in stage I or II, 3–4 × ABVD + IF-RT: Hodgkin’s lymphoma, stage I or II.

6  Response Assessment In principle, all imaging studies performed in the initial staging process will be repeated after completion of treatment. Imaging during treatment may also be indicated to decide on further management.

F. A. Pameijer and R. L. M. Haas

450

In hematology, response used to be assessed by the “Cheson-criteria” (Cheson 2008), but these criteria have been refined in the so-called Lugano Classification (Cheson et al. 2014). This applies also to response evaluation in head and neck lymphoma. This is in contrast to response assessment in solid tumors, such as carcinomas and sarcomas, for which the RECIST criteria are used (Therasse et al. 2000). FDG PET-CT has now become part of the routine diagnostic workup of the majority of ­oncologic patients, including patients with lymphoma. A major advantage of FDG PET-CT, over conventional imaging techniques such as CT and MRI, is its ability to distinguish between viable tumor and residual mass lesions that contain only necrosis or fibrosis. The widespread use of PET in lymphoma patients warranted a reassessment of the previously established response criteria. In these “Deauville response criteria for lymphomas” recommendations for the use of PET or PET-CT both for pretreatment imaging and for posttreatment response evaluation are included (Meignan et al. 2009) (Fig. 2).

7  Nodal Disease Per definition, this is lymphatic malignancy occurring in preexisting nodal chains (in the head and neck). In hematology, especially in Hodgkin Lymphoma, lymph nodes are typically involved

a

by contiguous spread. This is in contrast to solid tumors (i.e., squamous cell carcinoma of the head and neck) where specific lymph node groups are at a higher risk for metastatic involvement based on the preferential lymphatic drainage from the primary tumor site.

7.1  The Common Sites For all imaging modalities depicting nodal disease, including US, CT, and MRI, the cervical lymph nodes are described in terms of levels (Som et al. 2000; UICC 2016). In head and neck nodal disease, levels I–V are most frequently involved (Fig. 3). In clinically suspected neck adenopathy, US may be helpful because it can be used to guide fine needle aspiration cytology (FNAC). An experienced cytopathologist can reliably diagnose lymphoid lesions from FNAC-smears, especially in the differential diagnosis from carcinomas. CT and MRI are better equipped for local tumor mapping and staging. For further subtyping of the lymphoma, surgical removal of an entire lymph node is subsequently required. For practical purposes, nodal involvement with HL cannot be exclusively differentiated from nodes with NHL whatever imaging modality is used. In the head and neck, both entities typically present with multiple, uni- or bilateral non-necrotic enlarged lymph nodes involving

b

Fig. 3 (a, b) Nodal B-NHL. Axial contrast-enhanced CT shows bilateral enlarged level I–V cervical nodes without central necrosis

Neck Lymphoma

usual and unusual nodal chains. However, certain imaging features are more characteristic for HL than for NHL. General Features in Favor of Hodgkin lymphoma Nodes • HL presents in lymph nodes in about 98% of cases. When multiple nodes are present, the involved nodal groups are contiguous in about 90% of cases. This suggests a unifocal origin of the disease and subsequent dissemination through lymphatic pathways. • Levels III and IV most frequently involved. • Combination with mediastinal lymphadenopathy is frequent. • Involvement of extranodal sites such as Waldeyer’s ring is uncommon.

451

(Fig.  1). Other nodal sites, not included in the UICC nodal classification, include occipital, facial, buccal, periparotid, and intraparotid lymph nodes. All these nodes may be sites of origin of lymphoma (Vega et  al. 2005; Aiken and Glastonbury 2008; Watal et al. 2018). There are no differences in imaging appearance between these nodal groups and the level I–V nodes discussed above (Table 3).

8  Extranodal Disease

Per definition, this is lymphatic malignancy occurring outside preexisting nodal chains. Extranodal lymphoma can be separated in extranodal lymphatic and extranodal extralymphatic disease (Hermans 2004). Extranodal areas predisposed to develop lymphoma are sites that are normally rich in lymphoid tissue such as General Features in Favor of Non-­Hodgkin Waldeyer’s ring (extranodal lymphatic disease). lymphoma Nodes Extranodal extralymphatic sites include the orbit, parotid gland, nasal cavity, paranasal • NHL presents in a generalized fashion (includ- sinuses, and the thyroid gland. Although anaing bone marrow infiltration) much more tomically in close proximity, lymphomas arising often than HL. in these sites have distinct clinical characteris• Levels II–IV often involved. tics. Factors that appear to influence the disease • NHL often involves extranodal tissue, both pattern include concurrent conditions, such as Waldeyer’s ring (extranodal lymphatic) and Sjögren’s syndrome, and geographic factors parextranodal extralymphatic (orbit, salivary ticularly concerning sinonasal lymphomas (see glands, sinonasal, etc.) tissue. also Sect. 8.4). The clinical presentation of patients with Specific lymph node imaging features for US, extranodal NHL often mimic those of patients CT, and MRI are shown in Table 3. with squamous cell carcinoma (SCC) arising at Note: A minority of patients has enlarged the same site. Approximately 10% of patients lymph nodes with central necrosis on cross-­ with NHL present with extranodal primary sectional imaging. This imaging pattern is much lesions in the head and neck. more frequently seen in patients with head and The treatment and prognosis of patients with neck squamous cell cancer. Central necrosis in head and neck lymphoma depends on the patholymphoma patients usually indicates a higher logical subtype of disease and extent of involvemalignancy grade. ment at time of presentation. The most common lymphoma is the diffuse large B cell lymphoma (65%) in an early stage at presentation. In the parotid gland, however, the indolent histologies 7.2  The Uncommon Sites (marginal zone lymphoma and follicular lymManifestations of nodal disease may also be seen phoma) prevail. Predominantly patients between in more uncommon sites including pre-tracheal 50 and 60 years of age are affected. The male-to-­ nodes (level VI) and retropharyngeal nodes female ratio is 1.6:1, with the exception of lym-

F. A. Pameijer and R. L. M. Haas

452

phomas of the salivary glands, orbit, and thyroid, which occur equally or more frequently in women. The imaging findings of extranodal head and neck lymphomas are essentially the same as those found in SCC. In 80% of patients with primary extranodal NHL in the head and neck, additional nodal involvement is present. If a primary lesion is associated with large, homogeneously enhancing lymph nodes without central necrosis, the possibility of lymphoma may be suggested. However, this combination of imaging findings may also be encountered in SCC. Therefore, the diagnosis of lymphoma remains based on histologic tissue examination.

8.1  W  aldeyer’s Ring and the Upper Aerodigestive Tract Waldeyer’s ring comprises the lymphoid tissues located superiorly in the nasopharynx, laterally in both tonsillar fossae and inferiorly in the base of tongue. More than half of extranodal head and neck lymphomas occur in Waldeyer’s ring, the tonsil being the most prevalent subsite (40–50% of cases). All subtypes of NHL can be found in Waldeyer’s ring, but deposits of Hodgkin’s lymphoma are rare (Yuen and Jacobs 1999; Vega et al. 2005).

a

b

Fig. 4  Nasopharyngeal NHL. Adult patient with bilateral neck swelling and conductive hearing loss. (a) Axial T2-weighted image at the level of the nasopharynx shows a bulky mass (M) without parapharyngeal and/or clival invasion. Note: Bilateral mastoid effusion (arrows) explaining the conductive hearing loss. (b) Coronal contrast-­enhanced T1-weighted image with fat saturation

8.1.1  Nasopharynx Patients with nasopharyngeal NHL may present with intermittent hearing loss secondary to unilateral or bilateral Eustachian tube obstruction (Fig. 4) or may have persistent epistaxis or nasal obstruction. Imaging of nasopharyngeal NHL shows two morphological types: a circumscribed, bulky, mucosa-based mass that has not invaded the deep structures around the pharynx (Fig. 4), and a mass with a much more infiltrative spread pattern. The latter type of primary nasopharyngeal lymphoma may invade the skull base and spread along nerves in a manner similar to that of SCC (King et  al. 2003; Aiken and Glastonbury 2008). Both diseases are commonly associated with cervical adenopathy (Fig. 4). In the diagnostic phase of a patient with a nasopharyngeal mass, diffusion-weighted MR imaging may be used as an adjunct in differentiation between nasopharyngeal carcinoma (NPC) and lymphoma. The ADC values in lymphoma were found to be significantly lower than in NPC (Fong et al. 2010) (Fig. 4). In children and adolescents nasopharyngeal lymphoid tissue is prominent. These so-called “adenoids,” or nasopharyngeal tonsils are located in the roof of the nasopharynx. Such tissue should not be mistaken as an abnormal mass. Normal lymphoid tissues show uniform enhancement after contrast medium injection and, on MRI,

c

of the neck shows bilateral lymphadenopathy (arrows). Note: Non-necrotic enhancing node on the left side and necrotic non-enhancing lymphadenopathy on the right side. (c) b600 axial diffusion-weighted image showing diffuse increased signal intensity of the nasopharyngeal mass indicative of (marked) diffusion restriction

Neck Lymphoma

display a typical high signal intensity without invasion of adjacent structures.

8.1.2  Tonsillar Fossa Patients with NHL of the tonsil usually have a unilateral sore throat or obstructive symptoms. Clinically, especially in younger patients, these lesions are difficult to differentiate from reactive hypertrophy. Frequently the diagnosis of NHL is delayed while patients are treated with prolonged courses of antibiotics or even with incision for a suspected tonsillar abscess. Hermans et al. (1994) described two patterns of tonsillar involvement: (a) Involvement of only the tonsillar fossa. Imaging in these cases shows a unilateral or bilateral enlarged tonsil (Fig.  5), without infiltration of surrounding tissues. (b) Tonsillar involvement with extension to the pharyngeal wall. In this subgroup, two types of extension have been described. An infiltrating growth pattern, in which there is extension in the parapharyngeal fat plane and into the extrapharyngeal deep core tissues (masticator space, parotid space, etc.); this pattern mimics the growth pattern of carcinomas. The other subtype in this group consists

Fig. 5 Tonsillar NHL.  Axial T2-weighted sequence (TIRM) demonstrates bilateral high signal intensity (SI) soft-tissue masses in the tonsil. Note that there is no invasion of the surrounding deep-tissue planes. There is an associated enlarged level II lymph node on the right with the same high SI appearance

453

of an exophytic pattern of growth in which the tumor stayed inside the pharyngeal walls but extended outside of the tonsillar margins expanding into the oral cavity (Fig.  6). Complete circumferential thickening of the pharyngeal wall without deep infiltration has been reported as quite specific for lymphoma (Hermans et al. 1994). The signal intensity on T2-weighted, T1-weighted, and T1-weighted contrast-­ enhanced MRI images is as a rule homogeneous and similar to that of normal tonsils. Large tumors can be mildly heterogeneous showing small foci of necrosis. Lymphadenopathy in the ipsilateral upper internal jugular chain can be seen and in those cases, the nodes will be of similar signal intensity to the primary tumor (Fig. 5). In 20% of tonsillar NHL gastrointestinal involvement can be diagnosed simultaneously (Hanna et al. 1997).

8.1.3  Base of Tongue Patients with lymphoma in the base of tongue may have obstructive symptoms, sore throat and occasionally dysphagia or a change in voice. Clinically, these lesions are often submucosal, bulging and soft on palpation. As mentioned, the high concentration of lymphoid tissue in the base of tongue (tonsilla lingualis) predisposes this site to develop lymphoma. The typical imaging appearance on CT (Fig. 7) or MRI is a bulky, exophytic mass centered in the tongue base. Enhancement pattern and signal intensities are identical to tonsillar NHL. 8.1.4  Larynx Laryngeal involvement with NHL is very rare. Laryngeal lymphomas tend to have a large submucosal component most frequently centered in the supraglottis (Aiken and Glastonbury 2008). The imaging characteristics are nonspecific and comparable to SCC.  Multiple lesions strongly suggest the diagnosis of NHL (Hermans et  al. 1994) (Fig.  8). In addition, a laryngeal tumor with a large supraglottic submucosal component (Fig. 9) should alert the radiologist to the possibility of NHL (King et al. 2004).

F. A. Pameijer and R. L. M. Haas

454

a

Fig. 6  Tonsillar NHL. Contrast-enhanced multislice CT images. (a) Axial contrast-enhanced CT shows a soft-­ tissue mass situated in the right tonsil (arrows). There is anterior extension (within the pharyngeal wall) along the glossotonsillar sulcus with slight displacement of the right tongue base. A grossly enlarged lymph node is present in

a

Fig. 7  Base of tongue base NHL. A female patient with dysphagia. Contrast-enhanced multislice CT images. (a) Axial contrast-enhanced CT shows a large homogeneously enhancing mass centered in the left base of

b

level II (arrowhead). (b) Coronal reformatting shows the medial extension of the mass (arrows) with involvement of the soft palate. The level II lymph node contains a large area of central necrosis (arrowhead). (Courtesy R. Hermans, MD, PhD, Leuven, Belgium)

b

tongue. Note that there is no invasion of the surrounding deep-tissue planes. (b) Sagittal reformatting demonstrating the exophytic nature of the mass

Neck Lymphoma

8.2  Orbit There are a variety of tumors and tumor-like lesions in the orbit, both benign and malignant. In an unselected series of 1264 orbital lesions, 810 (64%) were benign and 454 (36%) were malignant. The percentage of malignant lesions was

Fig. 8 Laryngeal NHL.  Axial contrast-enhanced CT image at the lower border of the hyoid bone shows bilateral large submucosal masses. (Courtesy I.M. Schmalfuss, MD, Gainesville, FL)

a

Fig. 9  Laryngeal NHL. On endoscopy bulky supraglottic mass; no mucosal abnormalities. Contrast-enhanced CT images. (a) Axial and (b) coronal section showing a large

455

20% in children, 27% in young adults and middle-­aged patients, and 58% in older patents (age range, 60–92  years). Lymphoma was the most common malignancy in older patients, representing 10% of cases (Shields et  al. 2004). Lymphoid tumors of the orbit in children are extremely rare. Besides malignant lymphoma, a spectrum of less malignant lymphoid orbital tumors exists ranging from the benign pseudolymphoma or pseudotumor (also known as idiopathic orbital inflammatory syndrome; IOIS) to the reactive and atypical lymphoid hyperplasia. These two latter entities mimic orbital lymphoma, both from an imaging and from a histological point of view. The incidence of IOIS and lymphoid tumors seems to be equal (Yan et al. 2004). Any orbital tissue or combination of tissues may be involved in malignant lymphoma, as well as the less malignant lymphoid orbital tumors. Anatomically, true lymphoid tissue is found in the subconjunctival and lacrimal glands, predisposing these sites for development of lymphoma. Orbital lymphomas can be subdivided into three groups: conjunctival, lacrimal, and intra-orbital (true intra-ocular lymphomas are very rare). Orbital lymphomas are mostly indolent, low-­ grade lesions. Marginal zone lymphoma (MZL),

b

right-sided submucosal supraglottic Nasogastric tube (arrow in a)

mass.

Note:

F. A. Pameijer and R. L. M. Haas

456

synonym MALT lymphoma, followed by follicular lymphoma as the most prevalent diagnosis (Lee et  al. 2005). If MZL is diagnosed, special attention should be focused to the other MZL-­ prevalent sites like the salivary glands, Waldeyer’s ring, and the stomach. Treatment of orbital lymphomas is moderate dose irradiation (25 Gy for indolent histologies, such as MZL and 36 Gy for aggressive subtypes such as DLBCL) to the entire orbit with over 90% of local control (Yahalom et al. 2015). For imaging of the orbit, CT and MRI, or the combination of both, can be used (Priego et  al. 2012). When using MRI, diffusion-weighted imaging (DWI) can be used to differentiate orbital lymphomas from other pathologies because these display the lowest ADC among orbital mass lesions (Politi et al. 2010).

8.2.1  Conjunctiva Patients with lymphoma arising in the conjunctiva complain of local irritation, itching, ptosis, or the sensation of a mass. Imaging shows unilateral or bilateral symmetrical swelling of the conjunctival tissues. These are smooth, sharply marginated ovoid lesion that shows moderate to strong enhancement (Fig.  10). In case of conjunctival

a

Fig. 10  Conjunctival NHL. T1-weighted MR images. (a) Axial image at the cranial margin of the orbit shows an ovoid, sharply marginated, lesion located anterior to the

involvement, differentiation has to be made from preseptal cellulitis, which is usually unilateral, and of sinonasal origin; preseptal IOIS is rare (Hermans et al. 1994).

8.2.2  Intra-orbital Lymphoma Painless proptosis, diplopia, and visual disturbance are the common presenting symptoms of patients with intra-orbital lymphoma. On imaging, orbital lymphomas are homogeneous usually sharply marginated intra- or para-conal masses that occur most often in the anterior portion of the orbit, the retrobulbar area (Fig. 11), or in the superior orbital compartment. Mild to moderate enhancement is usually present. Based on the imaging findings alone, there is a differential diagnosis for these types of lesions including malignant lymphoma, pseudotumor (IOIS), reactive hyperplasia, lacrimal gland tumors, optic nerve tumors, and Graves’ orbitopathy. This differential can be narrowed when the imaging features are interpreted in conjunction with the clinical signs and symptoms. A somewhat characteristic feature of orbital lymphoid tumors is the tendency to mold themselves around intra-­ orbital structures without (bony) destruction (Tailor et al. 2013) (Fig. 11).

b

right globe. (b) Post-contrast sagittal image showing mild homogenous enhancement of the mass located in the upper eyelid

Neck Lymphoma

a

Fig. 11 Intra-orbital NHL.  T1-weighted axial MR images. (a) Non-enhanced image at the level of the optic nerve shows diffuse infiltration of the intraconal space of the left orbit. (b) Post-contrast fat-suppressed image

457

b

showing moderate homogeneous enhancement of the intraconal mass. Note that the lesion is “molding” around the optic nerve

using MRI, the addition of diffusion-weighted imaging (DWI) is recommended because lacrimal gland (and other orbital) lymphomas typically display very low ADC values (Politi et al. 2010) (Fig. 13).

8.3  Salivary Glands Fig. 12  Lacrimal gland NHL.  Axial contrast-enhanced CT image through the superior parts of the orbits shows symmetric enlargement of both lacrimal glands (arrows). There is mild medial and forward displacement of the globes. (Courtesy R.  Hermans, MD, PhD, Leuven, Belgium)

8.2.3  Lacrimal Gland Orbital lymphoma has a predilection for the lacrimal gland, which is involved in about 40% of cases (Tailor et  al. 2013). Lacrimal gland lymphoma may present with epiphora or as a painless mass noticed for cosmetic reasons. If lymphoma is restricted to the gland itself, the lesion is characteristically located in the supero-lateral orbital compartment. Cross-sectional imaging shows smooth, unilateral or bilateral (Fig. 12), enlargement of the lacrimal gland. If the gland is grossly enlarged, medial and forward displacement of the globe is the clue to the primary location. When

Primary lymphoma of the salivary glands is an uncommon tumor. Primary MZL (marginal zone lymphoma) may occur in Sjögren’s syndrome, as a consequence of continuous antigenic stimulation. The parotid gland is affected most often (80% of reported cases) followed by the submandibular gland (20%). There are only isolated reports of primary lymphoma in minor salivary glands; most commonly involved sites include the palate and gingiva.

8.3.1  Parotid Gland Lymphoma of the parotid gland may arise within the parotid parenchyma (primary) or within the intraparotid lymph nodes (secondary). Primary lymphoma of the parotid gland is rare, accounting for less than 5% of parotid tumors. These lymphomas are classified as MALT (mucosa-associated lymphoid tissue) lymphomas (see also Table  1). These low-grade lymphomas may involve any portion of the gastrointestinal

F. A. Pameijer and R. L. M. Haas

458

a

b

c

d

Fig. 13  Lacrimal gland NHL (marginal zone B cell lymphoma). (a) Axial T1-weighted image shows smooth enlargement of the left lacrimal gland (arrow). (b) Post-­ contrast fat-suppressed image reveals homogeneous enhancement of the gland (arrow). (c) b1000 diffusion-­

weighted image showing increased signal intensity of the left lacrimal gland (arrow). (d) ADC map at the same level displays decreased ADC indicative of (marked) diffusion restriction (arrow)

tract where lymphoid tissue is part of the mucosal defense system. Lymphoid tissue is normally present in the parotid gland and absent in the submandibular and sublingual gland. When MALT lymphoma occurs in the salivary gland, as in other extranodal sites such as the stomach, it is usually an indolent neoplasm that tends to remain localized for long periods of time (Abbondanzo 2001; Palacios et  al. 2004; Tonami et  al. 2002, 2003). Patients with primary lymphoma of the parotid gland typically present with a painless parotidregion mass or with parotitis with progressive enlargement of the gland. After systemic workup most patients are found to have early stage localized disease, i.e., Stage IIE (see also Table 2). Patients with Sjögren’s syndrome have an increased risk of developing primary parotid lymphoma. Sjögren’s syndrome, or sicca syndrome, is an autoimmune disease characterized by keratoconjunctivitis sicca, dryness of mucous

membranes, telangiectasias of the face, and bilateral parotid enlargement (Fig. 14). The syndrome is often associated with rheumatoid arthritis and Raynaud’s phenomenon. The risk of lymphoma in these patients has been estimated to be approximately 44 times the incidence expected in a generally healthy population (Yuen and Jacobs 1999). It may affect any area of the head and neck but may have its first manifestations in the parotid gland. Often these patients are scanned to survey for the possibility of lymphoma. Lymphoma may also involve the salivary glands (usually the parotids) secondary to systemic lymphoma. For imaging of suspected salivary gland tumors, US, CT, as well as MRI can be used. Because of their superficial position, the parotid, the submandibular, and the sublingual glands can be imaged with high-resolution US transducers (Gritzmann et al. 2003). This is especially helpful

Neck Lymphoma

a

459

b

Fig. 14  Primary MALT lymphoma of the parotid gland. A female patient with Sjögren’s syndrome with progressive parotid enlargement. T2-weighted (TIRM) MR images. (a) Coronal image confirms bilateral parotid swelling. This is caused by innumerable small cysts indicative of chronic sialadenitis with cystic enlargement of

in the diagnostic phase because US can be used to perform fine needle aspiration cytology. MRI and CT are better equipped for local tumor mapping and staging purposes. Primary lymphoma is characterized by, partial or total, replacement of the parotid gland by an infiltrating soft tissue mass (Fig.  14). Multiple solid, intraparotid masses are suggestive of secondary lymphoma (Fig. 15). These enlarged intraparotid lymph nodes are usually sharply marginated round to ovoid lesions with homogeneous attenuation. However, nodal necrosis can occur. On MRI, primary and secondary lymphomas are characterized by homogeneous intermediate signal intensity (SI) on all imaging sequences. There is also a tendency to “fade” into the SI of the parotid gland on T2-weighted and contrast-­ enhanced, fat-suppressed T1-weighted MR images. Low ADC values of the mass on diffusion-­ weighted imaging (DWI) may help to suggest the diagnosis of lymphoma (Kato et al. 2012). In the setting of multiple intraparotid and/or periparotid lesions, the age of the patient is important in establishing a differential diagnosis. In younger patients, acute viral or bacterial infections should be considered. Acute otitis

terminal salivary ducts. Note: Multiple non-enlarged level IV nodes bilaterally. (b) Axial plain T1-weighted image shows an additional infiltrative mass located in the superficial lobe of the right parotid gland invading the masseter muscle. Biopsy of this mass revealed MALT lymphoma

Fig. 15  Secondary parotid lymphoma. Axial contrast-­ enhanced CT image shows multiple sharply marginated masses within and around both parotid glands representing enlarged intra- and periparotid lymph nodes. (Courtesy R. Hermans, MD, PhD, Leuven, Belgium)

media can present with reactive lymphadenopathy of periparotid lymph nodes. Multiple intraparotid cystic lesions should alert the radiologist to the possibility of HIV infection. These cystic lesions (also called benign lymphoepithelial cysts; BLEC) can serve as the initial clinical manifestation of HIV seropositivity (Sekikawa and

F. A. Pameijer and R. L. M. Haas

460

Hongo 2017). Most commonly, with HIV infection there is also a diffuse cervical adenopathy. The differential diagnosis for elderly patients with multiple intraparotid masses should include lymphoma. Other lesions to be considered include Warthin’s tumors and intraparotid metastases from SCC of the skin for unilateral involvement, and Warthin’s tumors and melanoma for bilateral involvement.

8.4  Sinonasal Cavities The incidence of nasal and paranasal lymphoma in the United States and Europe is low. However, lymphomas of the paranasal sinuses and nasal cavity are relatively prevalent in certain parts of Central and South America and East Asia, including Korea, and seem to be associated with Epstein–Barr virus infections (Koom et al. 2004). Lymphomas arising in the sinonasal cavities are NHL and frequently are observed in patients who have disseminated lymphoma or AIDS. In the nasal cavity the extranodal natural killer (NK)/T cell lymphoma, nasal type (see also Sect. 3.2; entity 2(e)) is the most prevalent type (Yahalom et al. 2015). Sinonasal lymphoma most frequently occurs in the nasal cavity and maxillary sinus. Less a

often, lymphoma is found in the ethmoid sinuses, and only rarely it is found in the sphenoid and frontal sinuses. Patients with maxillary sinus lymphoma present with facial fullness, symptoms of sinusitis, or odontogenic pain. Initial symptoms of nasal cavity lymphoma include nasal obstruction and/or epistaxis. Proptosis and epiphora, i.e., orbital and lacrimal gland extension indicate more widespread disease. Rarely, malignant lymphoma can present as a mass in the canine fossa. If isolated, this most likely represents lymphoma arising in the infraorbital lymph node group, i.e., nodal disease (Tart et al. 1993). On CT and MRI imaging studies, lymphomas of the sinonasal cavities must be differentiated from the much more common entities of sinusitis, polyposis, as well as granulomatous processes such as Wegener’s granulomatosis, and benign and malignant neoplasms (Aiken and Glastonbury 2008). The imaging appearance is nonspecific. Lymphomas in the sinonasal cavities tend to be bulky soft tissue masses with moderate enhancement. The growth pattern may be expansile. However, in more advanced cases, aggressive bone erosion may be observed. Based on imaging alone, this more aggressive appearance of lymphoma cannot be differentiated from carcinoma (Fig. 16).

b

Fig. 16  Sinonasal NHL. Contrast-enhanced CT images. (a) Axial image shows a soft-tissue mass with homogenous enhancement centered in the left maxillary sinus. Extensive bone destruction. The mass is extending into the nasal cavity and the soft tissues of the cheek. Subtle

bony erosion of the postero-lateral sinus wall with beginning infiltration of the infratemporal fossa. (b) Coronal image (bone window setting) showing the extensive bony destruction. Note: Retention cyst in the right maxillary antrum

Neck Lymphoma

461

8.5  Thyroid Lymphoma of the thyroid gland is an uncommon disease occurring primarily in older women. It may arise primarily from the thyroid or involve the gland as part of a systemic disease. There is a strong association between thyroid lymphoma and Hashimoto’s thyroiditis. Hashimoto’s thyroiditis is an autoimmune disease characterized by circulating antithyroglobulin resulting in follicular atrophy, fibrosis, and enlargement and firmness of the gland. Surveillance of these patients by yearly physical examination and US will facilitate discovery of lymphoma at an early stage. The most common histologies are the diffuse large B cell lymphomas (up to 70% of cases) and MALT lymphomas (in 6–27% of cases). Most patients have a short history of an enlarging thyroid or a neck mass causing tracheal compression or swallowing problems. Treatment for MALT lymphomas includes radiation therapy alone. In case of diffuse large B cell lymphoma, patients will be treated by a combination of radiation with chemotherapy (Yahalom et al. 2015). The imaging approach is based on prior clinical evaluation. Small lesions are ideally assessed with US, which is capable of discriminating between solid and cystic nodules. US-guided fine needle aspiration cytology/biopsy provides tissue for pathologic examination of thyroid nodules. On US, thyroid lymphoma may present as a focal mass or diffuse replacement of the thyroid gland. The mass is typically diffusely hypoechoic. CT and MRI are indicated for larger tumors, especially when extending outside the gland ­ (Weber et al. 2000). On CT, thyroid lymphoma may be seen as a focal mass, but more commonly as diffuse enlargement of the gland (Kim et al. 2002). On non-contrast CT the tumor demonstrates low attenuation. Following contrast administration, these tumors moderately enhance (Fig.  17). Associated cervical adenopathy may occur. Calcification and necrosis are unusual (Aiken and Glastonbury 2008). These two latter features are much more frequently seen in thyroid goiter.

Fig. 17  Thyroid lymphoma. Axial contrast-enhanced CT image shows diffuse enlargement of the thyroid gland by a relatively homogenous mass. (Courtesy I.M. Schmalfuss, MD, Gainesville, FL)

Fig. 18 Thyroid lymphoma. Axial contrast-enhanced MR image shows diffuse enlargement of the thyroid gland with homogenous enhancement. Extensive retropharyngeal, as well as intralaryngeal extension. Note: Enlarged level III lymph node on the right with similar enhancement as the thyroid lymphoma. (Courtesy R.  Hermans, MD, PhD, Leuven, Belgium)

The MRI appearance of thyroid lymphoma is usually isointense on T1-weighted images with moderate homogeneous enhancement following contrast administration (Fig. 18). On T2-weighted sequences, the signal intensity is homogeneously bright (Takashima et al. 1995). Frequently occurring vascular invasion, associated lymphadenopathy and extension into the retropharyngeal area

F. A. Pameijer and R. L. M. Haas

462

and mediastinum are depicted adequately by both CT and MRI (Fig. 18).

8.6  Bone Bony disease in the setting of head and neck lymphoma is uncommon. Secondary infiltration and/ or destruction (mimicking SCC) may occur (Fig.  16). However, there are some primary hematological bone disease entities in the head and neck.

8.6.1  Primary Lymphoma of Bone The most common lymphoma affecting bone is the diffuse large B cell lymphoma. Primary lymphoma of bone occurs most frequently in the mandible followed by the maxilla. In the mandible, most often there are ill-defined lytic destructive areas of variable size. These imaging features are nonspecific and the differential diagnosis includes other primary neoplasm of bone, such as osteosarcoma and Ewing’s sarcoma. The jaw is a common site of presentation for the African type of Burkitt’s lymphoma. This B cell lymphoma occurs most frequently in children and young adults with 50% of cases arising in the maxilla or mandible (Fig. 19). This disease is found especially in Africa and other underdeveloped countries, much less frequently in the United States and Europe. Association with Epstein–Barr Virus is postulated (Muwakkit et al. 2004; Rao et al. 2000; Weber et al. 2003). 8.6.2  M  ultiple Myeloma (Kahlers’ Disease) Multiple myeloma (MM) is characterized by a malignant proliferation of plasma cells with monoclonal immunoglobulin or immunoglobulin fragments in the patient’s urine. This proliferation of neoplastic cells is associated with bone destruction and involves the bone marrow of the axial skeleton. Although single lesions occasionally occur (solitary plasmacytoma of bone), MM typically shows diffuse involvement of multiple bones. The typical radiographic appearance is that of “punched-out” round to ovoid, regular radiolucencies in the skull and long bones with

Fig. 19  Primary Burkitt’s lymphoma of the mandible. Axial T1-weighted non-enhanced MR image shows replacement of the normal fatty marrow by a low signal intensity mass in the median and right part of the mandible, with expansion of the medullary cavity

no circumferential bone sclerosis (Faria et  al. 2018). In the mandible, these lesions show a ­predilection for the angle, ramus, and molar tooth regions. Solitary plasmacytomas of bone usually occur in the vertebra and skull.

8.6.3  Extramedullary Plasmacytoma Extramedullary plasmacytoma (EMP) is a rare soft-tissue malignancy composed of plasma cells. Eighty percent of EMPs arise in the head and neck, most often occurring in the nasal cavity followed by the paranasal sinuses (Mendenhall et al. 2003). On CT, EMPs of the sinonasal cavities are smooth, homogeneous enhancing polypoid masses that remodel surrounding bone (Fig. 20). On MR imaging, they have intermediate signal intensity on all imaging sequences. EMPs show moderate to marked heterogeneous enhancement (Ching et al. 2002).

8.7  Skin NHL in the head and neck may also be localized in the skin. The cutaneous lymphomas can be distinguished in B and T cell types. Specific for the

Neck Lymphoma

463

imaging patterns can strongly suggest the diagnosis of lymphoma. However, it should be kept in mind that lymphoma is a possible cause in any infiltrative soft tissue mass in the head and neck, irrespective of the location. In the workup of a patient with a head and neck lesion, the radiologist may be the first one to suggest this diagnosis.

References

Fig. 20  Sinonasal plasmacytoma. Axial non-enhanced CT image (bone window setting) shows a smooth lobulated mass centered in the left nasal cavity. Note: Subtle remodelling of the medial wall of the left maxillary sinus

cutaneous T cell lymphomas are the so-called mycosis fungoides and the Sezary syndromes, but they are more likely to be seen on the trunk and extremities than the head and neck (Foss 2004). Crosti’s syndrome should be mentioned. Originally called the reticulohistiocytoma of the dorsum by Crosti in 1951, this disease is nowadays considered a primary cutaneous B cell lymphoma of follicular center cell origin. This localized skin disease on the back of the head and trunk has a very slowly progressive course, with many patients showing no systemic involvement even after prolonged follow-up (Ziemer et  al. 2008). When palpation is felt to be insufficient for local staging, CT or MRI can be used for mapping pre-therapeutic disease extent and document response post-therapy.

9  Conclusion Lymphoma should be approached as a systemic disease that can manifest itself in many forms in the head and neck. Frequently, the imaging findings are nonspecific and tissue sampling remains the mainstay of making the diagnosis. Sometimes, involvement of specific subsites or specific

Abbondanzo SL (2001) Extranodal marginal-zone B-cell lymphoma of the salivary gland. Ann Diagn Pathol 5:246–254 Abdel Razek AAK, Soliman NY, Elkhamary S et  al (2006) Role of diffusion-weighted MR imaging in cervical lymphadenopathy. Eur Radiol 16:1468–1477 Aiken AH, Glastonbury C (2008) Imaging Hodgkin and non-Hodgkin lymphoma in the head and neck. Radiol Clin North Am 46:363–378 Cartwright R, Brincker H, Carli PM et al (1999) The rise in incidence of lymphomas in Europe 1985–1992. Eur J Cancer 35:627–633 Cheson BD (2008) New response criteria for lymphomas in clinical trials. Ann Oncol 19:iv35–iv38 Cheson BD, Fisher RI, Barrington SF et  al (2014) Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol 32:3059–3067 Cheson BD, Pfistner B, Juweid ME et al (2007) Revised response criteria for malignant lymphoma. J Clin Oncol 25:579–586 Ching AS, Khoo JB, Chong VF (2002) CT and MR imaging of solitary extramedullary plasmacytoma of the nasal tract. AJNR Am J Neuroradiol 23:1632–1636 Faria KM, Brandão TB, Silva WG et al (2018) Panoramic and skull imaging may aid in the identification of multiple myeloma lesions. Med Oral Pathol Oral Cir Bucal 23:e38–e43. https://doi.org/10.4317/medoral.22123 Fong D, Bhatia KSS, Yeung D et  al (2010) Diagnostic accuracy of diffusion-weighted MR imaging for nasopharyngeal carcinoma, head and neck lymphoma and squamous cell carcinoma at the primary site. Oral Oncol 46:603–606 Foss F (2004) Mycosis fungoides and the Sezary syndrome. Curr Opin Oncol 16:421–428 Gritzmann N, Rettenbacher T, Hollerweger A et  al (2003) Sonography of the salivary glands. Eur Radiol 13:964–975 Hanna E, Wanamaker J, Adelstein D et  al (1997) Extranodal lymphomas of the head and neck. A 20-year experience. Arch Otolaryngol Head Neck Surg 123:1318–1323 Harnsberger RH (2004) Diagnostic imaging head and neck, 1st edn. Amirsys, Salt Lake City, UT

464 Harris NL, Jaffe ES, Diebold J et al (2000) World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: report of the Clinical Advisory Committee meeting-Airlie House, Virginia, November 1997. J Clin Oncol 17: 3835–3849 Harris NL, Jaffe ES, Stein H et  al (1994) A revised European-American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood 84:1361–1392 Hermans R (2004) Extranodal lymphoma—neck. Cancer Imaging 4:1–5 Hermans R, Horvath M, De Schrijver T et  al (1994) Extranodal non-Hodgkin lymphoma of the head and neck. J Belg Radiol 77:72–77 Kato H, Kanematsu M, Goto H et  al (2012) Mucosa-­ associated lymphoid tissue lymphoma of the salivary glands: MR imaging findings including diffusion-­ weighted imaging. Eur J Radiol 81:e612–e617. https:// doi.org/10.1016/j.ejrad.2011.12.035 Kim HC, Han MH, Kim KH et al (2002) Primary thyroid lymphoma: CT findings. Eur J Radiol 46:233–239 King AD, Ahuja AT, Yeung DKW et al (2007) Malignant cervical lymphadenopathy: diagnostic accuracy of diffusion-weighted MR imaging. Radiology 245:806–813 King AD, Lei KI, Richards PS et al (2003) Non-Hodgkin’s lymphoma of the nasopharynx: CT and MR imaging. Clin Radiol 58:621–625 King AD, Yuen EH, Lei KI et  al (2004) Non-Hodgkin lymphoma of the larynx: CT and MR imaging findings. Am J Neuroradiol 25:12–15 Koom WS, Chung EJ, Yang WI et al (2004) Angiocentric T-cell and NK/T-cell lymphomas: radiotherapeutic viewpoints. Int J Radiat Oncol Biol Phys 59:1127–1137 Kwee TC, Quarles van Ufford HME, Beek FJ et al (2009) Whole-body MRI, including diffusion-weighted imaging, for the initial staging of malignant lymphoma. Investig Radiol 44:683–690 Kwee TC, Takahara T, Ochiai R et  al (2008) Diffusion-­ weighted whole-body imaging with background body signal suppression (DWIBS): features and potential applications in oncology. Eur Radiol 18:1937–1952 Lee JL, Kim MK, Lee KH et al (2005) Extranodal marginal zone B-cell lymphomas of mucosa-associated lymphoid tissue-type of the orbit and ocular adnexa. Ann Hematol 84:13–18 Lister TA, Crowther D, Sutcliffe SB et al (1989) Report of a committee convened to discuss the evaluation and staging of patients with Hodgkin’s disease: Cotswolds meeting. J Clin Oncol 7:1630–1636 Littooij AS, Kwee TC, Barber I et al (2014) Whole-body MRI for initial staging of paediatric lymphoma: prospective comparison to an FDG-PET/CT-based reference standard. Eur J Radiol 24:1153–1165 Meignan M, Gallamini A, Haioun C (2009) Report on the first international workshop on interim-PET scan in lymphoma. Leuk Lymphoma 50:1257–1260

F. A. Pameijer and R. L. M. Haas Mendenhall WM, Mendenhall CM, Mendenhall NP (2003) Solitary plasmacytoma of bone and soft tissues. Am J Otolaryngol 24:395–399 Muwakkit SA, Razzouk BI, Shabb NS et al (2004) Clinical presentation and treatment outcome of children with Burkitt lymphoma in Lebanon: a single institution’s experience. J Pediatr Hematol Oncol 26:749–753 Palacios E, Larusso G, Rojas R et al (2004) Lymphoma of the parotid gland in Sjőgren’s syndrome. Ear Nose Throat J 83:156 Petralia G, Padhani AR (2018) Whole-body magnetic resonance imaging in oncology: indications. Magn Reson Imaging Clin N Am 26:495–507 Pickhardt PJ, Wippold FJ 2nd (1999) Neuroimaging in posttransplantation lymphoproliferative disorder. AJR Am J Roentgenol 172:1117–1121 Politi LS, Forghani R, Godi C et al (2010) Ocular adnexal lymphoma: diffusion-weighted MR imaging for differential diagnosis and therapeutic monitoring. Radiology 256:565–574 Priego G, Majos C, Climent F et al (2012) Orbital lymphoma: imaging features and differential diagnosis. Insight Imaging 3:337–344 Rao CR, Gutierrez MI, Bhatia K et al (2000) Association of Burkitt’s lymphoma with the Epstein-Barr virus in two developing countries. Leuk Lymphoma 39:329–337 Sekikawa Y, Hongo I (2017) HIV-associated benign lymphoepithelial cysts of the parotid glands confirmed by HIV-1 p24 antigen immunostaining. BMJ Case Rep. https://doi.org/10.1136/bcr-2017-221869 Shankland KR, Armitage JO, Hancock BW (2012) Non-­ Hodgkin lymphoma. Lancet 380(9844):848–857. https://doi.org/10.1016/S0140-6736(12)60605-9 Shields JA, Shields CL, Scartozzi R (2004) Survey of 1264 patients with orbital tumors and simulating lesions. Ophthalmology 111:997–1008 Smithers DW (1971) Summary of papers delivered at the Conference on Staging in Hodgkin’s Disease (Ann Arbor). Cancer Res 31:869–870 Som P, Curtin H, Mancuso A (2000) Imaging-based classification for evaluation of neck metastatic adenopathy. Am J Roentgenol 174:837–844 Sumi M, Ichikawa Y, Nakamura T (2007) Diagnostic ability of apparent diffusion coefficients for lymphomas and carcinomas in the pharynx. Eur Radiol 17:2631–2637 Swerdlow SH, Campo E, Pileri SA et al (2016) The 2016 revision of the World Health Organization classification of lymphoid neoplasms. Blood 127:2375–2390 Tailor TD, Gupta D, Dalley RW et al (2013) Orbital neoplasms in adults: clinical, radiologic and pathologic review. Radiographics 33:1739–1758 Takashima S, Nomura N, Noguchi Y et al (1995) Primary thyroid lymphoma: evaluation with US, CT and MRI. J Comput Assist Tomogr 19:282–288 Tart RP, Mukherji SK, Avino AJ et  al (1993) Facial lymph nodes: normal and abnormal CT appearance. Radiology 188:695–700

Neck Lymphoma Therasse P, Arbuck SG, Eisenhauer EA et al (2000) New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 92:205–216 Thoeny HC, De Keyzer F, King AD (2012) Diffusion-­ weighted MR imaging in the head and neck. Radiology 263:19–32 Tonami H, Matoba M, Kuginuki Y et al (2003) Clinical and imaging findings of lymphoma in patients with Sjőgren’s syndrome. J Comput Assist Tomogr 27:517–524 Tonami H, Matoba M, Yokota H et  al (2002) Mucosa-­ associated lymphoid tissue lymphoma in Sjőgren’s syndrome: initial and follow-up imaging features. Am J Roentgenol 179:485–489 UICC (2016) TNM classification of malignant tumours, 8th edn. Wiley Blackwell, Hoboken Vega F, Lin P, Medeiros J (2005) Extranodal lymphomas of the head and neck. Ann Diagn Pathol 9: 340–350 Vose JM, Chiu BC, Cheson BD et  al (2002) Update on epidemiology and therapeutics for non-Hodgkin’s lymphoma. Hematology 1:241–262 Walter C, Ziebart T, Sagheb K et  al (2015) Malignant lymphomas in the head and neck region—a retrospective, single-center study over 41 years. Int J Med Sci 12:141–145. https://doi.org/10.7150/ijms.10483. eCollection 2015

465 Watal P, Bathla G, Thaker S et  al (2018) Multimodality imaging spectrum of the extranodal lymphomas in the head and neck—a pictorial review. Curr Probl Diagn Radiol 47:340–352. https://doi.org/10.1067/j. cpradiol.2017.07.007 Weber AL, Rahemtullah A, Ferry JA (2003) Hodgkin and non-Hodgkin lymphoma of the head and neck: clinical, pathologic, and imaging evaluation. Neuroimaging Clin N Am 13:371–392 Weber AL, Randolph G, Aksoy FG (2000) The thyroid and parathyroid glands. CT and MR imaging and correlation with pathology and clinical findings. Radiol Clin North Am 38:1105–1129 World Health Organization Classification of Tumors (2001) In: Jaffe ES, Harris HL, Stein H et  al (eds) Tumors of haematopoietic and lymphoid tissues. Oxford University Press, Oxford, pp 162–167 Yahalom J, Illidge T, Specht L et al (2015) Modern radiation therapy for extranodal lymphomas: field and dose guidelines from the international lymphoma radiation oncology group. Int J Radiat Oncol Biol Phys 92:11–31 Yan J, Wu Z, Li Y (2004) The differentiation of idiopathic inflammatory pseudotumor from lymphoid tumors of orbit: analysis of 319 cases. Orbit 23:245–254 Yuen A, Jacobs C (1999) Lymphomas of the head and neck. Semin Oncol 26:338–345 Ziemer M, Bauer HI, Fluhr J et al (2008) Primary cutaneous follicle center lymphoma – ‘Crosti lymphoma’. Am J Clin Dermatol 9:133–136

Positron Emission Tomography in Head and Neck Cancer Ilona M. Schmalfuss

Contents

Abstract

1

Introduction

 468

2 2.1  2.2  2.3  2.4 

Clinical Applications Pretreatment Treatment Planning Treatment Surveillance Special Considerations for Some Histological Tumor Types

 468  470  482  483

References

 490  492

I. M. Schmalfuss (*) Department of Radiology, North Florida/South Georgia Veterans Administration and University of Florida College of Medicine, Gainesville, FL, USA e-mail: [email protected]

PET imaging complements CT and MRI in the staging process of head and neck malignancies. It has been shown to be in particular helpful in detection of early nodal metastasis, distant metastasis, and unknown primary tumors leading to change in TNM classification in up to 20% of patients. Tumor stage and CT tumor volumes have been well-established variables in the stratification process of patients into favorable and unfavorable treatment groups. PET imaging with its ability to capture biological activity and areas of hypoxia within tumors adds other important variables to consider with high FDG activities and hypoxia representing unfavorable risk factors. Utilization of FDG activity or hypoxia tumor maps facilitates more individualized treatment planning with the ability to deliver higher radiation doses to unfavorable tumor areas. Post treatment, PET imaging can help to decide if lymph node dissection is needed and to detect persistent or recurrent tumors in particular when significant alterations of the soft tissues planes are present. The strengths, weaknesses, and limitations of PET imaging prior, during, and post treatment will be discussed in this chapter.

Med Radiol Diagn Imaging (2020) https://doi.org/10.1007/174_2020_233, © This is a U.S. government work and not under copyright protection in the U.S.; foreign copyright protection may apply Published Online: 18 August 2020

467

I. M. Schmalfuss

468

1

Introduction

The prognosis of patients with head and neck cancer is closely related to tumor stage with advanced tumors carrying the most unfavorable prognosis. In addition, human papilloma virus (HPV) associated head and neck cancers carry a better prognosis which led into separation of HPV positive from HPV negative tumors in the recent TNM classification system (O’Sullivan et al. 2016; Malm et al. 2017). CT and or MRI are well-established tools in staging of head and neck malignancies that are routinely used in the clinical practice. These, however, have been shown to be deficient in detection of early nodal metastasis, distant metastasis, unknown primary tumors, and second primary cancers (Yongkui et  al. 2013; Xu et  al. 2011; Golusinski et  al. 2019). Positron emission tomography’s (PET) ability to measure the metabolic activity of tumors complements the anatomical information provided by CT and or MRI studies. PET has been reported to yield higher detection rates of occult primary tumors with sensitivity ranging from 15 to 73% and synchronous second primary tumors and unsuspected metastatic disease with an incidence of 4–8% and 7–15% than CT and or MRI (Abouzied et  al. 2017). Consequently, PET is gaining increasing clinical importance and utilization in the workup of patients with head and neck cancer, in particular when used as integrated PET/ CT scanner. The technical aspects of PET and PET/CT are reviewed in the chapter on “Imaging Techniques.”

2

Clinical Applications

The most commonly performed FDG-PET imaging is more complex and requires more careful patient preparation than CT and or MRI as FDG uptake is greatly influenced by different metabolic processes within the body. Blood glucose level >145  mg/dl (>8.0  mmol/l), fasting time

liver—lymph node ­ dissection or at least biopsy is required due to a hazard ratio of persistent disease of 6.18 in these patients (Fig. 22) ◦◦ Diffuse mild FDG uptake  >  liver—“waitand-see” approach as the FDG uptake is likely inflammatory in nature ◦◦ FDG uptake   2. All patients with a tumour-­to-­muscle ratio > 1.6 experienced tumour recurrence. The authors also reported temporally inconsistent 18FMISO distribution in that the location of regions which met the criteria for hypoxic changed with time (Eschmann et al. 2005). The potential of 18F-FDG-PET, 18FMISO-­PET, DW-MRI and dynamic contrast-enhanced (DCE)MRI to provide a potential BTV for dose painting was evaluated by Dirix and colleagues. After intravenous injection of a paramagnetic contrast agent, DCE-MRI shows enhancements in signal intensity on T1-weighted scans. Changes in signal intensity on DCE-MRI are related to tumour permeability, perfusion and interstitial pressures, which may influence treatment response. Fifteen patients with locally advanced HNC participated in a study of sequential CT/PET with 18F-FDG PET, 18FMISO PET and 1.5  T MRI performed before, during and after radical RT. GTVs delineated by the treating radiation oncologist were retrospectively retrieved while the GTVPET was

S. Nuyts and S. Deschuymer

automatically segmented based on a source-tobackground ratio. The GTVMRI was manually delineated by a radiation oncologist and radiologist in consensus. There was an excellent correlation between CT- and MRI-­based volumes, and all volumes showed significant shrinkage during RT, by approximately 50%. Both the GTVPET and GTVDW-MRI were significantly smaller than the GTVCT. Over a median follow-up of 30.7 months, disease recurred in seven patients; all recurrences were located within the area of overlap of the GTVCT, GTVMRI and GTVPET. There was little residual hypoxia on follow-up 18FMISO PET scan during the fourth week of treatment. DW-MRI showed residual disease in three patients and all developed locoregional recurrence. At 8 weeks after the end of treatment, 18FFDG PET suggested residual disease in two patients, both of whom ultimately recurred. DFS correlated negatively with baseline maximum tissue-to-blood 18FMISO ratio (T/Bmax), size of the hypoxic volume at baseline, and with T/Bmax on the 18FMISO scan during treatment. Three of the locoregional recurrences were outside the hypoxic volume defined by baseline 18FMISO PET. Compared with lesions that remained controlled, those which recurred had significantly lower ADC on DW-MRI during and after RT, and a significantly higher initial slope on baseline DCE-MRI (Dirix et al. 2008, 2009). Temporal and geographic stability is a significant concern for hypoxia-based dose escalation. Two distinct mechanisms can cause hypoxia. The first is the lack of sufficient numbers of tumour blood vessels causing chronic or diffusion-­ limited hypoxia. The second is related to the poor functionality of tumour vasculature with highly tortuous and poorly organised architecture, called the acute or perfusion-limited hypoxia. The oxygen concentration from subregions in the tumour can rapidly change if caused by ‘acute’ vessel instability or more slowly by reoxygenation in the course of the RT treatment (Fig.  5) (Joiner and van der Kogel 2009). Currently, there is only a weak correlation between the hypoxia PET tracers and the genetic hypoxia profiles (Löck et al. 2019). So although hypoxia is being investigated for many years,

Use of Imaging in Radiotherapy for Head and Neck Cancer

there is no clear consensus on the optimal dose required to eliminate hypoxic subpopulation, the optimal subvolume, the ideal identification method or PET tracer. How to handle the intratumoural hypoxia changes during the RT treatment is also still under investigation. 4.2.2.4  Apoptosis Apoptosis can be an indicator of intrinsic radiosensitivity because it is a major pathway of cell

Fig. 5  Temporal change in hypoxia. A 57-yearold patient presented with stage T3N2b squamous cell carcinoma of the oropharynx. Baseline 18 F-FDG PET (a) shows uptake at the primary tumour and regional lymph node. 18FMISO PET acquired one day later (b) indicates hypoxia in both the primary tumour and lymph node. Repeat 18 FMISO PET after 4 weeks of radiotherapy (c) and after the end of radiotherapy (d) shows decreased 18FMISO uptake

a

b

c

d

507

death after ionising irradiation. Annexin V is a protein which binds to membrane-bound phosphatidyl serine and is exposed on the surface of cells undergoing apoptosis. 99mTc-radiolabelled annexin V can, therefore, quantify apoptosis (Green and Steinmetz 2002). Vermeersch et  al. investigated in 18 patients the potential of 99mTc-­ radiolabelled annexin V scintigraphy to visualise the primary head and neck carcinoma in comparison to CT. The 99mTc-radiolabelled annexin V

508

allowed for the visualisation of all primary HNSCC identified by CT scan but failed to identify most of the sites of LN involvement (Vermeersch et  al. 2004). Van de Wiele et  al. (2003) showed that quantitative tumour apoptosis uptake values correlated well with the number of apoptotic cells on assays for apoptosis-induced DNA fragmentation. The authors concluded that this method could be used to monitor treatment response. In another publication, the radiolabelled annexin V tumour-to-background ratio derived from SPECT of 29 patients provided independent prognostic information on disease-­ free survival and overall survival (Loose et  al. 2008). Currently, this image tracer has no implications in the diagnostic process or treatment decision. 4.2.2.5  Receptor Status The proliferation of many tumours is regulated by factors that bind to membrane or intracellular receptors to activate signal transduction pathways. Imaging of receptor status, therefore, might provide information that can be used to guide treatment and prognosis. For example, epidermal growth factor receptor (EGFR) is a transmembrane glycoprotein involved in activating several pathways associated with proliferation, migration, stromal invasion, angiogenesis and resistance to cell death-inducing signals (Dancey 2004). EGFR is overexpressed in 80–100% of HNC, and this overexpression is associated with both increased metastatic potential and poor prognosis (Rubin Grandis et al. 1996; Ang et al. 2002). To date, contrast agents to visualise EGFR such as radiolabelled cetuximab are only used in preclinical models (Benedetto et al. 2019). In summary, functional imaging to guide radiotherapy planning is promising, and many clinical studies are running to establish their added value before it can be incorporated in daily practice.

4.3  Treatment Verification Treatment verification refers to the use of imaging to ensure that the planned RT treatment plan

S. Nuyts and S. Deschuymer

corresponds to the daily given treatment throughout the entire treatment period, taking into consideration factors such as patient immobilisation and positioning. Commonly in head and neck RT, individualised masks are fixed to the treatment table to prevent movement during treatment (intrafraction) and daily (interfraction). However, several factors such as changes in tumour size and shape, organ movement by respiration or swallowing, patient’s anatomy changes by weight loss and patient set-up errors make it essential to verify the patient’s position each day. Tumour and nodal volumes shrink by up to 3.0% per day, changing size, shape and position, sometimes asymmetrically (Castadot et al. 2010). Outer face and neck contour modifications may be seen as patients lose muscle mass and weight by treatment or tumour-induced dysphagia. Parotid glands not only decrease in volume but shift medially into the high-dose region with time (Barker et  al. 2004; Dirix et  al. 2008; Castadot et al. 2011). Resolving postoperative changes or even disease progression during treatment can also be seen. Thus, during RT treatment, the anatomy and geometry of the disease in relation to normal critical structures can alter significantly. Spatiotemporal instability of the target and normal structures and geometric uncertainty of patient positioning are critical in IMRT because of the sharp dose gradients involved (Castadot et  al. 2010). As a result of these changes, the received dose may differ significantly from what was planned. To avoid this, image-guided radiotherapy (IGRT) defined as the use of on-board imaging to improve patient set-up accuracy is essential. IGRT allows verification of the correct position before treatment initiation. Different methods can be used. Traditionally, a ‘port film’ used the fraction of the accelerator beam exiting from the patient to expose commercially available radiographic film. Especially bony anatomy is seen, while the tumour cannot be visualised. These images have less than diagnostic quality but are sufficient for comparison to planning images for gross errors. The next generation was called electronic portal imaging (EPI), which uses fluorescent screens, 2D ion chambers or matrix flat panel imagers and is again compared

Use of Imaging in Radiotherapy for Head and Neck Cancer

with a reference image. These 2D online electronic images have made it possible to correct patient positioning in real time, just before radiation is delivered although ‘matching’ is still performed on the bony anatomy of the patient (Fig.  6). On-board cone beam CT giving 3D information also allows visualisation of the soft tissue such as organs at risk. These better imaging techniques reduce the possible daily error and can lead to smaller PTV margins. The next step in reducing the daily error is adapting the treatment plan to the patient’s ‘anatomy of the day’, the so-called adaptive RT. However, treatment adaptation is very time consuming since all steps (delineation and treatment

a

b

Fig. 6 Comparison between digitally reconstructed radiograph derived from CT simulation images (a), and a port film, obtained during irradiation (b), which are compared to verify patient positioning and size and shape of the treatment portal

509

planning) must be redone and often requires multiple manual interventions. Therefore, patient selection might be essential and several authors have investigated which patients benefit the most from this adaptive RT approach. Brown et  al. (2015) identified predicting factors for the need for adaptive RT in oropharyngeal and nasopharyngeal cancer patient based on the N-status, pretreatment largest involved LN size and initial body weight. Furthermore, the ideal re-treatment planning for oropharyngeal and nasopharyngeal cancer patient differs. Replanning could be considered at the commencement of week 4 for patients with oropharyngeal cancer and in week 3 of RCT for patients with nasopharyngeal carcinoma (Brown et al. 2016).

4.4  Response Prediction Recognition of nonresponders or early progressors after/during radiotherapy is essential as these patients might benefit from intensified treatment (dose-escalation) or early surgical salvage (Farrag et al. 2010). Conversely, it is also important to identify the subgroup of very good responders for who the treatment could be de-­ intensified with the same oncologic outcome. Many studies are currently trying to establish the optimal treatment regimen for patients with Human Papillomavirus positive oropharyngeal cancer since these patients have, in general, a better prognosis. A less toxic treatment could greatly improve the quality of life of these patients (Deschuymer et al. 2018, b). Farrag et  al. investigated whether 18F-FDG PET during treatment can predict the outcome for HNC patients treated with primary (C) RT.  Forty-three consecutive patients underwent 18F-FDG-PET at baseline and at the end of the fourth week of therapy (approx. 47  Gy). Scans were analysed by two physicians with access to complete clinical information. 63% of patients had either hypopharynx or oropharynx cancer, 37% received cisplatin-based chemotherapy, and 44% had T3 or T4 disease. Median follow-up was 12.7  months. Two-year overall and disease-free survival was 66% and 52%, respectively. Median

510

SUVmax decreased from baseline to follow-up PET. Low SUVmax at baseline and on follow-up PET were significantly correlated with overall survival. There was a non-significant trend for patients with a complete metabolic response at follow-up PET to have better outcomes (Farrag et al. 2010). One of the challenges with the use of 18 F-FDG-PET during RT, especially after the fourth week, is that interpretation becomes more difficult due to the inflammation of both the tumour and the normal tissue (Geets et al. 2007). Baseline and mid-treatment 18FMISO PET/CT scanning have also been prospectively investigated (Lee et  al. 2009). Twenty HNC patients, 90% with locally advanced oropharyngeal primaries, underwent concurrent CRT.  All had pretreatment 18F-FDG and 18FMISO-PET/CT scans, with a mid-treatment 18FMISO PET scan 4 weeks after the start of treatment. However, the authors found that 18FMISO findings on the mid-­treatment scan did not correlate with patient outcome (Lee et al. 2009). A retrospective study of DW-MRI in response prediction included 38 patients with pathologically confirmed HNC treated to at least 60  Gy. Most received concurrent chemotherapy. Salvage treatment (neck dissection and/or intra-arterial chemotherapy) was ultimately delivered to eight patients. Pretreatment DW-MRI was performed a median of 7  days prior to initial treatment. The primary lesion region of interest was determined by consensus between two radiologists and one radiation oncologist without knowledge of locoregional disease status. Tumour volume, T stage, treatment and ADC correlated significantly with local disease status. ADC had a sensitivity of 77.8%, specificity of 100%, PPV of 100%, and NPV of 80% for the prediction of local failure in stage T3/T4 disease. Low ADC values pretreatment was correlated with high local control (Hatakenaka et al. 2011). Lambrecht et al. (2014) found more disease recurrences in tumours with a higher pretreatment ADC value. Kim et  al. reported the usefulness of pretreatment ADC for predicting treatment response in 33 cases. They calculated the ADC values of the neck nodes, not from the primary lesions. Pretreatment ADC value of complete responders was significantly

S. Nuyts and S. Deschuymer

lower than that from partial responders (Kim et al. 2009). The authors also performed a repeat MRI after 1 week of radiotherapy. The patients with a complete response showed a significantly higher increase in ADC than the partial responders by the first week of chemoradiotherapy. These data suggest that ADC can be used as a marker for prediction and early detection of response to therapy. In general, tumours showing a high ADC increase have a better treatment response than those with little or no ADC increase. The authors speculated that the apoptosis and necrosis of responding tumours present fewer microstructural barriers, thereby increasing ADC (Kim et al. 2009). In Vandecaveye et al. (2010), 30 patients received DW-MRI before and during CRT, a high delta ADC was predictive for better response and correlated significantly with 2-year locoregional control. DW-MRI has also been used prospectively to assess salivary gland function before and at a mean of 9  months after parotid-sparing RT in eight patients. RT was delivered to a dose of 60–72  Gy. The mean total dose to the spared parotid gland was 20  Gy, below the 25  Gy required for functional sparing. There were no dose constraints for the other parotid or either submandibular gland; typically these glands received >40 Gy. Before RT, a biphasic response to stimulation was confirmed in both parotids of all patients, identical to the pattern seen in healthy volunteers. In unspared glands, ADC was significantly higher after RT, possibly due to fibrosis, necrosis or both. In the spared parotid, ADC value after RT was not significantly different compared to before RT.  A comparable response to stimulation as before RT was observed in the spared but not the unspared parotid (Dirix et  al. 2008). This suggestion of preserved function in the spared gland was confirmed by low patient-­rated xerostomia scores and salivary gland scintigraphy results. The authors hypothesise that based on this data, DW-MRI allows non-invasive assessment of salivary gland functional changes with a sufficiently high spatial resolution that it could be used to improve models of dose-­response relationships (Dirix et al. 2008).

Use of Imaging in Radiotherapy for Head and Neck Cancer

4.5  Follow-Up Long-term follow-up after radical radiotherapy for HNC aims to evaluate treatment response, manage toxicity, exclude persistent disease, detect locoregional recurrence, and monitor for distant metastases or second primary tumours (Manikantan et  al. 2009). History, physical examination and endoscopy are the cornerstones of follow-up. Thyroid function studies should be performed after RT of the lower neck. More intense surveillance is recommended in the first 3 years after RT since most recurrences occur quite early. The optimal management of the nodal neck and imaging after chemoradiotherapy for head and neck squamous cell carcinoma with advanced nodal disease (stage N2–N3) was a matter of debate for a long time. A prospective multicentre randomised controlled trial was conducted to assess the noninferiority of PET-CT guided surveillance compared to planned neck dissection. If the PET-CT (12  weeks after the end of CRT) showed intense of mild FDG uptake of enlarged lymph nodes in the neck, neck dissection was still performed. This trial, known as the PET-­ NECK trial, showed similar overall survival among both groups, but there were considerably fewer surgical interventions in the PET-CT guided surveillance group. In addition, this treatment strategy was also more cost-effective (Mehanna et al. 2016). Unfortunately, imaging after (C)RT is challenging. Tissue changes after radiotherapy may be striking as they evolve over time (Glastonbury et al. 2010). The complexities of image interpretation after RT were pictorially reviewed by Glastonbury et  al. Radiation-induced changes may mask or mimic residual disease. Acute RT-related changes include mucositis, interstitial oedema, ischemia, and inflammation. Late side effects are associated with fibrosis and atrophy and can appear years after treatment. Severe but uncommon late effects include osteoradionecrosis and aerodigestive tract stenosis. Deep ulcers, new enhancing masses, lymphadenopathy or bone or cartilage destruction must be considered recurrence until proven otherwise. Nevertheless,

511

analysis of follow-up imaging is not always straightforward since anatomical imaging is limited by their dependency on size criteria in the detection of malignancy. Functional imaging modalities such as DWI-MRI can help distinguish persistent tumours from post-RT changes. Vandecaveye et al. investigated 26 patients with persistent or recurrent HNSCC after CRT planned for salvage surgery. Patients underwent DW-MRI as well as contrast-enhanced CT and 18F-FDG PET.  ADC values of the suspected primary or lymph node sites were significantly lower than the non-tumoural tissue with a sensitivity of 94.6%, specificity 95.9% and accuracy of 95.5%. Thus, RT-associated tissue changes show lower cellularity (and therefore higher ADC), compared to tumours which have high cellularity (with lower ADC). When compared with 18F-FDG PET (n = 17), DW-MRI correctly excluded tumour in three patients with hypermetabolic areas on PET, and correctly identified subcentimeter pathological lymph nodes without FDG uptake in two patients. The authors concluded that DW-MRI is optimally interpreted in conjunction with anatomic imaging (Vandecaveye et al. 2007).

5  Conclusion and Challenges No single imaging intervention is perfectly sensitive, specific, inexpensive, safe and convenient (Manikantan et  al. 2009). Anatomic and functional imaging provide complementary information which has already improved staging, radiotherapy delivery, response prediction and follow-up. Knowledge about the capabilities and limitations of all imaging modalities is essential for their rational clinical use, especially in the setting of imaging-intensive investigational RT protocols such as adaptive therapy. There remain hurdles to be overcome, including optimal timing of mid- and post-treatment imaging for each modality and accounting for spatiotemporal changes in tumour motion and hypoxia. Restrictions on PET use include the preferred method of segmentation and implementation of automated techniques. Many studies published to date have included small numbers of

512

patients with a wide variety of disease sites, biology, stages and treating institutions. This has the potential of confounding results due to different image acquisition protocols, instrumentation and interpretation methods (Kyzas et  al. 2008). Benefits of any imaging modality vary according to the local radiology experience available (Moeller et  al. 2010). Interpretation of certain imaging modalities and especially imaging performed after high-dose radiotherapy may require special expertise (Manikantan et al. 2009). Use of emerging technologies may address current challenges. For example, DW-MRI avoids the contrast required by dynamic MRI and CT that cannot be used in renal dysfunction (Hatakenaka et al. 2011). It may also be used as a non-invasive tool to select patients who could potentially benefit from treatment intensification such as dose painting. Further development of imaging markers which can predict radioresistance or outcome could assist with prescription of RT based on tumour phenotype or genotype. The successful introduction of functional imaging into routine clinical practice will depend on continued clinical studies with sound methodology and adequate followup (Dirix et al. 2009).

References Adams S, Baum RP, Stuckensen T, Bitter K, Hör G (1998) Prospective comparison of 18F-FDG PET with conventional imaging modalities (CT, MRI, US) in lymph node staging of head and neck cancer. Eur J Nucl Med 25:1255–1260 Ang KK, Berkey BA, Tu X, Zhang H-Z, Katz R, Hammond EH, Fu KK, Milas L (2002) Impact of epidermal growth factor receptor expression on survival and pattern of relapse in patients with advanced head and neck carcinoma. Cancer Res 62:7350–7356 Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tân PF, Westra WH, Chung CH, Jordan RC, Lu C, Kim H, Axelrod R, Silverman CC, Redmond KP, Gillison ML (2010) Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 363:24–35. https://doi.org/10.1056/ NEJMoa0912217 Barker JL, Garden AS, Ang KK, O’Daniel JC, Wang H, Court LE, Morrison WH, Rosenthal DI, Chao KSC, Tucker SL, Mohan R, Dong L (2004) Quantification of volumetric and geometric changes occurring during

S. Nuyts and S. Deschuymer fractionated radiotherapy for head-and-neck cancer using an integrated CT/linear accelerator system. Int J Radiat Oncol Biol Phys 59:960–970. https://doi. org/10.1016/j.ijrobp.2003.12.024 Benchaou M, Lehmann W, Slosman DO, Becker M, Lemoine R, Rufenacht D, Donath A (1996) The role of FDG-PET in the preoperative assessment of N-staging in head and neck cancer. Acta Otolaryngol 116:332–335 Benedetto R, Massicano AVF, Crenshaw BK, Oliveira R, Reis RM, Araújo EB, Lapi SE (2019) 89Zr-DFO-­ Cetuximab as a molecular imaging agent to identify cetuximab resistance in head and neck squamous cell carcinoma. Cancer Biother Radiopharm 34:288. https://doi.org/10.1089/cbr.2018.2616 Bernier J, Cooper JS, Pajak TF, van Glabbeke M, Bourhis J, Forastiere A, Ozsahin EM, Jacobs JR, Jassem J, Ang K-K, Lefèbvre JL (2005) Defining risk levels in  locally advanced head and neck cancers: a comparative analysis of concurrent postoperative radiation plus chemotherapy trials of the EORTC (#22931) and RTOG (#9501). Head Neck 27:843–850. https://doi. org/10.1002/hed.20279 Bhide SA, Kazi R, Newbold K, Harrington KJ, Nutting CM (2010) The role of intensity-modulated radiotherapy in head and neck cancer. Indian J Cancer 47:267– 273. https://doi.org/10.4103/0019-509X.64719 Biau J, Lapeyre M, Troussier I, Budach W, Giralt J, Grau C, Kazmierska J, Langendijk JA, Ozsahin M, O’Sullivan B, Bourhis J, Grégoire V (2019) Selection of lymph node target volumes for definitive head and neck radiation therapy: a 2019 update. Radiother Oncol 134:1– 9. https://doi.org/10.1016/j.radonc.2019.01.018 Blanchard P, Garden AS, Gunn GB, Rosenthal DI, Morrison WH, Hernandez M, Crutison J, Lee JJ, Ye R, Fuller CD, Mohamed ASR, Hutcheson KA, Holliday EB, Thaker NG, Sturgis EM, Kies MS, Zhu XR, Mohan R, Frank SJ (2016) Intensity-modulated proton beam therapy (IMPT) versus intensity-modulated photon therapy (IMRT) for patients with oropharynx cancer – a case matched analysis. Radiother Oncol 120:48–55. https://doi.org/10.1016/j.radonc.2016.05.022 Brierley JD, Gospodarowicz MK, Wittekind C (2017) UICC TNM classification of malignant tumours, 8th edn. Wiley Blackwell, Oxford Brown E, Owen R, Harden F, Mengersen K, Oestreich K, Houghton W, Poulsen M, Harris S, Lin C, Porceddu S (2015) Predicting the need for adaptive radiotherapy in head and neck cancer. Radiother Oncol 116:57–63. https://doi.org/10.1016/j.radonc.2015.06.025 Brown E, Owen R, Harden F, Mengersen K, Oestreich K, Houghton W, Poulsen M, Harris S, Lin C, Porceddu S (2016) Head and neck adaptive radiotherapy: predicting the time to replan. Asia Pac J Clin Oncol 12:460– 467. https://doi.org/10.1111/ajco.12516 Buck AK, Halter G, Schirrmeister H, Kotzerke J, Wurziger I, Glatting G, Mattfeldt T, Neumaier B, Reske SN, Hetzel M (2003) Imaging proliferation in lung tumors with PET: 18F-FLT versus 18F-FDG.  J Nucl Med 44:1426–1431

Use of Imaging in Radiotherapy for Head and Neck Cancer

513

Bussink J, van Herpen CML, Kaanders JHAM, Oyen Dandekar P, Partridge M, Kazi R, Nutting C, Harrington K, Newbold K (2010) Challenges in integrating WJG (2010) PET-CT for response assessment 18FDG PET-CT into radiotherapy planning of head and treatment adaptation in head and neck cancer. and neck cancer. Indian J Cancer 47:260–266. https:// Lancet Oncol 11:661–669. https://doi.org/10.1016/ doi.org/10.4103/0019-509X.64717 S1470-2045(09)70353-5 Castadot P, Geets X, Lee JA, Grégoire V (2011) Adaptive Deschuymer S, Dok R, Laenen A, Hauben E, Nuyts S (2018) Patient selection in human papillomavifunctional image-guided IMRT in pharyngo-laryngeal rus related oropharyngeal cancer: the added value squamous cell carcinoma: is the gain in dose distribuof prognostic models in the new TNM 8th edition tion worth the effort? Radiother Oncol 101:343–350. era. Front Oncol 8:273. https://doi.org/10.3389/ https://doi.org/10.1016/j.radonc.2011.06.011 fonc.2018.00273 Castadot P, Lee JA, Geets X, Grégoire V (2010) Adaptive radiotherapy of head and neck cancer. Semin Deschuymer S, Mehanna H, Nuyts S (2018) Toxicity reduction in the treatment of HPV positive oropharynRadiat Oncol 20:84–93. https://doi.org/10.1016/j. geal cancer: emerging combined modality approaches. semradonc.2009.11.002 Front Oncol 8:439. https://doi.org/10.3389/ Chao KS, Bosch WR, Mutic S, Lewis JS, Dehdashti F, fonc.2018.00439 Mintun MA, Dempsey JF, Perez CA, Purdy JA, Welch MJ (2001) A novel approach to overcome hypoxic Differding S, Sterpin E, Hermand N, Vanstraelen B, Nuyts tumor resistance: Cu-ATSM-guided intensity-­ S, de Patoul N, Denis J-M, Lee JA, Grégoire V (2017) Radiation dose escalation based on FDG-PET driven modulated radiation therapy. Int J Radiat Oncol Biol dose painting by numbers in oropharyngeal squamous Phys 49:1171–1182 cell carcinoma: a dosimetric comparison between Chapman JD, Baer K, Lee J (1983) Characteristics TomoTherapy-HA and RapidArc. Radiat Oncol 12:59. of the metabolism-induced binding of misonidahttps://doi.org/10.1186/s13014-017-0793-0 zole to hypoxic mammalian cells. Cancer Res 43: Differding S, Sterpin E, Janssens G, Hanin F-X, Lee JA, 1523–1528 Grégoire V (2016) Methodology for adaptive and Chapman JD, Bradley JD, Eary JF, Haubner R, Larson robust FDG-PET escalated dose painting by numbers SM, Michalski JM, Okunieff PG, Strauss HW, Ung in head and neck tumors. Acta Oncol 55:217–225. YC, Welch MJ (2003) Molecular (functional) imaging https://doi.org/10.3109/0284186X.2015.1046997 for radiotherapy applications: an RTOG symposium. Dirix P, De Keyzer F, Vandecaveye V, Stroobants S, Int J Radiat Oncol Biol Phys 55:294–301 Hermans R, Nuyts S (2008) Diffusion-weighted magCobben DCP, van der Laan BFAM, Maas B, Vaalburg W, netic resonance imaging to evaluate major salivary Suurmeijer AJH, Hoekstra HJ, Jager PL, Elsinga PH gland function before and after radiotherapy. Int J (2004) 18F-FLT PET for visualization of laryngeal Radiat Oncol Biol Phys 71:1365–1371. https://doi. cancer: comparison with 18F-FDG PET. J Nucl Med org/10.1016/j.ijrobp.2007.12.011 45:226–231 Cooper JS, Zhang Q, Pajak TF, Forastiere AA, Jacobs J, Dirix P, Nuyts S (2010) Value of intensity-modulated radiotherapy in Stage IV head-and-neck squamous Saxman SB, Kish JA, Kim HE, Cmelak AJ, Rotman cell carcinoma. Int J Radiat Oncol Biol Phys 78:1373– M, Lustig R, Ensley JF, Thorstad W, Schultz CJ, Yom 1380. https://doi.org/10.1016/j.ijrobp.2009.10.018 SS, Ang KK (2012) Long-term follow-up of the RTOG 9501/intergroup phase III trial: postoperative concur- Dirix P, Vandecaveye V, De Keyzer F, Op de Beeck K, Poorten VV, Delaere P, Verbeken E, Hermans R, rent radiation therapy and chemotherapy in high-risk Nuyts S (2010) Diffusion-weighted MRI for nodal squamous cell carcinoma of the head and neck. Int staging of head and neck squamous cell carcinoma: J Radiat Oncol Biol Phys 84:1198–1205. https://doi. impact on radiotherapy planning. Int J Radiat Oncol org/10.1016/j.ijrobp.2012.05.008 Biol Phys 76:761–766. https://doi.org/10.1016/j. Daisne J-F, Duprez T, Weynand B, Lonneux M, Hamoir ijrobp.2009.02.068 M, Reychler H, Grégoire V (2004) Tumor volume in pharyngolaryngeal squamous cell carcinoma: com- Dirix P, Vandecaveye V, De Keyzer F, Stroobants S, Hermans R, Nuyts S (2009) Dose painting in radioparison at CT, MR imaging, and FDG PET and validatherapy for head and neck squamous cell carcinoma: tion with surgical specimen. Radiology 233:93–100. value of repeated functional imaging with (18)F-FDG https://doi.org/10.1148/radiol.2331030660 PET, (18)F-fluoromisonidazole PET, diffusion-­ Daisne J-F, Sibomana M, Bol A, Cosnard G, Lonneux weighted MRI, and dynamic contrast-enhanced MRI. J M, Grégoire V (2003) Evaluation of a multimodality Nucl Med 50:1020–1027. https://doi.org/10.2967/ image (CT, MRI and PET) coregistration procedure jnumed.109.062638 on phantom and head and neck cancer patients: accuracy, reproducibility and consistency. Radiother Oncol Duprez F, De Neve W, De Gersem W, Coghe M, Madani I (2011) Adaptive dose painting by num69:237–245 bers for head-and-neck cancer. Int J Radiat Oncol Dancey JE (2004) Epidermal growth factor receptor and Biol Phys 80:1045–1055. https://doi.org/10.1016/j. epidermal growth factor receptor therapies in renal ijrobp.2010.03.028 cell carcinoma: do we need a better mouse trap? J Clin Oncol 22:2975–2977. https://doi.org/10.1200/ Emami B, Sethi A, Petruzzelli GJ (2003) Influence of MRI on target volume delineation and IMRT planning JCO.2004.04.934

514 in nasopharyngeal carcinoma. Int J Radiat Oncol Biol Phys 57:481–488 Eschmann S-M, Paulsen F, Reimold M, Dittmann H, Welz S, Reischl G, Machulla H-J, Bares R (2005) Prognostic impact of hypoxia imaging with 18F-misonidazole PET in non-small cell lung cancer and head and neck cancer before radiotherapy. J Nucl Med 46:253–260 Farrag A, Ceulemans G, Voordeckers M, Everaert H, Storme G (2010) Can 18F-FDG-PET response during radiotherapy be used as a predictive factor for the outcome of head and neck cancer patients? Nucl Med Commun 31:495–501. https://doi.org/10.1097/ MNM.0b013e3283334e2b Geets X, Daisne J-F, Gregoire V, Hamoir M, Lonneux M (2004) Role of 11-C-methionine positron emission tomography for the delineation of the tumor volume in pharyngo-laryngeal squamous cell carcinoma: comparison with FDG-PET and CT. Radiother Oncol 71:267–273. https://doi.org/10.1016/j. radonc.2004.02.013 Geets X, Tomsej M, Lee JA, Duprez T, Coche E, Cosnard G, Lonneux M, Grégoire V (2007) Adaptive biological image-guided IMRT with anatomic and functional imaging in pharyngo-laryngeal tumors: impact on target volume delineation and dose distribution using helical tomotherapy. Radiother Oncol 85:105–115. https://doi.org/10.1016/j.radonc.2007.05.010 Ghosh-Laskar S, Yathiraj PH, Dutta D, Rangarajan V, Purandare N, Gupta T, Budrukkar A, Murthy V, Kannan S, Agarwal JP (2016) Prospective randomized controlled trial to compare 3-dimensional conformal radiotherapy to intensity-modulated radiotherapy in head and neck squamous cell carcinoma: long-term results. Head Neck 38(Suppl 1):E1481–E1487. https:// doi.org/10.1002/hed.24263 Glastonbury CM, Parker EE, Hoang JK (2010) The postradiation neck: evaluating response to treatment and recognizing complications. AJR Am J Roentgenol 195:W164–W171. https://doi.org/10.2214/ AJR.09.4122 Green AM, Steinmetz ND (2002) Monitoring apoptosis in real time. Cancer J 8:82–92 Grégoire V, Evans M, Le Q-T, Bourhis J, Budach V, Chen A, Eisbruch A, Feng M, Giralt J, Gupta T, Hamoir M, Helito JK, Hu C, Hunter K, Johansen J, Kaanders J, Laskar SG, Lee A, Maingon P, Mäkitie A, Micciche F, Nicolai P, O’Sullivan B, Poitevin A, Porceddu S, Składowski K, Tribius S, Waldron J, Wee J, Yao M, Yom SS, Zimmermann F, Grau C (2018) Delineation of the primary tumour Clinical Target Volumes (CTV-P) in laryngeal, hypopharyngeal, oropharyngeal and oral cavity squamous cell carcinoma: AIRO, CACA, DAHANCA, EORTC, GEORCC, GORTEC, HKNPCSG, HNCIG, IAG-­ KHT, LPRHHT, NCIC CTG, NCRI, NRG Oncology, PHNS, SBRT, SOMERA, SRO, SSHNO, TROG consensus guidelines. Radiother Oncol 126:3–24. https:// doi.org/10.1016/j.radonc.2017.10.016 Guo T, Goldenberg D, Fakhry C (2017) AHNS series: do you know your guidelines? Management of head

S. Nuyts and S. Deschuymer and neck cancer in the era of human papillomavirus: educating our patients on human papillomavirus. Head Neck 39:833–839. https://doi.org/10.1002/hed.24693 Gupta T, Agarwal J, Jain S, Phurailatpam R, Kannan S, Ghosh-Laskar S, Murthy V, Budrukkar A, Dinshaw K, Prabhash K, Chaturvedi P, D’Cruz A (2012) Three-­ dimensional conformal radiotherapy (3D-CRT) versus intensity modulated radiation therapy (IMRT) in squamous cell carcinoma of the head and neck: a randomized controlled trial. Radiother Oncol 104:343–348. https://doi.org/10.1016/j.radonc.2012.07.001 Gupta T, Kannan S, Ghosh-Laskar S, Agarwal JP (2018) Systematic review and meta-analyses of intensity-­ modulated radiation therapy versus conventional two-­ dimensional and/or or three-dimensional radiotherapy in curative-intent management of head and neck squamous cell carcinoma. PLoS One 13:e0200137. https:// doi.org/10.1371/journal.pone.0200137 Hatakenaka M, Nakamura K, Yabuuchi H, Shioyama Y, Matsuo Y, Ohnishi K, Sunami S, Kamitani T, Setoguchi T, Yoshiura T, Nakashima T, Nishikawa K, Honda H (2011) Pretreatment apparent diffusion coefficient of the primary lesion correlates with local failure in head-and-neck cancer treated with chemoradiotherapy or radiotherapy. Int J Radiat Oncol Biol Phys 81: 339–345. https://doi.org/10.1016/j.ijrobp.2010.05.051 Hermans R, Feron M, Bellon E, Dupont P, Van den Bogaert W, Baert AL (1998) Laryngeal tumor volume measurements determined with CT: a study on intraand interobserver variability. Int J Radiat Oncol Biol Phys 40:553–557 Hermans R, Meijerink M, Van den Bogaert W, Rijnders A, Weltens C, Lambin P (2003) Tumor perfusion rate determined noninvasively by dynamic computed tomography predicts outcome in head-and-neck ­cancer after radiotherapy. Int J Radiat Oncol Biol Phys 57:1351–1356 Hsu S-H, Cao Y, Huang K, Feng M, Balter JM (2013) Investigation of a method for generating synthetic CT models from MRI scans of the head and neck for radiation therapy. Phys Med Biol 58:8419–8435. https:// doi.org/10.1088/0031-9155/58/23/8419 Hustinx R, Lucignani G (2010) PET/CT in head and neck cancer: an update. Eur J Nucl Med Mol Imaging 37:645–651. https://doi.org/10.1007/ s00259-009-1365-9 Hwang HS, Perez DA, Orloff LA (2009) Comparison of positron emission tomography/computed tomography imaging and ultrasound in staging and surveillance of head and neck and thyroid cancer. Laryngoscope 119:1958–1965. https://doi.org/10.1002/lary.20594 Jansen JFA, Schöder H, Lee NY, Wang Y, Pfister DG, Fury MG, Stambuk HE, Humm JL, Koutcher JA, Shukla-Dave A (2010) Noninvasive assessment of tumor microenvironment using dynamic contrast-­enhanced magnetic resonance imaging and 18F-fluoromisonidazole positron emission tomography imaging in neck nodal metastases. Int J Radiat Oncol Biol Phys 77:1403–1410. https://doi. org/10.1016/j.ijrobp.2009.07.009

Use of Imaging in Radiotherapy for Head and Neck Cancer Joiner M, van der Kogel A (2009) Basic clinical radiobiology, 4th edn. Hodder Arnold, Great Birtain Kam MKM, Leung S-F, Zee B, Chau RMC, Suen JJS, Mo F, Lai M, Ho R, Cheung K, Yu BKH, Chiu SKW, Choi PHK, Teo PML, Kwan W, Chan ATC (2007) Prospective randomized study of intensity-­ modulated radiotherapy on salivary gland function in early-stage nasopharyngeal carcinoma patients. J Clin Oncol 25:4873–4879. https://doi.org/10.1200/ JCO.2007.11.5501 Kim S, Loevner L, Quon H, Sherman E, Weinstein G, Kilger A, Poptani H (2009) Diffusion-weighted magnetic resonance imaging for predicting and detecting early response to chemoradiation therapy of squamous cell carcinomas of the head and neck. Clin Cancer Res 15:986–994. https://doi.org/10.1158/1078-0432. CCR-08-1287 Koh WJ, Rasey JS, Evans ML, Grierson JR, Lewellen TK, Graham MM, Krohn KA, Griffin TW (1992) Imaging of hypoxia in human tumors with [F-18]fluoromisonidazole. Int J Radiat Oncol Biol Phys 22:199–212 Kubota R, Kubota K, Yamada S, Tada M, Takahashi T, Iwata R, Tamahashi N (1995) Methionine uptake by tumor tissue: a microautoradiographic comparison with FDG. J Nucl Med 36:484–492 Kyzas PA, Evangelou E, Denaxa-Kyza D, Ioannidis JPA (2008) 18F-fluorodeoxyglucose positron emission tomography to evaluate cervical node metastases in patients with head and neck squamous cell carcinoma: a meta-analysis. J Natl Cancer Inst 100:712–720. https://doi.org/10.1093/jnci/djn125 Lambrecht M, Van Calster B, Vandecaveye V, De Keyzer F, Roebben I, Hermans R, Nuyts S (2014) Integrating pretreatment diffusion weighted MRI into a multivariable prognostic model for head and neck squamous cell carcinoma. Radiother Oncol 110:429–434. https:// doi.org/10.1016/j.radonc.2014.01.004 Langendijk JA, Lambin P, De Ruysscher D, Widder J, Bos M, Verheij M (2013) Selection of patients for radiotherapy with protons aiming at reduction of side effects: the model-based approach. Radiother Oncol 107:267–273. https://doi.org/10.1016/j. radonc.2013.05.007 Lee N, Nehmeh S, Schöder H, Fury M, Chan K, Ling CC, Humm J (2009) Prospective trial incorporating pre-/mid-treatment [18F]-misonidazole positron emission tomography for head-and-neck cancer patients undergoing concurrent chemoradiotherapy. Int J Radiat Oncol Biol Phys 75:101–108. https://doi. org/10.1016/j.ijrobp.2008.10.049 Ling CC, Humm J, Larson S, Amols H, Fuks Z, Leibel S, Koutcher JA (2000) Towards multidimensional radiotherapy (MD-CRT): biological imaging and biological conformality. Int J Radiat Oncol Biol Phys 47:551–560 Löck S, Linge A, Seidlitz A, Bandurska-Luque A, Nowak A, Gudziol V, Buchholz F, Aust DE, Baretton GB, Zöphel K, Steinbach J, Kotzerke J, Overgaard J, Zips D, Krause M, Baumann M, Troost EGC (2019) Repeat FMISO-PET imaging weakly correlates with hypoxia-­

515 associated gene expressions for locally advanced HNSCC treated by primary radiochemotherapy. Radiother Oncol 135:43–50. https://doi.org/10.1016/j. radonc.2019.02.020 Loose D, Vermeersch H, De Vos F, Deron P, Slegers G, Van de Wiele C (2008) Prognostic value of 99mTc-HYNIC annexin-V imaging in squamous cell carcinoma of the head and neck. Eur J Nucl Med Mol Imaging 35:47– 52. https://doi.org/10.1007/s00259-007-0577-0 Manikantan K, Khode S, Dwivedi RC, Palav R, Nutting CM, Rhys-Evans P, Harrington KJ, Kazi R (2009) Making sense of post-treatment surveillance in head and neck cancer: when and what of follow-up. Cancer Treat Rev 35:744–753. https://doi.org/10.1016/j. ctrv.2009.08.007 Mehanna H, Wong W-L, McConkey CC, Rahman JK, Robinson M, Hartley AGJ, Nutting C, Powell N, Al-Booz H, Robinson M, Junor E, Rizwanullah M, von Zeidler SV, Wieshmann H, Hulme C, Smith AF, Hall P, Dunn J, PET-NECK Trial Management Group (2016) PET-CT surveillance versus neck dissection in advanced head and neck cancer. N Engl J Med 374:1444–1454. https://doi.org/10.1056/ NEJMoa1514493 Moeller BJ, Rana V, Cannon BA, Williams MD, Sturgis EM, Ginsberg LE, Macapinlac HA, Lee JJ, Ang KK, Chao KSC, Chronowski GM, Frank SJ, Morrison WH, Rosenthal DI, Weber RS, Garden AS, Lippman SM, Schwartz DL (2010) Prospective imaging assessment of mortality risk after head-and-neck radiotherapy. Int J Radiat Oncol Biol Phys 78:667–674. https://doi. org/10.1016/j.ijrobp.2009.08.063 Mohamed ASR, Cardenas CE, Garden AS, Awan MJ, Rock CD, Westergaard SA, Brandon Gunn G, Belal AM, El-Gowily AG, Lai SY, Rosenthal DI, Fuller CD, Aristophanous M (2017) Patterns-of-failure guided biological target volume definition for head and neck cancer patients: FDG-PET and dosimetric analysis of dose escalation candidate subregions. Radiother Oncol 124:248–255. https://doi.org/10.1016/j. radonc.2017.07.017 Mortensen LS, Johansen J, Kallehauge J, Primdahl H, Busk M, Lassen P, Alsner J, Sørensen BS, Toustrup K, Jakobsen S, Petersen J, Petersen H, Theil J, Nordsmark M, Overgaard J (2012) FAZA PET/CT hypoxia imaging in patients with squamous cell carcinoma of the head and neck treated with radiotherapy: results from the DAHANCA 24 trial. Radiother Oncol 105:14–20. https://doi.org/10.1016/j.radonc.2012.09.015 Nordsmark M, Overgaard J (2004) Tumor hypoxia is independent of hemoglobin and prognostic for loco-­ regional tumor control after primary radiotherapy in advanced head and neck cancer. Acta Oncol 43:396–403 Nutting CM, Morden JP, Harrington KJ, Urbano TG, Bhide SA, Clark C, Miles EA, Miah AB, Newbold K, Tanay M, Adab F, Jefferies SJ, Scrase C, Yap BK, A’Hern RP, Sydenham MA, Emson M, Hall E, PARSPORT Trial Management Group (2011) Parotid-­ sparing intensity modulated versus conventional

516 radiotherapy in head and neck cancer (PARSPORT): a phase 3 multicentre randomised controlled trial. Lancet Oncol 12:127–136. https://doi.org/10.1016/ S1470-2045(10)70290-4 Olteanu LAM, Duprez F, De Neve W, Berwouts D, Vercauteren T, Bauters W, Deron P, Huvenne W, Bonte K, Goethals I, Schatteman J, De Gersem W (2017) Late mucosal ulcers in dose-escalated adaptive dose-­ painting treatments for head-and-neck cancer. Acta Oncol 57(2):262–268. https://doi.org/10.1080/02841 86X.2017.1364867 Patel MR, Hudgins PA, Beitler JJ, Magliocca KR, Griffith CC, Liu Y, Bougnon K, El-Deiry M, Saba NF, Aiken AH (2018) Radiographic imaging does not reliably predict macroscopic extranodal extension in human papilloma virus-associated oropharyngeal cancer. ORL J Otorhinolaryngol Relat Spec 80:85–95. https:// doi.org/10.1159/000487239 Pow EHN, Kwong DLW, McMillan AS, Wong MCM, Sham JST, Leung LHT, Leung WK (2006) Xerostomia and quality of life after intensity-modulated radiotherapy vs. conventional radiotherapy for early-stage nasopharyngeal carcinoma: initial report on a randomized controlled clinical trial. Int J Radiat Oncol Biol Phys 66:981–991. https://doi.org/10.1016/j. ijrobp.2006.06.013 Rasch C, Keus R, Pameijer FA, Koops W, de Ru V, Muller S, Touw A, Bartelink H, van Herk M, Lebesque JV (1997) The potential impact of CT-MRI matching on tumor volume delineation in advanced head and neck cancer. Int J Radiat Oncol Biol Phys 39:841–848. https://doi.org/10.1016/S0360-3016(97)00465-3 Rathod S, Gupta T, Ghosh-Laskar S, Murthy V, Budrukkar A, Agarwal J (2013) Quality-of-life (QOL) outcomes in patients with head and neck squamous cell carcinoma (HNSCC) treated with intensity-modulated radiation therapy (IMRT) compared to three-dimensional conformal radiotherapy (3D-CRT): evidence from a prospective randomized study. Oral Oncol 49:634–642. https:// doi.org/10.1016/j.oraloncology.2013.02.013 Rubin Grandis J, Melhem MF, Barnes EL, Tweardy DJ (1996) Quantitative immunohistochemical analysis of transforming growth factor-alpha and epidermal growth factor receptor in patients with squamous cell carcinoma of the head and neck. Cancer 78:1284–1292. https://doi.org/10.1002/ (SICI)1097-0142(19960915)78:63.0.CO;2-X Shields AF, Grierson JR, Dohmen BM, Machulla HJ, Stayanoff JC, Lawhorn-Crews JM, Obradovich JE, Muzik O, Mangner TJ (1998) Imaging proliferation in  vivo with [F-18]FLT and positron emission tomography. Nat Med 4:1334–1336. https://doi. org/10.1038/3337 Simone CB, Ly D, Dan TD, Ondos J, Ning H, Belard A, O’Connell J, Miller RW, Simone NL (2011) Comparison of intensity-modulated radiotherapy, adaptive radiotherapy, proton radiotherapy, and adaptive proton radiotherapy for treatment of locally advanced head and neck cancer. Radiother

S. Nuyts and S. Deschuymer Oncol 101:376–382. https://doi.org/10.1016/j. radonc.2011.05.028 Stoeckli SJ, Steinert H, Pfaltz M, Schmid S (2002) Is there a role for positron emission tomography with 18F-fluorodeoxyglucose in the initial staging of nodal negative oral and oropharyngeal squamous cell carcinoma. Head Neck 24:345–349. https://doi. org/10.1002/hed.10057 Toustrup K, Sørensen BS, Nordsmark M, Busk M, Wiuf C, Alsner J, Overgaard J (2011) Development of a hypoxia gene expression classifier with predictive impact for hypoxic modification of radiotherapy in head and neck cancer. Cancer Res 71:5923–5931. https://doi.org/10.1158/0008-5472.CAN-11-1182 Troost EGC, Bussink J, Hoffmann AL, Boerman OC, Oyen WJG, Kaanders JHAM (2010) 18F-FLT PET/ CT for early response monitoring and dose escalation in oropharyngeal tumors. J Nucl Med 51:866–874. https://doi.org/10.2967/jnumed.109.069310 Troost EGC, Schinagl DAX, Bussink J, Oyen WJG, Kaanders JHAM (2010) Clinical evidence on PET-CT for radiation therapy planning in head and neck tumours. Radiother Oncol 96:328–334. https://doi. org/10.1016/j.radonc.2010.07.017 Van de Wiele C, Lahorte C, Vermeersch H, Loose D, Mervillie K, Steinmetz ND, Vanderheyden J-L, Cuvelier CA, Slegers G, Dierck RA (2003) Quantitative tumor apoptosis imaging using technetium-­ 99m-­ HYNIC annexin V single photon emission computed tomography. J Clin Oncol 21:3483–3487. https://doi. org/10.1200/JCO.2003.12.096 Van den Bosch S, Dijkema T, Kunze-Busch MC, Terhaard CHJ, Raaijmakers CPJ, Doornaert PAH, Hoebers FJP, Vergeer MR, Kreike B, Wijers OB, Oyen WJG, Kaanders JHAM (2017) Uniform FDG-PET guided GRAdient Dose prEscription to reduce late Radiation Toxicity (UPGRADE-RT): study protocol for a randomized clinical trial with dose reduction to the elective neck in head and neck squamous cell carcinoma. BMC Cancer 17:208. https://doi.org/10.1186/ s12885-017-3195-7 Van der Veen J, Gulyban A, Nuyts S (2019) Interobserver variability in delineation of target volumes in head and neck cancer. Radiother Oncol 137:9–15. https://doi. org/10.1016/j.radonc.2019.04.006 Vandecaveye V, De Keyzer F, Nuyts S, Deraedt K, Dirix P, Hamaekers P, Vander Poorten V, Delaere P, Hermans R (2007) Detection of head and neck squamous cell carcinoma with diffusion weighted MRI after (chemo)radiotherapy: correlation between radiologic and histopathologic findings. Int J Radiat Oncol Biol Phys 67:960–971. https://doi.org/10.1016/j. ijrobp.2006.09.020 Vandecaveye V, De Keyzer F, Vander Poorten V, Dirix P, Verbeken E, Nuyts S, Hermans R (2009) Head and neck squamous cell carcinoma: value of diffusion-weighted MR imaging for nodal staging. Radiology 251:134– 146. https://doi.org/10.1148/radiol.2511080128 Vandecaveye V, Dirix P, De Keyzer F, de Beeck KO, Vander Poorten V, Roebben I, Nuyts S, Hermans

Use of Imaging in Radiotherapy for Head and Neck Cancer R (2010) Predictive value of diffusion-weighted magnetic resonance imaging during chemoradiotherapy for head and neck squamous cell carcinoma. Eur Radiol 20:1703–1714. https://doi.org/10.1007/ s00330-010-1734-6 Vermeersch H, Loose D, Lahorte C, Mervillie K, Dierckx R, Steinmetz N, Vanderheyden J-L, Cuvelier C, Slegers G, Van de Wiele C (2004) 99mTc-HYNIC

517 Annexin-V imaging of primary head and neck carcinoma. Nucl Med Commun 25:259–263 Zukauskaite R, Hansen CR, Brink C, Johansen J, Asmussen JT, Grau C, Overgaard J, Eriksen JG (2017) Analysis of CT-verified loco-regional recurrences after definitive IMRT for HNSCC using site of origin estimation methods. Acta Oncol 56:1554–1561. https://doi.org/10.1080/0284186X.2017.1346384