Revision Spine Surgery: Pearls and Pitfalls 1498773826, 9781498773829

Revision spine surgery requires a unique skill set different from performing a primary operation. Understanding when a s

878 65 20MB

English Pages xviii+270 [289] Year 2019

Report DMCA / Copyright


Polecaj historie

Revision Spine Surgery: Pearls and Pitfalls
 1498773826, 9781498773829

Citation preview

Revision Spine Surgery Pearls and Pitfalls

Revision Spine Surgery Pearls and Pitfalls

Edited by

Gregory D. Schroeder, MD

Assistant Professor of Orthopaedic Surgery at Thomas Jefferson University, Spine Surgeon at The Rothman Institute, Philadelphia, Pennsylvania

Ali A. Baaj, MD

Associate Professor of Neurological Surgery Co-Director, Spinal Deformity and Scoliosis Program, Weill Cornell Medical College, New York-Presbyterian Hospital, New York City, New York

Alexander R. Vaccaro, MD, PhD, MBA

Richard H. Rothman Professor and Chairman, Department of Orthopaedic Surgery, Professor of Neurosurgery, Co-Director, Delaware Valley Spinal Cord Injury Center, Co-Chief of Spine Surgery, Sidney Kimmel Medical Center at Thomas Jefferson University, President, Rothman Institute, Philadelphia, Pennsylvania

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2020 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed on acid-free paper International Standard Book Number-13: 978-1-4987-7382-9 (Hardback) This book contains information obtained from authentic and highly regarded sources. While all reasonable efforts have been made to publish reliable data and information, neither the author[s] nor the publisher can accept any legal responsibility or liability for any errors or omissions that may be made. The publishers wish to make clear that any views or opinions expressed in this book by individual editors, authors or contributors are personal to them and do not necessarily reflect the views/opinions of the publishers. The information or guidance contained in this book is intended for use by medical, scientific or health-care professionals and is provided strictly as a supplement to the medical or other professional’s own judgement, their knowledge of the patient’s medical history, relevant manufacturer’s instructions and the appropriate best practice guidelines. Because of the rapid advances in medical science, any information or advice on dosages, procedures or diagnoses should be independently verified. The reader is strongly urged to consult the relevant national drug formulary and the drug companies’ and device or material manufacturers’ printed instructions, and their websites, before administering or utilizing any of the drugs, devices or materials mentioned in this book. This book does not indicate whether a particular treatment is appropriate or suitable for a particular individual. Ultimately it is the sole responsibility of the medical professional to make his or her own professional judgements, so as to advise and treat patients appropriately. The authors and publishers have also attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access ( or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Names: Vaccaro, Alexander R., editor. | Baaj, Ali A., editor. | Schroeder, Gregory D., editor. Title: Revision spine surgery : pearls and pitfalls / edited by Alexander R. Vaccaro, Ali Baaj, Gregory D. Schroeder. Other titles: Revision spine surgery (Vaccaro) Description: Boca Raton : CRC Press, 2019. | Includes index. Identifiers: LCCN 2019010037| ISBN 9781498773829 (hardback : alk. paper) | ISBN 9780429188848 (e-book) Subjects: | MESH: Spine--surgery | Reoperation--methods | Spinal Diseases--surgery Classification: LCC RD768 | NLM WE 725 | DDC 617.4/82--dc23 LC record available at Visit the Taylor & Francis Web site at and the CRC Press Web site at

This book is dedicated to my wife Katie, and our three sons, Leo, Henry and Grant. Your support is an incredible gift that I am blessed to have. Gregory D. Schroeder This book is dedicated to my beautiful daughter, Hannah, who has filled our hearts with endless joy. Ali A. Baaj This book is dedicated to my nephew Luke Vaccaro. There is a wise saying that the meaning of life is to find your gift and the purpose of life is to give that gift away. Luke your gift is your courage, tenacity, resiliency and love for family. Your parents are proud of you and you will always be a role model for your younger brother Drew. Alexander R. Vaccaro


Video list Contributors

xi xiii









The approach to revision procedures Joseph A. Weiner and Wellington K. Hsu How to dissect the plane between the scar of a laminectomy defect in the posterior cervical spine Ken Ishii How to dissect the plane between the scar of a laminectomy defect in the posterior thoracic and lumbar spine Nickul S. Jain and Raymond J. Hah Local muscle flaps in the setting of revision spine surgery: Indications, operative planning, principles, and postoperative management Briar L. Dent, Jaime L. Bernstein, and Jason A. Spector Revision and reimplantation of a spinal cord stimulator device Fadi Al-Saiegh, John M. DePasse, Francis J. Sirch IV, Gregory D. Schroeder, and Chengyuan Wu



27 35





7 8 9

Revision ACDF at the same level Fadi Al-Saiegh, George M. Ghobrial, and James S. Harrop Revision ACDF: Adjacent level Courtney Pendleton, Matthew S. Galetta, and Jack Jallo Converting a total disc replacement to an ACDF Joseph D. Smucker and Rick C. Sasso Treatment of adjacent segment disease after total disc replacement (TDR) Bruce V. Darden II

47 51 59


viii Contents






12 13 14 15

Revision suboccipital decompression for complex Chiari malformation Jacob L. Goldberg, Ibrahim Hussain, Ali A. Baaj, and Jeffrey P. Greenfield How to revise a failed occipital cervical fusion Joshua T. Wewel, Mazda K. Turel, Joseph E. Molenda, and  Vincent C. Traynelis How to revise a failed C1–C2 fusion Nizar Moayeri and Michael G. Fehlings Treatment of postlaminectomy kyphosis Christopher T. Martin and John M. Rhee Revision of failed posterior cervical fusions Trevor Mordhorst, Vadim Goz, and William Ryan Spiker Complications necessitating surgical intervention following cervical laminoplasty Michael J. Moses, Amos Z. Dai, and Themistocles S. Protopsaltis


83 89 101 109





17 18


Revision surgery for proximal junctional kyphosis following thoracolumbar fusion Sundeep S. Saini, Daniel Cataldo, Christopher R. Cook, Hamadi Murphy, Paul W. Millhouse, and Kris Radcliff Pedicle subtraction osteotomy (PSO) nonunion revision Jason W. Savage Treatment of a nonunion of a thoracolumbar deformity, not at the site of a three-column osteotomy Randall B. Graham, Tyler R. Koski, and Patrick A. Sugrue How to safely remove a pedicle screw abutting the aorta Kevin Savage, Paul W. Millhouse, Hamadi Murphy, Gregory D. Schroeder, and Alexander R. Vaccaro


133 145






22 23


Revision of an anterior lumbar interbody fusion (ALIF) nonunion Edward Delsole, Rishi Sharma, and Gregory D. Schroeder How to revise nonunion of a lateral lambar interbody fusion (LLIF) through a lateral approach Heeren S. Makanji, Jacqueline Koomson, Dhruv K.C. Goyal, and Gregory D. Schroeder How to surgically manage a recurrent lumbar herniated nucleus pulposus (HNP) Taylor Paziuk, Matthew S. Galetta, and Jeffrey A. Rihn How to perform revision lumbar decompression Jacob Hoffman, Ryan Murphy, Mark L. Prasarn, and  Shah-Nawaz M. Dodwad How to perform revision lumbar decompression at the index level through a minimally invasive (MIS) approach Aaron Hillis, Christoph Wipplinger, Sertac Kirnaz, Franziska A. Schmidt, and Roger Härtl


161 167


Contents ix



27 28

29 30 31

How to revise a transforaminal lumbar interbody fusion (TLIF) nonunion with recurrent stenosis at the index level (open) Jesse E. Bible and Gregory Pace How to revise a minimally invasive transforaminal lumbar interbody fusion (MIS TLIF) nonunion with recurrent stenosis at the index level through an MIS approach Fady Y. Hijji, Ankur S. Narain, Gregory D. Lopez, Krishna T. Kudaravalli, Kelly H. Yom, and Kern Singh How to revise a posterior lateral decompression and fusion at the index level Fadi Sweiss, Cristian Gragnaniello, Anthony J. Caputy, and Michael Rosner How to revise a posterior lumbar fusion that has developed adjacent-level stenosis with or without instability Patrick Curry and Mark F. Kurd Flat back deformity revision surgery Jefferson Wilson, Matthew S. Galetta, and Srinivas Prasad Revision high-grade spondylolisthesis surgery Peter D. Angevine Management of a ventrally displaced graft following ALIF, TLIF or DLIF Dhruv K.C. Goyal, Heeren S. Makanji, Gregory D. Schroeder, and  Brian W. Su




205 211 217 223





33 34 35 36

Treatment of symptomatic cervical and lumbar pseudomeningocoeles Joshua E. Heller and George Rymarczuk Treatment of a persistent cervical dural tear Jessica L. Block and D. Greg Anderson Treatment of a ventral thoracic dural defect Ibrahim Hussain, Peter F. Morgenstern, and Ali A. Baaj Treatment of a persistent lumbar dural tear Joseph S. Butler, Matthew S. Galetta, and Barrett I. Woods Treatment of a chronic postoperative cervical and lumbar spine infection Kamil Okroj and Christopher Kepler


241 245 251 255 261

Video list

Video 6.1

ACDF Revision Same Level

Video 7.1

ACDF Revision ASD

Video 11.1

Failed OC Fusion

Video 15.1

Protopsaltis French Door Laminoplasty

Video 16.1

PJK Correction Video



D. Greg Anderson md Departments of Orthopaedics and Neurological Surgery Rothman Institute Thomas Jefferson University Philadelphia, Pennsylvania Peter D. Angevine md, mph Associate Professor Quality Chair Department of Neurological Surgery Columbia University Daniel and Jane Och Spine Hospital New York City, New York Ali A. Baaj md Associate Professor of Neurological Surgery Co-Director Spinal Deformity and Scoliosis Program Weill Cornell Medical College New York-Presbyterian Hospital New York City, New York Jaime L. Bernstein md Division of Plastic Surgery New York–Presbyterian Hospital/Weill Cornell Medicine New York City, New York Jesse E. Bible md Assistant Professor Department of Orthopaedics and Rehabilitation Pennsylvania State University Milton S. Hershey Medical Center Hershey, Pennsylvania

Jessica L. Block bs Drexel University College of Medicine Philadelphia, Pennsylvania Joseph S. Butler phd, frcs Rothman Institute Thomas Jefferson University Philadelphia, Pennsylvania Anthony J. Caputy md Department of Neurosurgery George Washington University School of Medicine & Health Sciences George Washington University Washington, DC Daniel Cataldo do Orthopedic Spine Surgery Fellow Icahn School of Medicine at Mount Sinai New York City, New York Christopher R. Cook do Orthopedic Spine Surgeon Graves Gilbert Clinic Western Kentucky Orthopaedic and Neurosurgical Associates Bowling Green, Kentucky Patrick Curry md Western Orthopaedics Denver, Colorado Amos Z. Dai md Department of Orthopaedic Surgery New York University Langone Medical Center New York City, New York


xiv Contributors

Bruce V. Darden, II md OrthoCarolina Spine Center Charlotte, North Carolina Edward Delsole md Department of Orthopaedic Surgery Rothman Orthopaedic Institute Thomas Jefferson University Philadelphia, Pennsylvania Briar L. Dent md Division of Plastic Surgery New York-Presbyterian Hospital/Weill Cornell Medicine New York City, New York John M. DePasse md Department of Orthopaedic Surgery Rothman Institute Thomas Jefferson University Philadelphia, Pennsylvania Shah-Nawaz M. Dodwad md Department of Orthopaedic Surgery University of Texas Health Sciences Center at Houston Houston, Texas Michael G. Fehlings md, phd, frcsc Division of Neurosurgery Toronto Western Hospital University Health Network and Department of Surgery University of Toronto Toronto, Canada Matthew S. Galetta ba Department of Orthopaedic Surgery Rothman Institute Merion Station, Pennsylvania George M. Ghobrial md Department of Neurological Surgery Thomas Jefferson University Hospital Philadelphia, Pennsylvania Jacob L. Goldberg md Department of Neurological Surgery Weill Cornell Brain and Spine Center New York, New York

Dhruv K.C. Goyal ba Department of Orthopaedic Surgery Rothman Orthopaedic Institute Thomas Jefferson University Philadelphia, Pennsylvania Vadim Goz md Department of Orthopaedics University of Utah School of Medicine Salt Lake City, Utah Cristian Gragnaniello md, phd Department of Neurosurgery George Washington University Washington, DC Randall B. Graham md Department of Neurological Surgery Northwestern University Feinberg School of Medicine Chicago, Illinois Jeffrey P. Greenfield md, phd Pediatric Neurological Surgery Weill Cornell Brain and Spine Center New York, New York Raymond J. Hah md Assistant Professor of Orthopaedic Surgery and Neurosurgery Keck School of Medicine University of Southern California Los Angeles, California James S. Harrop md Professor of Neurological Surgery and Orthopedics Department of Neurological Surgery Division of Spine and Peripheral Nerve Surgery Thomas Jefferson University Hospital Philadelphia, Pennsylvania Roger Härtl md Department of Neurological Surgery Weill Cornell Brain and Spine Center New York-Presbyterian Hospital/Weill Cornell Medicine New York City, New York Joshua E. Heller md, mba Department of Orthopaedic Surgery Rothman Institute Thomas Jefferson University Philadelphia, Pennsylvania

Contributors xv

Aaron Hillis, md Department of Neurological Surgery Weill Cornell Brain and Spine Center New York-Presbyterian Hospital/Weill Cornell Medicine New York City, New York

Sertac Kirnaz md Department of Neurological Surgery Weill Cornell Brain and Spine Center New York-Presbyterian Hospital/Weill Cornell Medicine New York City, New York

Fady Y. Hijji bs Department of Orthopaedic Surgery Rush University Medical Center Chicago, Illinois

Jacqueline Koomson ms Department of Orthopaedic Surgery Rothman Orthopaedic Institute Thomas Jefferson University Philadelphia, Pennsylvania

Jacob Hoffman md Department of Orthopaedic Surgery University of Texas Health Sciences Center at Houston Houston, Texas Wellington K. Hsu md Clifford C. Raisbeck Distinguished Professor of Orthopaedic Surgery Northwestern University Feinberg School of Medicine Chicago, Illinois Ibrahim Hussain md Department of Neurological Surgery Weill Cornell Brain and Spine Center New York-Presbyterian Hospital New York City, New York Ken Ishii md, phd Keio University School of Medicine Tokyo, Japan Jack Jallo md Department of Neurological Surgery Jefferson Medical College Thomas Jefferson University Philadelphia, Pennsylvania Nickul S. Jain md Southern California Orthopedic Institute Bakersfield, California Christopher Kepler md Orthopaedic Surgery Thomas Jefferson University Hospital Philadelphia, Pennsylvania

Tyler R. Koski md Department of Neurological Surgery Northwestern University Feinberg School of Medicine Chicago, Illinois Krishna T. Kudaravalli bs Department of Orthopaedic Surgery Rush University Medical Center Chicago, Illinois Mark F. Kurd md Associate Professor of Orthopaedic Surgery Thomas Jefferson University Philadelphia, Pennsylvania Gregory D. Lopez md Department of Orthopaedic Surgery Rush University Medical Center Chicago, Illinois Heeren S. Makanji md Department of Orthopaedic Surgery Rothman Orthopaedic Institute Thomas Jefferson University Philadelphia, Pennsylvania Christopher T. Martin md Orthopaedic Surgery Emory Spine Center Emory University School of Medicine Atlanta, Georgia Paul W. Millhouse md, mba Thomas Jefferson University Philadelphia, Pennsylvania

xvi Contributors

Nizar Moayeri md, phd University Medical Center Utrecht Utrecht University Utrecht, Netherlands Joseph E. Molenda md Department of Neurosurgery Rush University Medical Center Chicago, Illinois Michael J. Moses md Department of Orthopaedic Surgery New York University Langone Medical Center New York City, New York Trevor Mordhorst bs University of Wyoming Laramie, Wyoming Peter F. Morgenstern md Department of Neurological Surgery Weill Cornell Brain and Spine Center New York-Presbyterian Hospital New York City, New York Hamadi Murphy md SIU School of Medicine Southern Illinois University Springfield, Illinois Ryan Murphy md Department of Orthopaedic Surgery University of Texas Health Sciences Center at Houston Houston, Texas Ankur S. Narain ba Department of Orthopaedic Surgery Rush University Medical Center Chicago, Illinois Kamil Okroj md Department of Orthopaedic Surgery Sidney Kimmel Medical College Jefferson University Philadelphia, Pennsylvania Gregory Pace md Orthopaedic Resident Department of Orthopaedics and Rehabilitation Milton S. Hershey Medical Center Pennsylvania State University Hershey, Pennsylvania

Taylor Paziuk md Department of Orthopaedic Surgery Sidney Kimmel Medical College Jefferson University Philadelphia, Pennsylvania Courtney Pendleton md Department of Neurological Surgery Sidney Kimmel Medical College Jefferson University Philadelphia, Pennsylvania Mark L. Prasarn md Department of Orthopaedic Surgery University of Texas Health Sciences Center at Houston Houston, Texas Srinivas Prasad md, ms Department of Neurological Surgery Sidney Kimmel Medical College Jefferson University Philadelphia, Pennsylvania Themistocles S. Protopsaltis md Chief, Division of Spine Surgery Associate Professor of Orthopaedic Surgery and Neurosurgery Department of Orthopedic Surgery NYU Langone Health New York City, New York Kris Radcliff md Departments of Orthopedic Surgery and Neurological Surgery Rothman Institute Thomas Jefferson University Philadelphia, Pennsylvania John M. Rhee md Orthopaedic Surgery Emory Spine Center Emory University School of Medicine Atlanta, Georgia Jeffrey A. Rihn md Rothman Orthopaedics Philadelphia, Pennsylvania Michael Rosner md Professor of Surgery F. Edward Hebert School of Medicine Bethesda, Maryland

Contributors xvii

George Rymarczuk md Department of Orthopaedic Surgery Rothman Institute Thomas Jefferson University Philadelphia, Pennsylvania

Francis J. Sirch IV ba Department of Orthopaedic Surgery Rothman Institute Thomas Jefferson University Philadelphia, Pennsylvania

Fadi Al-Saiegh md Department of Neurosurgery Surgery Thomas Jefferson University Hospital Philadelphia, Pennsylvania

Joseph D. Smucker md Indiana Spine Group Carmel, Indiana

Sundeep S. Saini do PGY-IV Department of Orthopaedic Surgery Rowan University School of Osteopathic Medicine Stratford, New Jersey Rick C. Sasso md Indiana Spine Group Carmel, Indiana Jason W. Savage md Associate Professor of Orthopaedic Surgery Cleveland Clinic Cleveland, Ohio Kevin Savage md Franziska A. Schmidt md Department of Neurological Surgery Weill Cornell Brain and Spine Center New York-Presbyterian Hospital/Weill Cornell Medicine New York City, New York Gregory D. Schroeder md Assistant Professor of Orthopaedic Surgery Thomas Jefferson University and Spine Surgeon The Rothman Institute Philadelphia, Pennsylvania

Jason A. Spector md, facs Division of Plastic Surgery New York–Presbyterian Hospital/Weill Cornell Medicine New York City, New York William Ryan Spiker md Department of Orthopaedic Surgery University of Utah Salt Lake City, Utah Brian W. Su md California Orthopedics and Spine Larkspur, California Patrick A. Sugrue md Department of Neurological Surgery Feinberg School of Medicine Northwestern University Chicago, Illinois Fadi Sweiss md Department of Neurological Surgery George Washington School of Medicine and Health Sciences George Washington University Washington, DC Vincent C. Traynelis md Department of Neurosurgery Rush University Medical Center Chicago, Illinois

Rishi Sharma bs Department of Orthopaedic Surgery Rothman Orthopaedic Institute Thomas Jefferson University Philadelphia, Pennsylvania

Mazda K. Turel md Department of Neurosurgery Rush University Medical Center Chicago, Illinois

Kern Singh md Department of Orthopaedic Surgery Rush University Medical Center Chicago, Illinois

Alexander R. Vaccaro md, phd, mba Department of Orthopedic Surgery Thomas Jefferson University Philadelphia, Pennsylvania

xviii Contributors

Joseph A. Weiner md Department of Orthopaedic Surgery Feinberg School of Medicine Northwestern University Chicago, Illinois Joshua T. Wewel md Department of Neurosurgery Rush University Medical Center Chicago, Illinois Jefferson Wilson md, phd Department of Neurosurgery University of Toronto Toronto, Ontario Christoph Wipplinger md Department of Neurological Surgery Weill Cornell Brain and Spine Center New York-Presbyterian Hospital/Weill Cornell Medicine New York City, New York

Barrett I. Woods md Rothman Institute Thomas Jefferson University Philadelphia, Pennsylvania and Department of Orthopaedic Surgery Feinberg School of Medicine Northwestern University Chicago, Illinois Chengyuan Wu md, msbme Department of Orthopaedic Surgery Rothman Institute Thomas Jefferson University Philadelphia, Pennsylvania Kelly H. Yom ba Department of Orthopaedic Surgery Rush University Medical Center Chicago, Illinois


1 The approach to revision procedures 3 Joseph A. Weiner and Wellington K. Hsu 2 How to dissect the plane between the scar of a laminectomy defect in the posterior cervical spine 13 Ken Ishi 3 How to dissect the plane between the scar of a laminectomy defect in the posterior thoracic and lumbar spine 19 Nickul S. Jain and Raymond J. Hah 4 Local muscle flaps in the setting of revision spine surgery: Indications, operative planning, principles, and postoperative management 27 Briar L. Dent, Jaime L. Bernstein, and Jason A. Spector 5 Revision and reimplantation of a spinal cord stimulator device 35 Fadi Al-Saiegh, John M. DePasse, Francis J. Sirch IV, Gregory D. Schroeder, and Chengyuan Wu

1 The approach to revision procedures JOSEPH A. WEINER AND WELLINGTON K. HSU Introduction Patient evaluation Clinical considerations

3 3 5

INTRODUCTION Spine surgery is frequently performed in the treatment of spine trauma, tumors, and complex degenerative disorders. With an estimated 413,000 fusion procedures performed in the United States annually, the number of procedures performed has increased 2.4-fold since 1998.1 As more patients undergo spine surgery and patients live longer, a greater proportion of these patients will require revision surgery. Apart from common etiologies such as stenosis and disk herniation, recurrent back pain after surgery can result from infection, iatrogenic fracture, failure of fusion, or adjacent segment pathology. Pseudarthrosis has been reported at rates as high as 48% in multilevel posterolateral lumbar fusions.2 Given that failed spine surgery is a major burden on our healthcare system, substantial research has been dedicated to identifying factors that contribute to this problem and developing methods for successful revision surgery. Revision surgical procedures are challenging undertakings for both surgeons and patients alike. Multiple host and surgical issues often complicate revision spine procedures, including impaired biology secondary to the primary surgical procedure. Often, the factors that originally lead to pseudarthrosis are still present and must be

Conclusion References

8 8

addressed prior to repeat procedures. Compared with primary surgeries, these interventions have higher rates of return to the operating room, infection rates, and inferior patient-reported outcomes (according to the Owesty Disability Index/Visual Analog Scale [ODI/VAS]).3–5 Due to the risks of revision spine surgery, proper patient assessment and selection are critical for patients with failed spine surgery.

PATIENT EVALUATION Patient-reported satisfaction following both cervical and lumbar fusion surgeries is highly variable.6 Identification of factors leading to poor patientreported outcomes is a major focus of recent spine research. Studies have identified that preoperative factors such as smoking, poor mental health, obesity, low bone mineral density and workers compensation status are negative, independent predictors of patient reported outcomes.6,7 In addition, patient expectations are also a strong predictor of postoperative satisfaction.

History Recurrent neck/back pain and/or neurological symptoms following spine surgery require careful 3

4  The approach to revision procedures

assessment to determine the exact symptom generator. The first step in this process is obtaining a detailed history to establish the duration and character of their symptoms. The time course of symptoms helps to define whether there was a pain-free interval following the initial surgery. Absence of a pain-free interval may indicate an inadequate surgical decompression or failure of surgical technique. Conversely, in cases when the patient initially has immediate relief but symptoms recur during follow-up after a fusion procedure, one must consider pseudarthrosis, infection, or adjacent segment disease. Asking the patient about the character of symptoms can help to determine if the current symptoms are similar or different to their complaint before surgery. Mechanical back pain that is worse with movement can be indicative of pseudarthrosis, while nonmechanical pain associated with fever, chills and weight loss may suggest infection. A crucial element of history-taking in the postsurgical patient population is the mental health evaluation. Numerous studies have identified mental health as a strong predictor of patient-reported outcomes.8,9 Both preemptive and postoperative depression symptoms correlate with clinical outcomes after lumbar spine surgery.10 Hart et  al. recently reported that lumbar spine fusion has also been associated with symptoms of posttraumatic stress disorder (PTSD) during the first postoperative year in 11% of patients undergoing elective lumbar fusion. Postoperative PTSD was a stronger predictor of reduced clinical benefit than either preoperative psychiatric diagnosis or preoperative mental composite scores.11 Diagnostic criteria for depression includes sleep disturbances, loss of interest in daily activities, feelings of guilt, lack of energy, impaired cognition and concentration, loss of appetite, psychomotor retardation, and suicidal ideation. Addressing these mental health concerns prior to revision surgery is critical to ensure a good outcome.12

Physical exam A thorough physical exam should be performed to evaluate for the common causes of back pain. Special attention should be given to the neurological exam, including sensory, motor, deep tendon

Table 1.1  Waddell criteria of non-organic back pain Tenderness • Tender to superficial palpation • Non-anatomic distribution of tenderness (i.e. pain over pelvis, thoracic spine, etc.) Simulation • Low back pain with axial compression on cranium • Increase in low back pain with passive rotation of the shoulders and pelvis in the same plane Distraction • Difference in results of straight leg raise in sitting vs. supine positions • Inconsistent exam findings when patient is distracted Regional Disturbances • Motor: Generalized weakness of the lower extremities with cogwheel resistance on exam. • Sensory: Non-dermatomal distribution of sensory loss. Stocking and glove distribution. Overreaction • Disproportionate response to stimulus • Bracing: Both arms supporting body weight while seated • Clutching back for >3 seconds • Dramatic grimacing • Collapsing

reflexes, and gait pattern. During the evaluation of any patient with neck and/or back pain, the surgeon must always consider nonorganic etiologies for their symptoms. Waddell’s signs are a group of well-described physical exam findings that suggest a nonorganic and/or psychosocial etiology to low back pain. These include superficial or diffuse nonanatomic tenderness, overreaction to nonpainful stimuli, and change in exam findings when the patient is distracted13 (see Table 1.1).

Imaging After completion of a thorough history and physical examination, imaging studies are generally warranted to evaluate for pseudarthrosis or new

Clinical considerations  5

pathology. Plain radiography is generally used for the initial assessment of pseudarthrosis because of its widespread availability and low cost. However, these studies may significantly overestimate the likelihood of fusion. Brodsky et  al. demonstrated a 64% correlation rate between postoperative anteroposterior and lateral radiographs and surgical exploration.14 Furthermore, the time to radiographic presentation of a pseduarthrosis can vary between patients. Although many clinical studies use 1 year as the end point for a fusion study,15,16 Kim et  al. reported an average time of 3.5 years (range 12–131 months) before the detection of pseudarthrosis using plain radiographs.17 It is well known that the appearance of a fusion radiographically can change even after the 1-year milestone. Another option for evaluation of arthrodesis is utilizing flexion-extension films to assess for motion of the fused segment. Pseudarthrosis is likely when motion is present; however, there is not a consensus regarding the amount of motion that is considered a solid fusion. Criteria for approval by the U.S. Food and Drug Administration (FDA) of spine fusion systems includes evidence of bridging trabecular bone between the involved motion segment, translational motion 90% chance of improvement. There is only a fair chance of improvement of axial neck pain, and thus that should not be the primary indication for surgery. Biomechanically, the construct should approach the kinematics of the native vertebral segments. However, the literature is too sparse and immature to definitively comment on the effect of TDR on adjacent segment disease. Also, conflicting thoughts exist on whether there is an increased or diminished risk of developing heterotopic ossification with multilevel constructs. Finally, patients need to be aware that if a TDR is planned, but intraoperative radiographic visualization of the target level is inadequate, the procedure will have to be converted to an ACDF.

PRINCIPLES OF REVISION SURGERY The key principles of revision anterior cervical surgery are careful preoperative planning and meticulous dissection. Preoperative imaging should assure the surgeon that there are no anatomical nor index surgery anomalies. These anomalies can be planned for and hopefully avoided. Next is the approach dissection. Typically, little significant scarring occurs anterior to the pretracheal fascia. Posterior to the pretracheal fascia, the esophagus and the carotid sheath are vulnerable to injury. To avoid these structures, the surgeon should extend the dissection either proximal or distal to that of the index procedure. This extended

62  Treatment of adjacent segment disease after total disc replacement (TDR)

dissection will allow a more normal plane between these structures to expose the anterior aspect of the vertebral bodies between the longus colli muscle masses. The dissection should be done bluntly to avoid visceral or vascular injuries. If exposure is initially required distally, the omohyoid muscle may be transected. The vascular supply for the omohyoid is in the cranial third of the muscle. Once the midline is exposed, sharp dissection can be used to elevate the vertebral body scar tissue. From there, surgery ­proceeds as with the primary procedures.

PREOPERATIVE PLANNING/ OPERATING ROOM (OR) SETUP Prior to recommending surgery to the patient, the radiographic studies must be thoroughly evaluated, as must the factors listed here in indications. If cervical TDR is indicated, the patient must be counseled that if the operative segment cannot be adequately visualized radiologically in the operating room (OR), the procedure will have to be converted to an ACDF. With placement of an adjacent TDR, the size of the vertebral bodies should be evaluated. In patients with smaller vertebral bodies, especially if a keeled TDR has been previously implanted, a nonkeeled TDR should be considered to avoid a split vertebral body fracture. The side of approach will have to be determined. If an ipsilateral approach to the primary surgery is recommended, then the surgery can proceed. If a contralateral approach is desired, an ear, nose, and throat (ENT) evaluation by indirect laryngoscopy is necessary to rule out an occult recurrent laryngeal nerve injury. If this injury is present, consideration of an ipsilateral approach is recommended to avoid the increased risk of a bilateral recurrent laryngeal nerve injury, and thus dysphonia. Finally, thorough evaluation of the radiographic studies is important to avoid vertebral artery anomalies and other anatomic anomalies. Intraoperatively, meticulous setup is another key to surgical success, especially with TDR. With anesthesia, intubation should be performed with the neck in a neutral position, especially in the myelopathic patient. Mean arterial pressure should be maintained at or above 90 mmHg to ensure adequate spinal cord perfusion. I prefer using an esophageal stethoscope to allow palpation

of the esophagus during exposure, which helps avoid the potentially devastating complication of esophageal perforation. The stethoscope use for esophageal localization outweighs the potential risk of recurrent laryngeal nerve injury theoretically occurring by entrapment between the trachea and t he ­now-rigid esophagus. Lastly, corticoste­roids should be considered for ­spinal cord ­protection and to potentially diminish p ­ ostoperative swelling and dysphagia. We give appropriate prophylactic antibiotics based on a preop methicillin-resistant Staphylococcus ­aureus/­methicillin-susceptible S. aureus (MRSA/ MSSA) screen. For TDR, a radiolucent OR table is required to permit anterioposterior (AP) as well as lateral visualization of the cervical spine. The Mayfield head holder cannot be used for a TDR because it interferes with AP radiographic evaluation. The neck is positioned in a neutral position, with the head resting on a round gel pad and a rolled sheet under the neck to avoid neck movement during the surgery. The head is then taped to the bed, with padding placed over the brow ridge. The shoulders are also taped to the bed to improve distal cervical radiographic visualization. The C-arm and image intensifier are brought in. The surgeon must ensure that there are true AP and lateral radiographic images of the operative segments. The uncovertebral joints are better landmarks than the spinous processes to identify the midline on AP images. Intraoperative neurophysiological monitoring is important in my regimen. I include somatosensory evoked potentials (SSEPs), motor-evoked potentials (MEPs), and free-run electromyographies (EMGs). While not the standard of care, it allows real-time neurological assessment and the potential to take protective steps if spinal cord/nerve root injury is suspected. MEPs require total intravenous anesthesia and obviate the use of muscle relaxants.

OPERATIVE TECHNIQUE The surgical procedure involves a standard transverse approach, on the side decided preoperatively by the surgeon. Blunt dissection is emphasized deep to the pretracheal fascia. This dissection minimizes the risks to the carotid sheath and the esophagus. The exposure can be facilitated by the

Complications 63

use of Kittner dissectors and a Freer elevator. Once these structures are identified, the esophagus is mobilized and retracted; sharp dissection in the midline between the longus colli muscle masses can be carried out. The longus colli muscle can be elevated bilaterally to allow a Cloward-type retractor placement. A marker is then placed adjacent to the presumed operative disc space. A C-arm lateral radiograph is obtained to verify the level. The C-arm is moved superiorly between uses to allow unimpeded access to the surgical site. The next step is to apply the Caspar-type distraction pins, either for the TDR or ACDF. Some TDR instrument sets have their own distraction pins. To facilitate implantation of a TDR, placement of the distraction pins should be done using the C-arm, assuring placement in the true midline. Midline placement will allow quicker and more accurate TDR preparation later. Anterior discectomy and decompression is done as is typical for the surgeon and the specific pathology. If TDR is to be used, it is important to err on the aggressive side because it is not possible to depend on distraction for decompression. Also, the pathology can recur because there will be continued motion. Use of a high-speed burr should be minimized with a TDR, and only with copious irrigation, due to the potential increased risk of heterotopic ossification. Typically, the posterior longitudinal ligament is resected, facilitated by operative microscope use. At this point, the TDR is implanted per the protocol of the manufacturer. Care should be taken to choose the optimal-sized implant. The implant should cover the end plates as much as possible. In the lateral plane, the implant must be placed as posteriorly as possible, especially with constrained implants, to reestablish normal kinematics. The end plates prior to implantation should be parallel on the lateral plane, to avoid the TDR being placed in extension, potentially limiting motion and allowing abnormal implant contact. After implantation, any bleeding bony surfaces should be treated with bone wax. Copious irrigation is carried out and a standard closure applied, with a drain placed at the surgeon’s discretion. If ACDF is being contemplated, several techniques may be used. If the surgeon wants to use a standard interbody graft/cage and plate, the plate

should be as short as possible to avoid anterior ligamentous ossification disease. Diverging screws in the plate ensure a stable construct. I typically avoid the zero-profile integrated cage/screw devices in revision cases if the index procedure is an ACDF. However, if the index case is a functioning TDR, less stress is transferred to the cage/screw device, with potentially less risk of pseudarthrosis. These devices having a lower profile may diminish postoperative dysphagia compared to ACDF plates.

POSTOPERATIVE MANAGEMENT Postoperative care of these patients is essentially no different from the typical care for the index procedure. Prophylactic antibiotics are continued only for 24 hours. No collar is used except for rare circumstances in which a soft collar is provided for comfort. A drain is used at the surgeon’s discretion; in obese patients or patients with difficult dissections, drainage is probably necessary. Our criteria for doing these cases as an outpatient include minimal bleeding with the dissection, age ≤ 60 years, body mass index (BMI)  ≤ 30. ASA ≤ 2, and a nonsmoker. We observe the patient for 4 hours to ensure that there are no swallowing issues. If the patient meets these criteria, then he or she is fine for discharge. Any other concerns warrant an overnight observation stay. We prescribe a 3-week course of nonsteroidal anti-inflammatory drugs (NSAIDs) to diminish the risk of heterotopic ossification (HO) after TDR implantation. However, it has been observed in our patients that heterotopic ossification may progress for up to 4 years postprocedure. Keeled implants have a higher rate of HO, so this factor needs to be considered in the preoperative planning. Lastly, no restrictions are recommended to postoperative activity. The patient may resume his or her lifestyle, as tolerated.

COMPLICATIONS The complications from surgery for adjacent segment disease are typically no different from the index procedure except for the approach. Though still rare, there is a slightly higher risk of injury to the esophagus in the redo approach. This risk can be minimized by carrying out the approach from

64  Treatment of adjacent segment disease after total disc replacement (TDR)

the contralateral side to the index procedure or by slow, blunt dissection with early localization of the esophagus during an ipsilateral approach. High suspicion should be taken for these injures; an intraoperative ENT consult should be requested if the injury occurs. The risk of recurrent laryngeal nerve injury (and thus dysphonia) is higher than in primary surgery, but the risks can be mitigated by preoperative ENT evaluation and meticulous dissection. Dysphagia is likewise more common in redo surgery, but intraoperative corticosteroids may lower the risk. Vertebral artery injuries are exceedingly rare and are minimized by preoperative evaluation for vascular anomalies and early identification of the midline during dissection. Nerve root and spinal cord injury should be no more common than in the index procedure.

Pearls and Pitfalls The key to success in redo anterior cervical surgery is meticulous preoperative planning. The other strategies are fairly straightforward. In TDR implantation, keeled implants should be avoided in patients with small vertebral bodies (typically smaller men and most women). The TDR implant should be sized to adequately cover the vertebral body, thus avoiding subsidence. The prosthesis should be undersized in the cranial-caudal dimension. Too large a prosthesis can limit range of motion of the

segment. A constrained prosthesis placed too anteriorly can alter the center of rotation, transferring more stress to the facet joints. Ensuring that both vertebral end plates are parallel prevents implantation of the TDR in extension, which can limit range of motion and potentially accelerate implant wear. Copious irrigation, minimal use of the burr, and generous use of bone wax and NSAIDs help prevent HO after TDR implantation. With ACDF, similar principles to the index procedure apply. A zero-profile device is acceptable adjacent to a TDR. When using a plate for the ACDF, the plate should be applied as short as possible to avoid anterior ligamentous ossification disease, potentially limiting range of motion of the adjacent TDR. Finally, with adequate planning, adjacent anterior cervical surgery can be performed safely, with satisfactory clinical results.

REFERENCE 1. Barbagallo GMV, Assietti R, Corbino L et al. Early results and review of the literature of a novel hybrid technique combining cervical arthrodesis and disc arthroplasty for treating multilevel degenerative disc disease: opposite or complimentary techniques? Eur Spine J 2009;18(suppl. 1):S29–S39, doi: 10.1007/ s00586-009.0978-9.


Part     Posterior Cervical

10 Revision suboccipital decompression for complex Chiari malformation Jacob L. Goldberg, Ibrahim Hussain, Ali A. Baaj, and Jeffrey P. Greenfield 11 How to revise a failed occipital cervical fusion Joshua T. Wewel, Mazda K. Turel, Joseph E. Molenda, and  Vincent C. Traynelis 12 How to revise a failed C1–C2 fusion Nizar Moayeri and Michael G. Fehlings 13 Treatment of postlaminectomy kyphosis Christopher T. Martin and John M. Rhee 14 Revision of failed posterior cervical fusions Trevor Mordhorst, Vadim Goz, and William Ryan Spiker 15 Complications necessitating surgical intervention following cervical laminoplasty Michael J. Moses, Amos Z. Dai, and Themistocles S. Protopsaltis



83 89 101 109


10 Revision suboccipital decompression for complex Chiari malformation JACOB L. GOLDBERG, IBRAHIM HUSSAIN, ALI A. BAAJ, AND JEFFREY P. GREENFIELD Indications Relative contraindications Expectations Principles of revision surgery

67 68 69 70

INDICATIONS The Chiari malformations are a diverse group of anomalies in which a mismatch between the posterior fossa neural and bony elements may result in a relative descent of the cerebellar tonsils through the foramen magnum. This may lead to dorsal compression of the brainstem and disturbance of normal cerebrospinal fluid (CSF) circulation. Patients can present with a variety of symptoms, including but not limited to exertionally induced occipital headaches, neck pain, dysphagia, sleep apnea, dysmetria, and paraesthesia. Hydrocephalus and spinal cord syringomyelia can also develop, which can result in further neurologic deficits and lead to myelopathy and scoliosis. For patients in whom clinical and radiographic criteria are met, suboccipital decompression is an option to palliate symptoms or address syringomyelia or a combination of interdependent neurologic findings. A decompression of the foramen magnum, often paired with a C1 laminectomy and duraplasty, is

Preoperative planning and operating room (OR) setup Operative technique Postoperative management Complications

70 71 74 74

the standard procedure offered, although many variations in surgical technique exist (Figure 10.1). The degree of suboccipital and upper cervical decompression required should be determined by the patient’s clinical findings, in conjunction with the anatomy/radiographic findings, and the degree to which neural tissue is compressed. In certain scenarios, posterior fossa compression may be primarily due to bony involvement, and a duraplasty may not be offered initially. C1 laminectomy should usually be performed to assist with intradural exploration and duraplasty, particularly when compression is present well below the foramen magnum, although anatomic variation dictates that occasionally a C1 laminectomy may not be required, or conversely, a C2 laminectomy or laminoplasty in fact may be warranted for extremely caudally oriented compression. Patients with persistent symptoms despite initial surgical intervention pose a challenging clinical scenario. Further clinical assessment and radiographic evaluations may be useful to 67

68  Revision suboccipital decompression for complex Chiari malformation



Figure 10.1  Sagittal T2 MRI scans of a patient with Type 1 Chiari malformation. (a) Preoperatively, note the descent of the cerebellar tonsils through the foramen magnum, resulting in cervicomedullary compression. The dashed line is drawn from the dorsal-inferior tip of the clivus (basion) to opisthion, roughly approximating the anatomical position of the foramen magnum (also referred to as McRae’s line). The solid line represents the distance of cerebellar tonsil descent, with values of 5 mm or greater suggesting a diagnosis of Type 1 Chiari malformation. (b) Postoperatively seen following suboccipital decompression, C1 laminectomy, and autologous duraplasty. Note the reduction in crowding of the cerebellar tonsils characterized by the presence of T2 hyperintense CSF space dorsal to the cerebellum.

determine if a reoperation might be expected to further palliate or address recalcitrant symptomatology not addressed by a first surgery. Patients may have persistent posterior fossa crowding, inadequately decompressed syringomyelia, and/ or abnormal CSF flow within the central nervous system (CNS), resulting in persistent symptoms, pain, or clinical manifestations of CNS dysfunction. There are several reasons why Chiari decompressions fail, including (1) failure to diagnosis the actual cause of neurologic symptoms in the presence of an incidental Chiari malformation; (2)  failure to diagnose a CSF leak or idiopathic intracranial hypertension (IIH) as the etiology for cerebellar ectopia (i.e., unnecessary decompression); (3) proper diagnosis but inadequate decompression; (4) proper diagnosis but overly aggressive decompression; (5) adequate decompression but operative complications, such as CSF leak or infection, leading to scarring or pseudomeningocoele; or (6) adequate decompression without perioperative complications in the presence of abnormal skull base geometry or hereditary connective tissue disorder, leading to postoperative craniocervical instability (CCI) with ventral brainstem compression. Each of these etiologies presents a clinical conundrum with its own

challenges. In this chapter, we are going to focus on some of the technical aspects learned through attempting many of these revision surgeries and offer strategies to decide to whom to offer further surgery and how to minimize the inherent risks.

RELATIVE CONTRAINDICATIONS There are two critical relative contraindications for repeat suboccipital decompression in patients with Chiari 1 or Chiari 1.5, despite persistent symptoms following an initial surgery. First, patients with continued neck and occipital pain, particularly when upright or with activity, may be experiencing symptoms of CCI rather than static posterior fossa compression. For these patients, further decompression may exacerbate instability, and surgeons must be vigilant for patients with concomitant connective tissue disorders such as Ehlers–Danlos syndrome (EDS). Flexion-extension x-rays may demonstrate listhesis that confirms the diagnosis; however, even with negative flexion/extension plain films, a hard-cervical collar trial for 2–6 weeks may be warranted. These trials in patients with craniocervical instability (CCI) due to EDS may result in obvious self-reported improvements in pain and symptoms, providing additional support for a diagnosis of CCI.

Expectations 69

In these patients, occipitocervical fusion may be discussed as a more appropriate solution than revision decompression surgery. Another situation where repeat suboccipital decompression may be relatively contraindicated is when there is persistent or particularly worsened ventral compression of the brainstem. Patients with congenital skull base abnormalities such as platybasia and basilar invagination in association with Chiari 1, or even more commonly Chiari 1.5, malformations may have persistent symptoms of brainstem compression that are not relieved by posterior fossa decompression alone. Even when the cerebellum displays relative dorsal drift away from the medulla following a suboccipital craniectomy, persistent kinking of the brainstem, usually at the cervicomedullary junction, may continue to cause compression and abnormal CSF pulsations through the foramen magnum, contributing to hydrocephalus, syrinx, cranial neuropathies, and long tract signs (Figure 10.2). The clivo-axial angle (CXA), also known as the clivuscanal angle, is a helpful radiographic measurement determined from the angle at the intersection of the lines delineating the dorsal slope of the clivus and the posterior spinal line behind the vertebral body of C2. Typically, angles less than 125–130 degrees suggest significant ventral compression. In these situations, repeat posterior fossa decompression may provide little to no benefit. Therefore, vental decompression is required, either through transoral or transnasal (a)

approaches for the anterior arch of C1 resection, odontoidectomy, and dorsal clivusectomy when indicated to relieve symptoms. Of note, these patients require occipitocervical fusion prior to ventral decompression, given the instability that results from the anterior approach. There is also a rare but recognized subgroup of patients in whom neural imaging fails to explain symptoms of brainstem compression that are clinically apparent; in these patients, close attention to the contribution of connective tissue disorders and possible inclusion of dynamic imaging modalities may be necessary to identify the etiology of often very severe symptomatology.

EXPECTATIONS Patient expectations of revision suboccipital decompression surgery in patients with Chiari malformation require significant patient counseling. Many of the symptoms that these patients develop can be nonspecific with unpredictable temporal associations. These patients should have detailed workups for other neurologic causes of headaches, paresthesias, and dizziness prior to considering reoperation if initial decompression does not resolve or only temporarily relieves symptoms. Given that many of these patients are in the pediatric and adolescent population, the stress of going through additional surgeries and hospital stays should be respected and referrals for psychosocial support made as necessary.


Figure 10.2  Sagittal T2 MRI scans of a patient with Type 1 Chiari malformation and concomitant basilar invagination who did not improve clinically following initial posterior decompression surgery. The solid lines depict the pre- and post-operative clivoaxial angles (CXAs) (a) Postoperative imaging demonstrating persistent syrinx and kinking of the cervicomedullary junction despite adequate suboccipital decompression. Note the low CXA (95 degrees). (b) Following anterior decompression via endoscopic endonasal odontoidectomy, symptomatology and radiographic findings improved, with a postoperatve CXA of 135 degrees. Solid lines depict the pre- and post-operative clivoaxial angles (CXAs).

70  Revision suboccipital decompression for complex Chiari malformation

PRINCIPLES OF REVISION SURGERY The principles of revision surgery for failed suboccipital decompression follow similar principles as other revision surgeries in orthopedics and neurosurgery, in that normal anatomy should be identified first and then used as a guide to slowly identify previous surgical margins and persistent abnormal anatomy. Often, skin incisions needs to be extended to find normal anatomic margins. Identification of the C2 spinous process, which in most cases is preserved on initial surgeries, can serve as a relative indicator of where the C1 laminectomy site resides within scar tissue and allows the surgeon prevent inadvertent durotomy and spinal cord injury. Likewise, the superior aspect of the incision is usually superior to the cranial aspect of the previous craniectomy site, allowing normal occipital skull identification prior to working caudally toward the craniectomy site.

PREOPERATIVE PLANNING AND OPERATING ROOM (OR) SETUP Preoperatively, clinical assessments should be thoroughly documented for baseline status, and repeat imaging should be reviewed. Typically, a number of preoperative imaging studies can be formed. As mentioned previously, flexion/extension x-rays can help exclude the possibility of CCI, in which case occipitocervical fusion may also be required in addition to or in lieu of repeat (a)

suboccipital decompression. Magnetic resonance imaging (MRI) of the brain and cervical spine without contrast can assess the extent of persistent posterior fossa compression. Contrast studies can reveal scar tissue, and high-resolution T2 imaging, including Fast Imaging Employing Steady-state Acquisition (FIESTA) sequences, may help define adhesions, webs, or other intradural pathology contributing to deranged CSF flow. In cases of suspected idiopathic intracranial hypertension, ICP monitor placement is an additional study to consider, and in intracranial hypotension, computed tomography (CT) or MRI myelogram to look for possible causative CSF leaks should be discussed. Noncontrast CT scans can often better appreciate the degree of bony decompression originally performed. A corollary to this in very young children is the well-recognized phenomenon of bony reossification of the dura or ligaments in the posterior fossa, leading to the return of symptoms. Our group has recently adopted the use of high-resolution thin CT scans that can be translated to threedimensional (3D) models. These models can serve as excellent academic resources to help plan out the exact measurements of any additional craniectomy required intraoperatively, aid in patient/parent explanations, and help demonstrate the relation of the cranium to cervical spine for patients requiring occipitocervical fusions (Figure 10.3). Patients are brought into the OR supine. They should be intubated and have all intravenous (IV) (b)

Figure 10.3  3D-printed models of the subocciputal and upper cervical spine in a patient who failed initial suboccipital decompression and C1 laminectomy surgery for Type 1 Chiari malformation. (a) Dorsal/­ posterior view, which demonstrates an inadequate suboccipital craniectomy and excessive titanium mesh cranioplasty, which were contributing factors for representation. The C1 laminectomy was adequate. (b) Cranial to caudal view, again demonstrating the original foramen magnum plus minimal craniectomy. Note the dens is in the normal vertical position relative to the skull base, and thus ventral compression was not a precipitating factor.

Operative technique  71

lines placed in this position, including electrophysiological monitoring of motor-evoked potentials (MEPs) and somatosensory evoked potentials (SSEPs) if CCI is suspected and an occipitocervical fusion is planned. A Mayfield head clamp is applied to a pressure of 60–70 PSI. Patients are then gently log-rolled prone onto a flat OR table with chest and hip bolsters. The arms are tucked at the sides and kept in place by a sheet rolled and clamped onto the back. Care should be taken to pad all pressure points of the arms and any bulky anesthetic equipment. Padding should also be placed under the knees and the legs kept slightly flexed and supported. The head is flexed about 30 degrees, making sure that at least two fingers can fit between the chin and the chest and that there is no abnormal elevation in peak airway pressures. The shoulders can be taped down to increase skin turgor around the incision site; however, care should be taken to prevent brachial plexopathy. Suction tubing and monopolar and bipolar cautery cables are usually taken down to the legs to maintain sterility and ease of use. Depending on the size of the approach or previous incision, typically the hair is cut in a 5-cm width from the inion to the C2 spinous process protuberance. The wound is prepped with iodine wash or gel rather than chlorhexidine to minimize the risk of chlorhexidine dripping into the eyes in the prone position, which can cause retinal scarring and blindness. The operative microscope should be set up in a face-to-face orientation.

OPERATIVE TECHNIQUE After the patient is positioned and draped, the original incision is opened. Based on the goals of revision surgery, this incision can be extended as needed. As mentioned previously, identifying normal anatomy is performed first. At the cranial aspect, normal occipital bone is identified. A small two- or three-prong Weitlaner retractor is placed. Gentle retraction can help identify the midline or natural planes of the deeper soft tissues. Monopolar cautery can be used carefully to open the remainder of the superficial tissue. Once fascia is identified, extreme caution needs to be taken with monopolar cauterization to avoid incidental durotomy. At the caudal aspect, the spinous process of C2 is identified and superior aspect of the

C2 lamina is exposed, taking care not to unroof the C1/2 joints bilaterally, which can result in instability. Single- or double-cerebellar retractors are then placed for deeper and wider retraction and visualization. In the suboccipital region, dissection of the bony interface with dura/scar is accomplished using a combination of Penfield dissector, Woodson dissector, and straight or upgoing curettes. Once a clear plane has been defined around all regions of the prior craniectomy site, the drill is brought in. We typically prefer the use of a 4-mm round cutting burr for this portion of the case. The predetermined amount of additional bone requiring resection is thinned with the drill. The assistant can place a Penfield 1 or malleable retractor under the bone and above the dura to provide additional protection during drilling. Once the bone had been thinned enough in all directions, Kerrison rongeurs can be used to complete the bony resection (Figure  10.4a–d). The edges should be smoothed out and waxed to prevent dural injury when retracting the dura. If the C1 laminectomy needs to be widened, this is accomplished with careful subperiosteal dissection using a Penfield 1 instrument. Curettes can also be used to clear off the inferior portion of the lamina. Once freed up, Kerrison rongeurs can be used to widen the laminectomy. In portions where the bone is thick, a matchstick drill can be used to thin down the bone, followed by resection using the Kerrison rongeur. In some cases, additional partial C2 laminectomy may be required. Any bleeding bone edges should be addressed with bone wax, and any epidural bleeding can be packed with hemostatic matrix (e.g., FloSeal) or thrombin-soaked gelfoam and cottonoid patties. Thorough irrigation should be completed, making sure that all bone dust is washed out prior to opening the dura, as well as hemostasis to prevent postoperative aseptic meningitis. At this point, the operative microscope is brought in. Using a Woodson dissector, planes of the dura or scar are identified. Any bands or sheets of tissue that can be dissected inferiorly are done with the Woodson, above which the assistant sharply incises with a 15-blade scalpel. In rare situations, it is not possible to identify native dura below the scar. In this case, we advise using this entire width of tissue as essentially the new native

72  Revision suboccipital decompression for complex Chiari malformation




Cr an










u Ca


Figure 10.4  Microsurgical technique for extradural exploration and suboccipital craniectomy extension during a revision surgery for Type 1 Chiari malformation. Note that the orientation is the same for all figures as depicted in panel (a). (a) A straight curette is used to scrape soft tissue and scar from native dura. (b) A straight curette is used to dissect out the native dura from the previous craniectomy edge. (c) A Penfield 1 instrument is used to protect the dura during drilling for craniectomy extension. (d) The craniectomy is completed using Kerrison rongeurs. (e) Any remaining soft tissue or scar tissue bands across the dura are dissected using a slotted Woodson dissector and sharply incised with a 15-blade scalpel. (f) A 15-blade scalpel is used to perform the double-Y-shaped durotomy in a layer-by-layer fashion.

dura. A slotted Woodson dissector with a 15-blade scalpel are used to cut a double-Y-shaped opening in the dura (Figure 10.4e,f). Care must be taken while retracting the dural leaflets because previous scars may avulse vessels on the surface of the cerebellum or in the subarachnoid space, depending on the degree of prior exploration. Bipolar cautery, microscissor sharp dissection, and combinations of different Rhoton instruments can be used to ensure that the cerebellar surface is not injured during dural release and retraction. The leaflets are tacked up using 4–0 nylon sutures. At this point, intradural exploration should commence. Any arachnoid adhesions should be sharply excised using microscissors or an arachnoid knife. In situations where the cerebellar tonsils have extreme inferior migration into the upper

cervical canal, tonsillopexy can be performed. Bipolar cautery on the dorsal surface of the tonsil while protecting the medulla is used to retract this tissue without clinical consequence. A side-to-side inspection of the cerebellar hemispheres should be performed to continue resecting any arachnoid adhesions. Once complete, the cerebellar tonsils should be carefully dissected and spread at the midline, where one continues to cut adhesions as needed. It is important to visualize the fourth ventricle with the tonsils retracted. CSF pulsatile flow consistent with inspirations should be noted, which confirms adequate CSF flow in this region. Sometimes adhesions are not encountered until deep in this area approaching the fourth ventricle, so careful inspection needs to be performed, and if necessary, a web or veil of arachnoid covering

Operative technique  73






ud a












a Cr

Figure 10.5 Microsurgical technique for intradural dural exploration during a revision surgery for Type 1 Chiari malformation. Note that the orientation is the same for all figures as depicted in panel (a). (a) Intradural exploration with lysis of adhesions tethering the cerebellar tonsils and hemispheres to the inner layer of the dura using microscissors. (b) Dural edges tacked up and the descending cerebellar tonsils are directly visualized. (c) Tonsillopexy is performed, shrinking down the size of the cerebellar tonsils with bipolar cautery. (d) Exploration of the fourth ventricle ensuring no adhesions and good pulsatile CSF flow. (e) Final view of the posterior fossa following completed tonsillopexy and lysis of adhesions. (f) Watertight dural closure with allogeneic dural substitute.

the obex can be identified and lysed, often in cases with unresolved syringomyelia (Figure 10.5a–e). Once completed, thorough irrigation is performed and hemostasis achieved. The choice of duraplasty at this point is at the discretion of the surgeon and based on anatomical factors. Autologous pericranium is preferred and can be locally harvested from an extension of the incision cranially or a separate incision closer to the vertex; in cases where this is unsuccessful, an allogeneic dural substitute can be used (e.g., bovine pericardium and cadaveric skin). An appropriatesized piece is cut and tacked using interrupted 4–0 sutures. Then each leg of the graft is sutured in a running fashion to achieve a watertight dural closure (Figure 10.5f). Prior to completing the last suture, the intradural space should be irrigated to

ensure that there is no bleeding requiring reexploration and to reinsufflate the subarachnoid space. For any areas where there is persistent CSF leak, small muscle grafts can be harvested and sutured in place. Fibrin sealant can be used over the entirety of the dural closure for additional support. A dural substitute such as DuraForm or compressed gel foam may be placed over the dural closure, but we do not favor this approach. Similarly, small mesh cranioplasty can be utilized if needed for cosmesis following revision of aggressive decompressions, but more importantly, to provide structural support of the deep soft tissue, prevent scarring of muscle on the dura, and minimize postoperative pain. The deep tissue and superficial tissues are closed with interrupted sutures using 0, 2-0, and 3-0

74  Revision suboccipital decompression for complex Chiari malformation

absorbable sutures. A running 3-0 caprosyn or nylon suture is used for skin closure. Drains are typically not placed to prevent the CSF-cutaneous fistula.

POSTOPERATIVE MANAGEMENT Postoperatively, patients should be observed in an intensive care unit (ICU) setting for hourly neurologic and vital monitoring checks. Systolic blood pressure goals should remain less than 150 mmHg for adults or age-appropriate normotension for pediatric patients. Any coagulopathies should be immediately addressed. Careful monitoring of the wound for clear fluid leaking should be undertaken. The head of bed should be elevated to 30 degrees or higher. For patients with preexisting dysphagia, a formal swallow evaluation should be performed by speech-language therapists prior to initiating diet to prevent aspiration. Steroids (dexamethasone) should be initiated for the prevention of aseptic meningitis, which can occur from blood and bone dust irritation in the subarachnoid space. Most patients can initiate physical therapy and deep vein thrombosis (DVT) chemoprophylaxis on the first postoperative day. For patients with stable neurologic examination, further inpatient imaging is usually not required. Pain control is the primary issue that should be addressed. At our institution, we have instituted a postoperative pain protocol for all adult patients undergoing first-time or revision Chiari decompression surgeries, beginning with methadone in the OR and transitioning to hydromorphone patient-controlled analgesia (PCA) in the ICU. It is important that this regimen is instituted only on those with a reliable neurologic examination. A low-dose muscle relaxer such as cyclobenzaprine or diazepam is started as a standing dose and can be titrated up or given additional doses as needed. These medications should be held for any excessive sedating side effects. The PCA narcotic requirement should be transitioned to an oral regimen on postoperative day 1–2 as tolerated. Standing Tylenol (hold for patients with preexisting hepatic pathology), celecoxib (hold for patients with renal pathology and start 24 hours after surgery), and neuropathic pain medications (such as pregabalin and gabapentin) should also be initiated. Patients on narcotics should be discharged on a bowel regimen consisting of at senna and/or colace.

COMPLICATIONS The postoperative complications that may arise vary based on the exact revision surgery performed. For revisions with or without duraplasty, there is a risk of superficial or deep wound infections. Superficial infections can usually be treated with observation and oral antibiotics; however, deep or progressive infections may require surgical debridement and IV antibiotics to prevent epidural abscess formation, which can compress the posterior fossa structures. Patients that undergo duraplasty are at risk for pseudomeningocele formation, which can progress to gross CSF leaking. In these cases, a CT head should be obtained to assess for underlying hydrocephalus, in which case CSF diversion can be used to treat both conditions. Small pseudomeningoceles can remain contained and spontaneously resolve with time if the fascial closure is tight. However, as the collections grow, that can exert pressure on the closure layers, impairing wound healing and leading to superinfections and potentially meningitis. This is especially of concern when there is obvious CSF leaking from the incision itself. Conservative measures such as oversewing the wound are sometimes sufficient for treatment. However, in other situations, the collection may need to be percutaneously drained, or a lumbar drain may need to be placed to promote wound healing. If these measures fail to work, then exploration and identification of the leak site may be required. For revision surgeries that involve subarachnoid exploration, retraction injury or iatrogenic contusions can lead to edema and potentially infarct of the cerebellar tissue. In a similar regard, vessel or parenchymal injury can lead to bleeding complications and hematomas, which can exert a mass effect on surrounding structures. In these situations, reexploration with potential widening of the suboccipital craniectomy may be required for adequate decompression.

Pearls and Pitfalls ●●

Given the unpredictable interfaces of hard, bony edges with scar/dura/duraplasty, monopolar cautery should be limited when approaching these regions during revision surgeries. Alternatively, manual blunt

Complications 75


and sharp dissection with combinations of Penfield, Rhoton, Woodson, or other instruments should be used to avoid thermal injury, which can result in durotomy and neural injury even if not directly applied to tissue. MRI CSF flow sequences (e.g., CINE [short for cinematographic] or high-resolution T2 sequences) can demonstrate abnormal CSF flow through the foramen magnum, confirming persistent compression at that level. In some cases, when clinical history suggests worse symptoms when in extreme flexion or extension, neutral-alignment MRI images may not demonstrate abnormal CSF flow. In these situations, flexion/extension MRIs can be considered, in which medullary compression and resultant CSF flow dynamics can be assessed in patients during these alignment changes.



Additional craniectomy is required in some revision cases for Chiari decompression. Preoperative MRI should be carefully assessed to exclude a low-lying torcular, which can inadvertently be injured during bony resection or durotomy. 3D CT reconstruction has proven useful to counsel ­families, plan surgery, and teach in the OR. Ultrasonography can be a useful adjunct during revision surgeries prior to dural opening. This imaging can demonstrate the extent to which additional craniectomy is required, how low the cerebellar tonsils are relative to the new planned dural opening, the determination for whole or partial resection of C2 lamina, and the assessment of CSF flow in the subarachnoid space due to the presence of MRI occult adhesions.

11 How to revise a failed occipital cervical fusion JOSHUA T. WEWEL, MAZDA K. TUREL, JOSEPH E. MOLENDA, AND VINCENT C. TRAYNELIS Indications Relative contraindications Expectations Preoperative planning and operating room (OR) setup

77 78 78

Operative technique Postoperative management Complications

79 81 82


INDICATIONS The cranial-vertebral junction (CVJ) is a complex structure consisting of the occiput, atlas, and axis, along with an array of ligaments that allow mobility and stability of the articulating surfaces. The most common sources of pathology leading to instability at the CVJ include congenital, infectious, trauma, rheumatoid arthritis, malignancy, and iatrogenic factors. CVJ instability is frequently secondary to trauma such as atlanto-occipital dislocation (AOD), type III occipital condyle fractures, complex atlas/axis fractures, and ligamentous instability. The congenital anomalies of the CVJ that produce instability are invariably complex lesions that often require specific and unique strategies to correct and stabilize. Inflammatory processes such as rheumatoid arthritis can result not only in translational and axial instability, often in conjunction with deformity requiring fusion, but also decompression at times. Anterior instrumented fusion of the CVJ has been reported, but these

procedures are very rarely performed, as most fixation and fusion of this segment are accomplished via a dorsal procedure. The CVJ is a challenging region to fuse. There is great sagittal and axial plane rotation across the segment, and these motions must be minimized for a successful arthrodesis to occur. This dynamism places increased stress on the fixation. Fixation purchase points may be limited by factors such as a large suboccipital craniectomy, fractured vertebra, anomalous vascular anatomy, and poor bone quality. Every surgeon who actively manages pathology of the CVJ will encounter patients with failed fusions across this segment. A nonunion at the CVJ may be detected by hardware failure, resorption of bone graft, lucencies in bone surrounding screws, or overt instability, as detected by flexion/extension lateral radiographs. Nonunions that are symptomatic in terms of producing pain and those that are related to instability should be surgically addressed. Other indications for revision are infection and prominent hardware. 77

78  How to revise a failed occipital cervical fusion

It should be noted that this chapter does not include any discussion of revision to correct deformity that persisted after the primary procedure or that was created by the primary procedure. A number of the points presented here may be valuable in these cases, but these are very rare and unique events, each of which requires special, highly i­ ndividualized management.

RELATIVE CONTRAINDICATIONS The relative contraindications for revision of a CVJ nonunion are rare, given the fact that the primary procedure was performed for pathology that mandated a fusion. A few patients may be asymptomatic and have radiographs that show no significant instability; these individuals can be followed without surgery. Additionally, a small subset may have developed serious comorbidities that would preclude general anesthetic, which would make surgery unreasonable. The remainder of patients with CVJ nonunions should be offered surgical intervention. The senior author has not seen an external bone growth stimulator lead to a fusion in any patient who has a nonunion (CVJ and subaxial), and therefore, this is not recommended.

EXPECTATIONS Patients who were doing well initially following the primary procedure can expect to return to that state if the reoperation is successful in obtaining a solid arthrodesis. Those with instability also should expect benefits from revision surgery in terms of protecting the neural elements, as well as eliminating or decreasing pain if that is part of the symptomatology. Primary fusion rates when using rigid fixation have been reported to be as high as 90% by multiple authors. The best opportunity for success is with the first operation, so those needing revision surgery should understand that there is risk of persistent nonunion. Postoperative wound infections that occur in the early period after surgery do not require removal of the instrumentation. Those that are substantially delayed are more prone to be associated with the formation of biofilm, and removal of the fixation hardware is usually necessary to

e­ radicate the infection. If the fusion is not mature, then new instrumentation may need to be implanted. Patients should expect that the infection can be cured. Prominent hardware at the CVJ on very rare occasions will produce pain. Judicious removal of proud instrumentation in select cases of patients who have already fused should result in decreased pain. If the instrumentation is exposed, then it will need to be removed to prevent (or treat) infection. Nonunions will often require revision of the fixation. The previously placed instrumentation should be carefully examined, and often only simple modifications are necessary. If screws have become loose, then larger screws can be placed or screws can be positioned in different locations. For example, a loose screw in the pars interarticularis of C2 could be replaced with a largerdiameter screw, or C2 fixation could be achieved by putting a new screw in the pedicle. Loose cranial fixation can be managed in the same fashion. Often, cranial plates minimize the number of screws that can be placed and the options for where they are placed. If cranial fixation needs to be reestablished, it is often prudent to use multiple single-screw/rod fixators rather than a plate. Not only does this increase the options for fixation, but such a technique usually allows more of the occiput to be uncovered, which provides a larger surface area for fusion substrate to be placed. If there has been a rod fracture, then rods of greater diameter or one constructed with a stronger material should be considered. In rare occasions, a third rod could be placed. Graft selection is an important, yet often overlooked part of surgical planning. Broadly, two choices exist—autograft or allograft. Allograft is well known to heal very poorly under tension, and if that was what was utilized in the primary procedure, it is the most likely cause for failure. In such cases, it is imperative to employ another grafting strategy. We prefer to use a rib autograft, which nicely matches the curvature of the CVJ in all cases. The senior author has not seen great success with calvarial autograft, although it is recognized that others have reported good results with this bone source. Iliac crest is an excellent source of autograft. If allograft is chosen, then it should

Operative technique  79

be used in conjunction with bone morphogenetic protein (BMP). This is an off-label use of BMP, but it will have a good chance of producing a successful fusion.

PREOPERATIVE PLANNING AND OPERATING ROOM (OR) SETUP The previous section includes information that is important for preoperative planning, and these issues will not be repeated. It is important to carefully assess the airway. Those patients with prior CVJ fusions will usually have limited motion in this region, even if there is a nonunion, and that can make intubation difficult. Certainly, those with instability will require extreme care in terms of minimizing motion during intubation. Once intubated, all revision patients are fixed in position using the Mayfield three-point skeletal fixation device and carefully turned prone. We use neuromonitoring in a very selective fashion. If there is instability or significant neural compression, monitoring is critical for the positioning phase. Baseline studies are obtained and repeated rapidly following final positioning. Prompt radiographic evaluation of the CVJ following the turn to prone position is also very important to minimize neurologic injury in the unstable patient. It is a quicker assessment than can be achieved with electrophysiological monitoring, and if the region is in normal alignment, the risk of creating a neurological deficit based on patient position approaches zero. Radiographic evaluation of alignment in patients without instability or compression is used as the sole means of monitoring, and electrophysiological studies are not used in this group. Neuromonitoring is of no value after the position has been determined because the CVJ will not move if using the Mayfield. This is different than in some primary procedures, where deformity correction is part of the surgical plan. When there is no need for deformity correction, careful and precise surgical technique will not result in a neurological deficit. Poor operative execution can result in deficit, but in this case, monitoring is of no value because it will not prevent injury from events such as inadvertent striking of the cord. Electrophysiological monitoring can alert one to

risk due to hypotension. The more serious factor in these cases is the anesthesiologist—monitoring is no replacement for a skilled, experienced, and vigilant anesthesiologist. Intraoperative image guidance is a very useful adjunct in those cases where screw fixation needs to be changed. This is particularly important when nonstandard fixation is desired, such as in the occipital condyle or calvarial diploe. It can also be helpful even in standard C2 pedicle screws, as the surface anatomy may be altered by the prior surgery. Use of an image guidance system that acquires the data with a radiographic assessment after positioning will minimize the errors of ­manual registration; thus, it is preferable. If there is uncertainty about the type of instrumentation that is implanted, a universal cervical instrumentation removal set should be available.

OPERATIVE TECHNIQUE Once positioned and sterilely prepped, the original incision is reopened. Great care is required in the exposure if there has been a previous decompression. Firm avascular scarring will inhibit the exposure at times because it can be exceedingly difficult to retract. Resection of the medial portion of scars in these select cases will improve exposure. If the approach results in a dural violation, then all possible maneuvers to obtain a primary dural closure should be employed. Reliance on the many fibrin glue adjuncts and dural onlay grafts to prevent the development of a pseudomeningocele without primary closure will often d ­ isappoint both the patient and the surgeon. All the previously placed instrumentation and the posterior occiput must be fully exposed. Avascular, scarred encased graft material should be removed. Any necessary decompression should be performed once the exposure is completed. At this point, attention is directed to the instrumentation, and poor screw purchase is corrected by either using larger-diameter screws or placing screws in new and different trajectories. Only a single option is available for C1, but C2 can be fixed with screws in the pars, pedicle, or lamina. If one chooses the C2 lamina, then it is best to extend the fixation at least one level caudally, as these screws

80  How to revise a failed occipital cervical fusion

are in line with axial rotation and thus are biomechanically disadvantaged to minimize movement in that axis. Limitation of fixation points in C2 can be overcome by extending the fusion inferiorly, but this is not possible if occipital fixation has failed. Changing individual screw fixation, as described earlier, is a good strategy. This may not be possible, though, if there has been a large suboccipital craniectomy. In these cases, establishing good cranial fixation can be challenging. Two potential targets for screw placement are available: the occipital condyle and the diploe. Occipital condyle screw placement was first described about 10 years ago. Biomechanical studies suggest that it provides as strong a point of fixation as the occiput itself, but these studies consisted of only acute testing, and it is possible that with time, the purchase will become compromised. This is particularly worrisome because the condyle is so close to the axis of rotation of occiput–atlas sagittal plane motion, which places it at a biomechanical disadvantage. The condyle is relatively deep in the exposure, and the condylar canal, through which traverses the hypoglossal nerve, is in the anterior portion of this structure. Despite these challenges and concerns, a number of surgeons have reported good results with this fixation, and although it should not be considered a primary technique, it has great value as a secondtier point of purchase. The technique often requires skeletalization and mobilization of the V3 segment of the vertebral artery (VA). The VA is traced along the condylar fossa until the condylar foramen and emissary vein are identified, delineating the lateral

extent of condyle exposure. The boundaries of the condylar fossa are defined laterally by the condyle emissary vein, inferiorly by the lateral portion of V3, and medially by the junction of the condyle and ­occipital bone. The entry point for condyle screw placement is 5 mm lateral to the posteromedial edge of the condyle and 2 mm inferior to the skull base floor (Figure  11.1). A pilot hole is drilled with a 10–33-degree medial angulation and 10–30 degrees caudally, attempting to pass the screw along the longest axis of the condyle with the goal of staying parallel to the skull base floor (Figure  11.2). Preoperative evaluation of the occipital condyle dimensions will determine the length of the screw placed, bearing in mind that 11–14 mm of the unthreaded portion of the screw will need to remain proud, such that it can reach the rod projecting from the C1, C2 instrumentation (Figure  11.3). It is recognized that there is great variability in the angular drilling parameters, and therefore, it is optimal to utilize intraoperative navigation to assist in achieving optimal screw placement in the condyle. Condylar screws do not block the occipital surface, which increases the area available for fusion, which may be very important if there has been a craniectomy. The other potential technique if there has been a large posterior fossa craniectomy is to place screws into the diploe of the calvarium (Video 11.1). This is akin to iliac crest screw placement. Guide holes are made through the diploic space, and screws are placed. The screw diameter should be tailored to each patient such that cortical bone is engaged yet not violated by screw threads. The

Figure 11.1  The appropriate starting point for the condyle screw on a sawbones model.

Postoperative management  81

Figure 11.2  The intraoperative planning using navigation for the placement of the condyle screw.

Figure 11.3  The successful placement of a condyle screw utilizing navigation.

diploic screws can then be fixed to the cervical instrumentation. The procedure requires careful preoperative planning and is best performed using image guidance.

POSTOPERATIVE MANAGEMENT Following the completion of the procedure, the patient is turned supine and extubated. Upright

anterioposterior (AP) and lateral cervical spine x-rays are obtained as an inpatient to establish baseline films. Cervical radiographs are then obtained at the 3-, 6-, and 12-month follow-up visits. A computed tomography (CT) scan of the CVJ is usually performed 1 year after surgery to verify fusion. CT is necessary because the hardware ­frequently obscures the view of the fusion mass on plain radiographs.

82  How to revise a failed occipital cervical fusion

COMPLICATIONS When occipital screws are being replaced, care must be taken to choose appropriate sites for new screw placement, preferably below the superior nuchal line along the occipital ridge. If screws must be placed superior to the superior nuchal line, dural venous structures and bleeding may be encountered. If bleeding is encountered during screw placement, it is best to continue with placement of the screw. Minor penetration of a screw into a venous sinus does not appear to cause any harm. The real risk of occipital screws is a posterior fossa hematoma, which can occur following a minor cerebellar surface vein or artery injury. Any patient who does not awaken promptly from surgery should be immediately evaluated with a CT, and if there is a posterior fossa hematoma, it should be evaluated absolutely as quickly as possible. The management of CSF leaks has been previously addressed. A VA injury may occur, and if this happens, all possible maneuvers to control the hemorrhage rapidly should be undertaken. Ideally, the vessel is preserved, but if this is not possible, the risk of neurological deficit with the loss of one VA is low. If the VA injury occurs on the first side being treated, then all possible efforts to preserve the contralateral VA must be undertaken. In some instances, this will mandate not operating on the patent VA side.

Pearls and Pitfalls ●●




●● ●●

The time invested in studying the images preoperatively is well spent. These patients should have plain radiographs, magnetic resonance imaging (MRI), and CT. There must be a clear understanding of the anatomy prior to surgery. Verify that the tools to remove the previously implanted instrumentation are available if hardware removal is necessary. Do not repeat exactly what was done before. The best opportunity for success lies with the first surgery, so try to determine why it failed and adjust accordingly. Consider overbuilding the construct. Additionally, if the patient’s age and body habitus permits, consider halo immobilization for 8–12 weeks, even if instrumentation is placed. Use image guidance when incorporating any nonstandard fixation. Do not rely on allografts alone.

Video 11.1 Failed OC Fusion (

12 How to revise a failed C1–C2 fusion NIZAR MOAYERI AND MICHAEL G. FEHLINGS Indications Relative contraindications Expectations Principles of revision surgery Preoperative planning and operating room (OR) setup

83 84 84 84

Operative technique 86 Postoperative management 86 Complications 87 Reference 88


INDICATIONS C1–C2 revision surgery is a technically demanding procedure, as normal anatomic landmarks are obscured, as well as the presence of vital organs: the transition from medulla oblongata to spinal cord and the vertebral artery (VA), with its complex geometric anatomy. A misplaced screw can result in devastating neurologic and vascular complications. A variety of surgical techniques have been utilized for revised internal fixation to treat ongoing atlantoaxial instability, with failure rates ranging from 5%‒50%. Patients with rheumatoid arthritis and os odontoideum have particularly high rates of nonunion after C1–C2 fixation, necessitating a revision surgery. The indication for revision of C1–C2 fusion is nonunion with persistent or progressive symptoms, due to pseudarthrosis. The overall incidence of atlantoaxial pseudarthrosis after atlantoaxial fusion is estimated to be around 2%–10%. However, the true incidence may be higher because it can be oligosymptomatic or asymptomatic. Main symptoms may include, but are not limited to, persistent pain, progressive or limiting neurological compromise, and instability. A wide range of other specific indications for this surgery also exist,

depending on patient and/or physician preferences and radiological ­findings. Examples are nonunion of odontoid fracture with development of odontoid granuloma or retro-odontoid pannus or progressive C1–C2 dislocation. In addition, pseudarthrosis following C1–C2 fusion is frequently seen after wire fixation, clamps or hooks, unilateral transarticular screw with cable fixation, or onlay bone graft with no internal fixation. There are several biomechanical reasons for C1– C2 nonunions. The C1–C2 segment has the widest range of motion of any spinal motion segment, and this motion is increased significantly when there is pathological instability present. The difficulty of achieving adequate control of C1–C2 motion during bone healing has led to a number of strategies to improve the fusion rate. Adjunctive use of a halo brace or internal fixation with rigid transarticular screws has been promoted to improve the success rate of surgery. Additional factors contributing to the risk of nonunion include malnutrition and the use of steroids, nonsteroidal anti-inflammatory drugs (NSAIDs), and cytotoxic and immunosuppressive drugs. Two independent risk modifiers include smoking and osteoporosis. 83

84  How to revise a failed C1–C2 fusion

RELATIVE CONTRAINDICATIONS A patient’s age and/or comorbidities may pose a relative contraindication as to what invasive technique can be used or whether an internal or external revision should be performed. The screw-rod constructs technique for posterior atlantoaxial fusion using C1 lateral mass screws and C2 pars/ pedicle screws, as described by Goel and Laheri and later modified by Harms and Melcher, offers excellent access to the C1–C2 joint and allows decortication and packing of the joint with graft material. However, the use of transarticular screws involves the alignment of C1 on C2, which can be challenging in obese patients or those with significant thoracic kyphosis. To capture both C1 and C2, adequate reduction in case of anterior atlantal subluxation is mandatory before hardware insertion. Moreover, many patients with fixed subluxations or absent or fractured posterior elements of C1 and/or C2 cannot be treated with the transarticular screws technique, making the screw-rod construct the technique of choice for C1–C2 revision surgery.

EXPECTATIONS As C1–C2 revision surgery is one of the most complex procedures in spine surgery, the performing surgeon must be aware of her or his own primary end point and have an elaborate, detailed discussion with the patient as to what expectations are upon agreeing to revision surgery. A thorough patient history and physical examination will be the first step in establishing the patient’s needs. Questions that should be asked include the following: ●● ●●




Why did you have your initial procedure? What symptoms were you having before your initial procedure? Following the initial procedure, did you get relief from some of or all your symptoms? If so, how long did this relief last? Are the symptoms that you are having now similar to those you had before your initial procedure? If not, how are the symptoms different? Are the symptoms related to a certain position of the head or neck, and does that change while sitting or lying down?

These questions will give the spine surgeon some sense of whether the initial problem was successfully treated and whether the current symptoms represent persistence of the initial problem, recurrence of the initial problem, or a new problem at the same or an adjacent level. Questions regarding constitutional chills, nausea, symptoms (i.e., fever, vomiting, unexplained weight loss, fatigue) should also be addressed during the history-taking to investigate whether problems such as infection or tumor may be present. Questions pertinent to the nature, duration, severity, and location of pain, numbness, and/or tingling, as well as questions relating to weakness, problems with balance and fine motor skills, and bowel and bladder function are also essential, as is the case when assessing any spine patient. Red flags such as progressive weakness, constitutional symptoms, unrelenting pain, weakness or numbness while the head is moved to a certain position, and difficulty breathing or gasping for breath without any physical effort are suggestive of an urgent (or even emergency) situation.

PRINCIPLES OF REVISION SURGERY The majority of C1–C2 pseudarthroses occur between C1 and the graft and are associated with failed hardware. Atlantoaxial instability due to failed surgery can be successfully salvaged if the pathological, pharmacological, biomechanical, and technical problems are effectively resolved. Careful exploration of the indications for surgery through an extensive history and careful management of the patient’s expectations is key in the decision-making phase for revision surgery. Preoperative understanding of the altered anatomy through imaging should be used to guide the spine surgeon in choosing the appropriate technique. Fusion can be achieved in a substantial proportion of revision surgeries. The greatest success can be gained by using autologous bone grafts, adequately controlling C1–C2 motion (in our view, using the C1 lateral mass screw and the C2 pars interarticularis screw is the optimum technique for technical and biomechanical reasons highlighted later in this chapter), compressing the bone graft between the arches of C1–C2, meticulously preparing the fusion bed, and optimizing the pharmacological and metabolic factors for promoting bone fusion.

Preoperative planning and operating room (OR) setup  85

PREOPERATIVE PLANNING AND OPERATING ROOM (OR) SETUP Imaging Appropriate imaging guides the spine surgeon in preoperative planning. Imaging techniques that must be present to evaluate a patient for C1–C2 revision surgery include plain radiography, computed tomography (CT) scan (with or without angiography), and magnetic resonance imaging (MRI). Plain radiography should typically include open-mouth view, anteroposterior, lateral, and flexion/extension views, if deemed safe and clinically possible. The status of an existing fusion should be assessed for the presence of bridging trabecular bone or continued motion. The location of instrumentation, along with any subtle loosening of existing screws in the form of haloing and implant failure in the form of screw pullout and screw and/or rod breakage, should be noted. Catastrophic failure of C1–C2 implants can be seen in the setting of trauma, tumor, or infection. Pseudarthrosis should also be considered when subtle or overt signs of instrumentation loosening or failure are noted on radiographs. Flexion and extension lateral cervical radiographs are helpful in assessing pseudarthrosis and instability. Flexion and extension radiographs may also show the movement of loosening screws and angular motion that may suggest pseudarthrosis and/or instability. CT scans are commonly used as the imaging modality of choice to assess a patient’s possible pseudarthrosis. Coronal and sagittal reconstructions are particularly helpful in assessing the fusion as well as the position and status of the C1–C2 instrumentation. Bridging trabecular bone in these areas indicates a solid fusion. Lucency that is typically linear in nature indicates a pseudarthrosis. Lytic bony destruction may be seen in cases of infection or tumor. In particular, when planning for a C1–C2 revision fusion, the status of the bone should be noted. Significant bony destruction precludes the placement of instrumentation and can alter the surgical plan, often leading to the involvement of the occiput and/ or C3 or placing the lower distal of the instrumentation into areas of preserved bony anatomy. CT angiography is important in planning a C1– C2 fusion revision, especially when the anatomy

is altered or anatomic variations involving the VA between C1 and C2 are present. In particular, V3 segment anomalies, which occur in up to 10% of the population, should be carefully studied. These include the persistent intersegmental artery, in which the VA courses abnormally below the C1 arch after leaving the transverse foramen of the C2 and enters the spinal canal without passing through the C1 transverse foramen; a VA fenestrated at the atlas level; and the posterior inferior cerebellar artery, originating from the VA between C1 and C2 and entering the spinal canal from the caudal side of C1. In addition, the presence of a dominant VA must be noted to avoid introducing any extra risks during screw placement. The presence of a high-riding VA (i.e., one that is more cranial than usual and medially located in the body of C2) occurs in up to 23% of patients and may increase the risk of vertebral artery injury (VAI). MRI provides great detail of the soft tissue structures, including spinal cord compression and epidural fluid collections that can signal infection and tumors. MRI can be performed with and without gadolinium to help differentiate recurrent disease and/or fluid collections from scar tissue. Scar tissue is vascular, and therefore will be enhanced on images after the intravenous (IV) administration of gadolinium, which has high signal intensity on T1-weighted images. In patients who cannot have an MRI, CT myelography is helpful in assessing the neurologic structures for evidence of compression. If there is a concern that the patient may have an infection in the process of working up for C1–C2 revision surgery, a complete white blood cell count with differentiation, an erythrocyte sedimentation rate (ESR), and determination of C-reactive protein (CRP) level is recommended. ESR and CRP are both nonspecific markers of inflammation. The normal concentration of CRP is 10 mg/L, which in the setting of an infection could increase from 40 mg/L to >200 mg/L. The ESR is also a measure of inflammation, with a normal ESR considered to be