Arthroscopic and Endoscopic Spinal Surgery

Arthroscopicand EndoscopicSpinal Surgery

Text and Atlas, Second Edition

Edited by

Parviz Kambin, MDProfessor of Orthopedic Surgery

and Endowed Chair of Spinal Surgery,Drexel University College of Medicine,

Philadelphia, PA

www.humanapress.com

All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by anymeans, electronic, mechanical, photocopying, microfilming, recording, or otherwise without written permission from thePublisher.

The content and opinions expressed in this book are the sole work of the authors and editors, who have warranted due diligencein the creation and issuance of their work. The publisher, editors, and authors are not responsible for errors or omissions or forany consequences arising from the information or opinions presented in this book and make no warranty, express or implied,with respect to its contents.

Production Editor: Tracy Catanese

Cover design by Patricia F. Cleary

Cover Illustrations: Figures 38A and 35B from Chapter 4, “Herniated Lumbar Disc and Lumbar Radiculopathy,” and Figures7A and 12B from Chapter 5, “Management of Discogenic Pain and Spinal Instability Using Minimally Invasive SurgicalTechniques,” by Parviz Kambin.

For additional copies, pricing for bulk purchases, and/or information about other Humana titles, contact Humana at the aboveaddress or at any of the following numbers: Tel.: 973-256-1699; Fax: 973-256-8341; E-mail: [emailprotected] or visitour website: http://humanapress.com

This publication is printed on acid-free paper.ANSI Z39.48-1984 (American National Standards Institute) Permanence of Paper for Printed Library Materials.

Photocopy Authorization Policy:Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted byHumana Press Inc., provided that the base fee of US $30.00 per copy is paid directly to the Copyright Clearance Center at 222Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license from the CCC, aseparate system of payment has been arranged and is acceptable to Humana Press Inc. The fee code for users of the TransactionalReporting Service is: [1-58829-522-2/05 $30.00].

Printed in the United States of America. 10 9 8 7 6 5 4 3 2 1

e-ISBN: 1-59259-904-4

Library of Congress Cataloging-in-Publication DataArthroscopic and endoscopic spinal surgery : text and atlas / edited by Parviz Kambin. p. ; cm. Includes bibliographical references and index. ISBN 1-58829-522-2 (alk. paper) 1. Spine--Endoscopic surgery. 2. Spine--Endoscopic surgery--Atlases.

[DNLM: 1. Spine--surgery--Atlases. 2. Arthroscopy--methods--Atlases. 3. Endoscopy--methods--Atlases. 4. SurgicalProcedures, Minimally Invasive--methods--Atlases. WE 17 A7875 2005] I. Kambin, Parviz. RD533.A785 2005 617.5'60597--dc22

2004026633

Dedication

The authors wish to dedicate this text and atlas to their families, colleagues, and studentsof minimally invasive spinal surgery.

vii

Preface

The term “minimally invasive spinal surgery” was coined in early 1990 following publi-cation of the first edition of this text entitled Arthroscopic Microdiscectomy: MinimalIntervention in Spinal Surgery, and subsequent establishment of the International Societyfor Minimal Intervention in Spinal Surgery (ISMISS) under the auspices of the InternationalSociety of Orthopaedic Surgery and Traumatology (SICOT) in April l990.

The orthopedic and neurological surgeons who participated in lectures and hands-on work-shops both in Philadelphia and abroad have witnessed the evolution of minimally invasivespinal surgery from blind nucleotomy to endoscopic fragmentectomy, decompression of lat-eral recess stenosis, foraminoplasty, and spinal stabilization.

In Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas, Second Edition, expertsdescribe and illustrate various techniques and approaches that are currently used in this field.In addition, the ongoing research for the betterment of spine care via minimally invasiveapproaches is briefly reviewed.

I would like to express my sincere appreciation to so many of my colleagues who sup-ported my efforts in the field of minimally invasive spinal surgery throughout the years.Many of them participated in our teaching symposiums and have provided valuable contri-butions to this text.

Parviz Kambin, MD

Contents

DEDICATION ............................................................................................................................ V

PREFACE .................................................................................................................................VII

CONTRIBUTORS ......................................................................................................................... XI

COLOR PLATES ....................................................................................................................... XIII

COMPANION DVD .................................................................................................................XV

1 History of Surgical Management of Herniated Lumbar DiscsFrom Cauterization to Arthroscopic and Endoscopic Spine Surgery

Parviz Kambin ............................................................................................................. 12 Arthroscopic and Endoscopic Anatomy of the Lumbar Spine

Parviz Kambin ........................................................................................................... 293 Instruments and Surgical Approaches for Minimally Invasive Spinal

Surgery Via Posterolateral AccessParviz Kambin ........................................................................................................... 49

4 Herniated Lumbar Disc and Lumbar RadiculopathyParviz Kambin ........................................................................................................... 61

5 Management of Discogenic Pain and Spinal InstabilityUsing Minimally Invasive Surgical Techniques

Parviz Kambin ......................................................................................................... 1196 Lateral Recess Stenosis of Lumbar Spine Foraminoplasty

Parviz Kambin ......................................................................................................... 1457 Role of Epidural and Radicular Veins in Chronic Back Pain

and RadiculopathyWesley W. Parke ..................................................................................................... 151

8 Diagnostic and Therapeutic Percutaneous Transpedicular Approachesto the Spine

Alexander G. Hadjipavlou, George M. Kontakis, Ioannis Gaitanisand Michael Tzermiadianos ............................................................................. 167

9 Selective Endoscopic Discectomy™: Twelve Years of ExperienceAnthony T. Yeung .................................................................................................... 205

10 Minimally Invasive Posterior Fusion and Internal FixationWith the Atavi® System

Richard D. Guyer and Terry P. Corbin ............................................................... 22711 Vertebral Augmentation for Osteoporotic Compression Fractures

Daisuke Togawa and Isador H. Lieberman ....................................................... 239

x Contents

12 Principles of Transthoracic, Transperitoneal, and RetroperitonealEndoscopic Techniques in the Thoracic and Lumbar Spine

Geoffrey M. McCullen and Hansen A. Yuan ...................................................... 25113 Use of Laser in Minimally Invasive Spinal Surgery

and Pain ManagementJohn C. Chiu and Martin H. Savitz ..................................................................... 259

14 Minimally Invasive Techniques in Pain ManagementJames Reynolds and Garrett Kine ....................................................................... 271

15 Experience With Minimally Invasive Nucleus ReplacementMichele Marcolongo, Parviz Kambin, Anthony Lowman

and Andrew Karduna ......................................................................................... 29516 Interspinous Process Implant for Treatment of Lateral and Central

Spinal Stenosis: Operative Technique and ResultsDouglas Wardlaw ................................................................................................... 315

17 Frameless Stereotactic Imaging Techniquesin Minimally Invasive Spine Surgery

Kern Singh, Laurence N. Fitzhenry, and Alexander R. Vaccaro ..................... 33518 The Rise and Fall of Chemonucleolysis

James W. Simmons, Jr. and Robert D. Fraser .................................................... 35119 Lumbar Microendoscopic Discectomy

Trent L. Tredway and Richard G. Fessler ........................................................... 359INDEX................................................................................................................................... 377

Contributors

JOHN C. CHIU, DSc, MD • Director, Neurospine Surgery Department, California SpineInstitute Medical Center Inc., Park, CA

TERRY P. CORBIN, BS • Clinical Outcomes Research Center, University of Minnesota,Maple Grove, MN

RICHARD G. FESSLER, MD, PhD • Professor and Chief, Section of Neurosurgery, Universityof Chicago, Chicago, IL

LAURENCE N. FITZHENRY, MD • Department of Orthopedic Surgery, Thomas JeffersonUniversity and the Rothman Institute, Philadelphia, PA

ROBERT D. FRASER, MD • Alamo Bone and Joint Clinic, San Antonio, TXIOANNIS GAITANIS • Orthopaedic Surgeon, University of Crete, Crete, GreeceRICHARD D. GUYER, MD • Associate Clinical Professor, Department of Orthopaedic

Surgery, UT Southwestern Medical School, Dallas, and Research Founder andChairman of the Board, Texas Back Institute, Plano, TX

ALEXANDER G. HADJIPAVLOU, MD • Professor of Orthopaedics-Traumatology, Universityof Crete, Crete, Greece

PARVIZ KAMBIN, MD • Professor of Orthopaedic Surgery, Drexel University Collegeof Medicine, Philadelphia, PA

ANDREW KARDUNA, PhD • Department of Physical Therapy, University of Oregon,Eugene, OR

GARRETT KINE, MD • Anesthesiologist, SpineCare Medical Group, Daly City, CAGEORGE M. KONTAKIS, MD • Assistant Professor of Orthopaedics-Traumatology,

University of Crete, Crete, GreeceISADOR H. LIEBERMAN, MD, MBA, FRCS (C) • Center Director and Director of Minimally

Invasive Surgery, Center for Advanced Skills Training; Department of OrthopaedicSurgery and Spine Institute, The Cleveland Clinic Foundation, Cleveland, OH

ANTHONY LOWMAN, PhD • Assistant Professor, Department of Chemical Engineering,Drexel University, Philadelphia, PA

MICHELE MARCOLONGO, PhD • Assistant Professor, Department of Materials Scienceand Engineering, College of Engineering, Drexel University, Philadelphia, PA

GEOFFREY M. MCCULLEN, MD • Orthopedic Surgeon, Neurological and Spinal Surgery,LLC, Lincoln, NE

WESLEY W. PARKE, PhD • Professor and Chair Emeritus, Department of Anatomy,University of South Dakota School of Medicine, Vermillion, SD

JAMES REYNOLDS, MD • Orthopedic Surgeon, SpineCare Medical Group, Daly City, CAMARTIN H. SAVITZ, MD, FACS, FICS • Executive Director, American Academy of

Minimally Invasive Spinal Medicine and Surgery, and Adjunct Professor of Bioethics,MCP-Hahnemann School of Medicine, Philadelphia, PA

xii Contributors

JAMES W. SIMMONS, JR., MD • Alamo Bone and Joint Clinic, San Antonio, TXKERN SINGH, MD • Department of Orthopedic Surgery, Rush-Presbyterian-St. Luke’s

Medical Center, Chicago, ILDAISUKE TOGAWA, MD, PhD • Department of Orthopaedic Surgery and Spine Institute,

The Cleveland Clinic Foundation, Cleveland, OHTRENT L. TREDWAY, MD • Assistant Professor, Department of Neurosurgery, University

of Washington, Seattle, WAMICHAEL TZERMIADIANOS • Orthopedic Surgeon, University of Crete, Crete, GreeceALEXANDER R. VACCARO, MD • Professor of Orthopedic Surgery, Thomas Jefferson

University and the Rothman Institute, Philadelphia, PADOUGLAS WARDLAW, ChM, FRCS(Ed) • Consultant, and Orthopaedic and Spinal Surgeon,

Grampion University Hospitals Trust–Woodend Hospital, Aberdeen, ScotlandANTHONY T. YEUNG, MD • Voluntary Associate Clinical Professor, Department of

Orthopaedics, University of California San Diego School of Medicine, San Diego, CA,and Arizona Institute for Minimally Invasive Spine Care, Phoenix, AZ

HANSEN A. YUAN, MD • Professor of Orthopedic and Neurological Surgery, StateUniversity of New York Upstate Medical University, Syracuse, NY

Color Plates

Color plates 1–39 appear in an insert following p. 144.

Plate 1 Fig. 7 from Chapter 1; for full caption see p. 10.Plate 2 Fig. 16B from Chapter 1; for full caption see p. 17.Plate 3 Fig. 4A from Chapter 2; for full caption see p. 32.Plate 4 Fig. 5A from Chapter 2; for full caption see p. 33.Plate 5 Fig. 6A,B from Chapter 2; for full caption see p. 34.Plate 6 Fig. 7 from Chapter 2; for full caption see p. 35.Plate 7 Fig. 8 from Chapter 2; for full caption see p. 35.Plate 8 Fig. 9C from Chapter 2; for full caption see p. 37.Plate 9 Fig. 10 from Chapter 2; for full caption see p. 38.Plate 10 Fig. 12A from Chapter 2; for full caption see p. 39.Plate 11 Fig. 13A from Chapter 2; for full caption see p. 40.Plate 12 Fig. 14A from Chapter 2; for full caption see p. 41.Plate 13 Fig. 15A from Chapter 2; for full caption see p. 42.Plate 14 Fig. 16 from Chapter 2; for full caption see p. 43.Plate 15 Fig. 18A,B from Chapter 2; for full caption see p. 44.Plate 16 Fig. 12A,B from Chapter 3; for full caption see p. 57.Plate 17 Fig. 5A,B from Chapter 4; for full caption see p. 67.Plate 18 Fig. 22 from Chapter 4; for full caption see p. 91.Plate 19 Fig. 23A from Chapter 4; for full caption see p. 92.Plate 20 Fig. 27B,C from Chapter 4; for full caption see p. 97.Plate 21 Fig. 28E,F from Chapter 4; for full caption see p. 99.Plate 22 Fig. 29A,B from Chapter 4; for full caption see p. 102.Plate 23 Fig. 30 from Chapter 4; for full caption see p. 103.Plate 24 Fig. 31A from Chapter 4; for full caption see p. 104.Plate 25 Fig. 33A from Chapter 4; for full caption see p. 106.Plate 26 Fig. 35A from Chapter 4; for full caption see p. 110.Plate 27 Fig. 36 from Chapter 4; for full caption see p. 111.Plate 28 Fig. 37A–C from Chapter 4; for full caption see p. 112.Plate 29 Fig. 38A from Chapter 4; for full caption see p. 114.Plate 30 Fig. 39A from Chapter 4; for full caption see p. 115.Plate 31 Fig. 1B from Chapter 5; for full caption see p. 120.Plate 32 Fig. 7C,D from Chapter 5; for full caption see p. 129.Plate 33 Fig. 3A from Chapter 6; for full caption see p. 148.Plate 34 Fig. 7B from Chapter 9; for full caption see p. 211.Plate 35 Fig. 8 from Chapter 9; for full caption see p. 212.Plate 36 Fig. 12 from Chapter 9; for full caption see p. 215.Plate 37 Fig. 13 from Chapter 9; for full caption see p. 215.Plate 38 Fig. 17 from Chapter 9; for full caption see p. 219.Plate 39 Fig. 5 from Chapter 13; for full caption see p. 264.

xiii

Companion DVD

The companion DVD to this volume contains three video segments from the editor. TheDVD can be played in any "set-top" DVD player attached to an NTSC standard definitiontelevision set. The DVD may also be viewed using any computer with a DVD drive andDVD compatible playback software such as Apple DVD Player, Windows Media Player 8or higher (Win XP), PowerDVD, or WinDVD.

1History of Surgical Management of Herniated

Lumbar Discs From Cauterization to Arthroscopicand Endoscopic Spinal Surgery

Parviz Kambin, M D

INTRODUCTION

A review of the history of the surgical management of herniated lumbar discs as acommon cause of sciatica is not complete without acknowledging the efforts of manyinvestigators and researchers who have contributed to the understanding of the anatomyand origin of the sciatic nerve and lumbar intervertebral discs. In addition, the efforts ofscientists and clinicians who have participated in developing the surgical management ofdisc herniation in the last seven decades should be recognized.

MANAGEMENT OF BACK AND LEG PAIN IN ANCIENT MEDICINE

Low back and sciatic pain has been one of the most common and disabling spinaldisorders recorded in medical history. The role of the spinal canal’s contents in extremityfunction is well demonstrated in the Dying Lioness (Fig. 1), a ca. 650 BC. Assyrian artwork.

In the writings of Hippocrates (460–370 BC) one can find references to the anatomyof the brain, brachial plexus, and sciatic nerve. In animal dissections it appears that hehad difficulty in differentiating tendons from peripheral nerves. However, he attributedthe development of paresthesia, weakness of the limbs, and fecal and urinary retentionto spinal cord compression (1).

On the basis of his animal and human dissections, Aristotle (384–322 BC) describedvertebrate anatomy (2). Erasistratus (250 BC) distinguished between the role of motorand sensory nerve fibers in his findings from cadaver dissections (3).

Avicenna (980–1037 AD), a Persian physician and philosopher who was born inBokhara, also wrote extensively on human anatomy and the peripheral nerves. How-ever, his writings make no clear reference to sciatic pain. His text Canon of Medicineformed the cornerstone of medical practice for ensuing centuries. Avicenna condemnedthe reliance on mysticism and astrology in medicine (4). His writings were translatedinto Latin and included in the medical curriculum of European universities. Avicenna’sprincipal method of treating spinal disorders by traction and manipulation remains anaccepted practice in many centers at present (Fig. 2), (5,6). A calligraphy (Fig.3), dating

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

2 Kambin

Fig. 1. The dying lioness, ca 650 BC. (Reprinted with permission from refs. 1 and 42.)

Fig. 2. Avicenna’s a “Method of Treating Spinal Disorders by Traction and Manipulation.”(Reprinted with permission from ref. 5.)

to 1400 AD demonstrates the depth of curiosity of the times, and the information thatwas gathered from cadaver dissections. Their illustrations show the presence of 6 cervi-cal, 12 thoracic, and 5 lumbar segments. The origins of the brachial plexus from thecervical segments, the intercostal nerves from the thoracic nerves, and the sciatic nerve

from the lumbar segments are described. In addition, the two divisions of the sciaticnerve as it extends into the lower extremities are shown.

In the ancient literature there is no reference to surgical management of sciatic pain.However, the use of traction, local cauterization (Fig. 4), cupping, bloodletting, andopioids was common in Arabic, Persian, and Islamic medicine and subsequently inEuropean medicine. Acupuncture has been practiced in Chinese medicine for centuries.

RECOGNITION OF SCIATICA AND ITS ASSOCIATED SYMPTOMATOLOGY

Domenico Cotugno (Fig. 5), an eighteenth century Italian physician (7), introducedthe term sciatica into the medical vocabulary. Without having knowledge of the commonetiology of this disabling spinal disorder, he described some of the signs and symptomscommonly seen in association with sciatic pain. Subsequently, Cotugno’s disease as anentity gained acceptance in European medicine. Associated clinical findings of sciatica

History of Lumbar Disc Surgery 3

Fig. 3. Persian miniature from about 1400 AD. (Reprinted with permission refs. 1 and 42.)

were further detailed and documented by the writings of Putti (8), Valleix (9), Lasègue (10),and Brissard (11) in later years.

IDENTIFICATION OF ANATOMICAL AND PATHOLOGICAL CONDITIONS OF THE INTERVERTEBRAL DISC

In the late nineteenth and early twentieth centuries, many investigators contributed tothe understanding of intervertebral disc anatomy. In 1857, Virchow (12) publishedautopsy findings on the intervertebral disc in a patient who was injured and laterexpired. In 1868, von Luschka (13) described posterior disc protrusion in cadaversfound in the course of routine autopsy procedures. Kocher (14) referred to his findingson intervertebral discs at L1-L2 in a patient who had a traumatic injury. Schmorl’s (15)contribution to anatomical structures of the intervertebral disc also deserves recogni-tion. In 1926, he reported on autopsy findings on 5000 intervertebral discs, 15 of whichshowed evidence of disc protrusion into the spinal canal. However, despite this signifi-cant anatomical finding, he had not yet established the causal connection between discherniation and sciatic pain.

HISTORY OF SURGICAL MANAGEMENT OF SCIATICA

In the early twentieth century, laminectomy was being performed for the treatment ofa variety of spinal disorders. In 1911, Goldthwait (16) described the management of a39-yr-old male who underwent spinal manipulation, and then he developed paralysis in

4 Kambin

Fig. 4. Cauterization points for spine and other disorders. (Reprinted with permission fromref. 5.)

the lower extremities. His conservative management included plaster immobilizationand rest. The patient failed to show improvement, and 6 wk later he underwent exten-sive decompressive laminectomy, extending from L1 to S2. The patient responded tothe operative procedure and showed improvement. Goldthwait (16) attributed thepatient’s neurological deficit to detachment and protrusion of the fibrotic annulus intothe spinal canal, slippage of the articular processes, and abnormality of the transverseprocess of the lumbar segment.

In 1913, Dr. Elsberg of the New York Neurological Institute and Mt. Sinai Hospital,reported on his findings on 60 consecutive laminectomies. However, he did not believedisc pathology was responsible for the presenting symptomatology in any of thepatients described (17). In 1928, in a paper entitled a “Extradural Spinal Tumors, Pri-mary, Seconary, Metastisis,” Dr. Elsberg attributed compression of the cauda equina tothe presence of cartilaginous tumors (chondromas) (18).

In 1927, Putti (8) reported on one of his patients who underwent laminectomy andfacetectomy to decompress the L5 and S1 nerve roots and relieve sciatic pain. He furtherelaborated on the contribution of Sicard, who performed laminectomy from L3 to thesacrum to provide relief from sciatic pain.

History of Lumbar Disc Surgery 5

Fig. 5. Potrait of Domenico Cotugno.

Other investigators, including Stookey in 1928 (19) and Bucy in 1930 (20), alsoreported on the removal of chondroma-type tumors from the intervertebral discs thatwere causing pressure on the neural structures. Alajouanine, a neurologist residing inParis, reported on two patients who underwent laminectomy and discectomy in 1928(21,22). A brief translation of his article is as follows:

It is a very specific type of radiculomedullary compression that we call “a fibrocartilagi-nous nodule of the posterior aspect of the intervertebral discs.” This compression is mani-fested by radicular signs, more rarely medullary, most often unilateral. Surgical ablation,although sometimes laborious, like all premedullary tumors, usually results in the rapidregression of compressive disorders. Their first presentation was made in 1928 to the Sur-gical Society of a unilateral cauda equina syndrome due to a curious formation related toan intervertebral disc (ref. 2: Bull et Mem de la Soc nat de chir 12 Oct 1928, 54: 1452).Now we have seen a second case, absolutely identical to the first.

Case 1. Male, 37 years old, complained of left lumbosacral pain with root, sen-sory and sphincter problems for 4 years. The flow of Lipiodol was blocked below L5-S1. Ablation of fibrocartilaginous nodule from L5-S1 intervertebral disc. Rapid andcomplete cure.

Case 2. Female, 20 years old, had a 3-year history of pain in the left leg and whilewalking. There was foot drop, absence of achilles and medial plantar reflexes. Anesthesiaof L4-L5 and all sacral roots. Positive Lipiodol test at L3-L4. On July 18, 1929, disc pro-trusion, transdural approach, removal of fibrocartilaginous nodule in comparison to thefirst case. Partial recovery of the foot drop but not the ankle reflex. Notes probable com-pression of nerve roots by rongeurs in the course of laminectomy.

These nodules are neither tumors, chondromas nor fibrochondromas and are dis-tinctly different from chordomas. Basically, they are always related to the intervertebraldisc. We have shown that these curious formations should be considered to result fromherniation of the central pulp of the disc across the latter, the hernia produced either bytrauma or by pathological changes in the disc; in addition, the effects of these twocauses can be combined.

The use of Lipiodol is indispensable, not only with radiography but also with fluo-roscopy. The prognosis depends upon surgical treatment which is midline through thedura. If the protrusion is very lateral, the dura mater should be incised laterally. There is aproblem with retraction of the spinal cord in the neck and thorax, particularly evidentwhen the nodule is calcified and embedded in the cord. Such nodules should be suspectedin refractory lumbalgia and sciatica.

In 1931, Crouzon et al. (23) gave credit to the contribution of Alajouanine and furtherdetailed and described the clinical outcome of patients who underwent laminectomy anddiscectomy. A translation of their publication is as follows.

This is a new example of a fibrocartilaginous nodule on the posterior aspect of the inter-vertebral disc, producing a very specific type of root compression that one of us, withAlajouanine, has called attention to in a recent report. [Alajouanine T, Petit-Dutaillis D.Le nodule fibrocartilageneux de la face posterieure des disques intervertebraux. PresseMedicale nos. 98 and 102 of 6 and 20 September 1930]. The favorable results obtainedby surgical intervention make it possible to emphasize once more its clinical and thera-peutic value in such a disorder.

[There follows a case history, summarized here.]

A brickmaker, 44 years old, was hospitalized by Dr. Crouzon for refractory sciatica thathad kept him out of all work for 6 months. There was no special precipitating factor, butthere was a history of an acute injury to the lumbar region 7 years earlier when he fell 4meters onto his back and kidneys. After severe pain immobilized him for some days, he

6 Kambin

was able to go back to work, but with intermittent episodes of “lumbago” making him restfor 3 to 4 days. Only after 7 years did he begin to have (In June 1930) pains in the left legthat became increasingly severe and frequent. Examination on 20 February 1931 showedareas of pain in the lumbar region, calf and left heel. These were aggravated by the slight-est movement, cough or strain. When he stood, his weight was placed on the intact rightextremity. There was an antalgic spasm of the lumbar muscles, but hypotonia of thequadriceps and calf on the left side. His body was held forward when he walked with obvi-ous pain. The spine was held flexed forward and to one side. There was some atrophy ofthe left thigh and calf, the latter measuring 3 cm less than the healthy calf. There was aslight decrease in strength of flexion and extension of the foot on the left side. Knee reflexeswere equal, but the achilles and medial plantar reflexes were absent on the left. Sensoryexam showed sharp pain on pressure all along the left sciatic nerve and sharp pain onLasègue’s maneuver.

There was pain on pressure and percussion over the spinous processes of L4 and L5.The sensory exam of the plantar aspect of the foot was consistent with anesthesia for allmodalities on the plantar aspect of the foot and posterior aspect of the left calf, extending5 cm onto the posterior of the thigh. There was also a band of sensory loss on the lateralaspect of the foot and adjacent leg, ascribed to L5, S1 and S2. There was some sphincterdysfunction with pain on defecation and difficulty in urination. Lumbar puncture onFebruary 25, 1931 showed normal fluid and normal pressure, slight dissociation betweenalbumin and cells (40 g albumin and 2 cells). Wasserman tests of blood and CSF were neg-ative. X-rays showed some narrowing at L4-L5. A Lipiodol study showed temporary block-age at L4-L5 under the fluoroscope, but by the time the patient reached the radiographyroom, the oil had all fallen to the bottom. The temporary blockage was pronouncedenough to induce Dr. Alajouanine to operate on the patient on 7 March 1931. Laminec-tomy of L3-L5 showed ossification of the ligamentum flavum at L4-L5; the dura wasindented, and the ligament was removed. The dura was opened to show displacement ofthe nerve roots by a whitish nodule compressing the left L5 root. The root was compressedto a thread at the level of the intervertebral foramen, as if it had been partially destroyedby stretching. In order to free it without further damage, the dura was cut transversely.This made it possible to displace the root of L5 to the left and the rest of the roots to theright. The dura was incised anteriorly over the nodule, and a specially designed spatulawas used to hold the root while the fibrocartilaginous nodule was removed. Because of thetransverse cut in the dura, no attempt was made to suture it, and the wound was closed inlayers with catgut and without drainage.

The postoperative course was uneventful; sutures were removed on Day 9. The outcomeof surgery was very good and recovery was rapid. The day after surgery the patient saidthe left leg no longer hurt, and re-examination showed a return of sensation in the areas ofL5, S1 and S2. He could now feel the bedsheets on his foot. Fifteen days after surgery hehad no complaints and could get out of bed; 25 days after surgery he stood straight andwalked normally without pain or fatigue.

Examination on April 25, 7 weeks after surgery, showed normal posture, with weightequally distributed on the two legs. Flexion and extension of the left foot were normal.Mild hypotonia persisted in the left thigh as did slight atrophy of the calf and thigh on theleft. The achilles and medial plantar reflexes were still absent. There was no pain on pres-sure over the course of the left sciatic nerve. There was no pain on straight-leg raising.Objective examination of sensation showed a slight decrease in tactile sensation on thelateral border of the left foot. The sphincter problems had resolved, and the patient’s gen-eral health was excellent.

Histological study of the specimen by Dr. I. Bertrand showed fibrocartilaginous tissuewith abundant interstitial stroma containing amorphous tissue with some collagen bun-dles. There were only a few cells, but those seen resembled cartilage cells. An examinationfor Virchow’s physaliferous cells was negative. There were few vessels, and in some placesthe absence of staining indicated some necrosis.

History of Lumbar Disc Surgery 7

This case should be added to similar cases published in France by Alajouanine andPetit-Dutaillis, by Robineau and, in the foreign press, by Adson, Stookey, then Bucyand P. Bailey, and, very recently, by Katzenborn, making a total of 23 operated cases. Thenew case reported appears to prove that this is not a very rare condition and that the num-bers will soon increase now that attention has been directed to these facts.

In view of this new case, it seems appropriate to emphasize certain points: the role oftrauma is beyond doubt, even though in this case it may be dismissed, for in this case theinjury occurred 7 years previously. Emphasis is placed on the occasionally long latentperiod before symptoms become manifest. Some temporary lumbar symptoms of an appar-ently common type may occur in this period, as if the lesion, only produced by the initialtrauma, gradually becomes more pronounced, undoubtedly affected by repeated strains inthose whose occupations are strenuous. There is a notable incidence of unilateral symp-toms. The lumbar region is not the only site of pathological disc changes; the first casesdealt with those in the neck. Although Stookey initially thought these fibrocartilaginouslesions were exclusively cervical, it is clear that they may occur elsewhere, although theydo appear to be rare in the thoracic region.

In addition to clinical signs and symptoms, compression is also manifested by a disso-ciation between albumin and cells and by a blockage of Lipiodol. The blockage of the oilmay be quite temporary and be seen only on fluoroscopy. For this reason the authorsemphasize the need for this diagnostic procedure as well as radiography. The absence ofthe disc in radiograms was similar to that in long-standing Pott’s disease. However, itshould be noted that there is no sign of herniation into the vertebral bodies. It seems likelythat in compression phenomena of traumatic origin the compression, or even absence, ofthe disc might promote the development of fibrocartilaginous nodule formation.

The histological study also shows that these nodules should not be considered to betumors (neoplasms) as has been thought to be the case by those authors who called themfibrochondromas, ecchondromas or even chondromas of the disc. These structures are anintegral part of the intervertebral disc with no neoplastic characteristics, but should beconsidered protrusions of the disc or of the nucleus pulposus across a break in the poste-rior part of the intervertebral disc into the spinal canal. This interpretation (Schmorl,Andrae) seems the only logical one.

It is more painstaking to surgically remove these pathological structures than otherintraspinal tumors. In the region of the cauda equina the compressed roots must be freedvery gently and very slowly. Even if the size of the nodule is small, its consistency is veryhard and it exerts a very firm compression. In our case the left root at L5 had already beenheavily compressed and stretched. Sometimes the root may be in contact with the lamina,and care must be taken in removing the lamina to avoid injuring the root.

Dandy (24) independently reported on the removal of a detached fragment of cartilagi-nous tissue from the intervertebral disc for treatment of sciatic pain.

Mixter and Barr are credited for establishing a clear causal connection between theherniated disc and sciatica. They provided a detailed description of disc herniation andpopularized laminectomy and discectomy for surgical management of herniated lumbardiscs (25).

Between the 1930s and 1950s, orthopedic and neurological surgeons followed thetraditional teaching of Mixter and Barr that consisted of wide exposure, bilateral dissec-tion of the paraspinal muscles, laminectomy, and extensive epidural hemostasis andcoagulation in the course of extraction of herniated disc fragments.

The traditional surgery described by Mixter and Barr was later modified andbecame less invasive with the introduction of the microscope to the surgical field byYasargil, a Turkish surgeon, in 1972 (26,27). This concept was further advanced by otherinvestigators (28).

8 Kambin

EMERGENCE OF THE MINIMALISTS’ CONCEPT

Annular Fenestration and Reduction of Hydrostatic Pressurein the Intervertebral Disc

The earliest recorded departure from the concept of traditional laminectomy and dis-cectomy in the treatment of a herniated lumbar disc is found in an article published byHult (29) in 1950, in which he advocated an anterior retroperitoneal annular fenestrationfor decompression of herniated lumbar discs. The relationship between hydrostatic pres-sure of the intervertebral disc and the size of the annular bulge and protrusion has been asubject of interest to many investigators. Virgen (30) demonstrated that the height of theintervertebral disc is decreased and the annulus bulged outward when intervertebral discswere subjected to axial loading. Brown et al. (31) showed that the annular bulge wasincreased on the side on which the spine was flexed and the annulus was flattened on theopposite side. Nachemson (32,33) also demonstrated bulging of the annulus associatedwith increased intradiscal pressure under load, particularly in the sitting position and withforward bending and lifting. Kambin and colleagues reported on their in vivo evaluationof hydrostatic pressure in the intervertebral disc prior to and following annular fenestra-tion via a 4.9-mm-outer diameter (od) trephine and partial nuclear resection. A consider-able reduction of intradiscal pressure was observed when patients were instructed toextend and rotate the trunk following annular venting (34,35). However, long-termpatency of the annular fenestration remains highly questionable. Although Sakamotoet al. (36) showed that the reduction of intradiscal hydrostatic pressure may be maintainedup to 21 mo postoperatively, Hampton et al. (37) reported healing and closure of the sur-gically created defect in the annulus between 3 and 12 wk after surgery. This phenomenonwas also confirmed in the my own experience when a repeated surgery was required a fewmonths following the original percutaneous arthroscopic discectomy. It was found thatthe original site of annular fenestration was closed with scar tissue.

Concept of Nuclear Mass Reduction

Lyman Smith should be recognized as a champion of the minimally invasive movement(38). Learning from the experience of Lewis Thomas in rabbits (39), he introduced theconcept of dissolving the nucleus pulposus by intradiscal injection of chymopapain. Thesimplicity of the procedure and the fact that the operative technique did not violate thecontent of the spinal canal attracted the attention of many orthopaedic and neurologicalsurgeons, both in the United States and abroad. This was followed by many presenta-tions, hands-on seminars, and publications in the ensuing years.

Encouraged by previously reported satisfactory outcomes of chemonucleolysis, in theearly 1970s, following institutional approval, Kambin (Fig. 6) initiated a feasibility studyon the efficacy of mechanical nuclear debulking for the treatment of herniated lumbardiscs via a Craig cannula inserted into the intervertebral disc dorsolaterally (40,41).

Clinical research conducted by my colleagues and I in the ensuing years was directedtoward establishing the effect of central nucleotomy on the size of the bulge or hernia-tion. In 1973, at The Graduate Hospital of Philadelphia, we combined the centralnucleotomy via a Craig biopsy cannula with laminectomy in patients who demonstratedsigns, symptoms, and imaging evidence of disc herniation (Fig. 7) (35). In 1973, a 60-yr-old male with myelographic and clinical evidence of disc herniation at L3-L4 and

History of Lumbar Disc Surgery 9

10 Kambin

Fig. 6. Authorization by Board of Governors of Doctors Hospital permitting use of Craigcannula for nucleotomy in management of disc herniation.

Fig. 7. Intraoperative photo demonstrating effect of nucleotomy via a Craig cannula oncontour and geometry of herniated disc, which was visualized through open laminatomy.

L4-L5 underwent open laminectomy at both levels. The larger disc herniation at L3-L4was removed through the laminectomy exposure. However, the smaller protrusion atL4-L5 was decompressed through the posterolaterally introduced Craig cannula. Thepatient had a satisfactory outcome with no complications.

In February 1974, a 52-yr-old male presented with right sciatica and was diagnosedwith herniated discs at L3-L4 and L4-L5. This patient underwent a combined operativeprocedure. The herniated disc at L4-L5 was removed through the laminectomy site;however, the L3-L4 intervertebral disc was decompressed through the cannula that hadbeen inserted dorsolaterally.

In April 1974, a 43-yr-old female with unremitting sciatic pain and myographic evi-dence of disc herniation at L4-L5 underwent percutaneous nucleotomy via a Craig can-nula. This patient failed to respond to the nucleotomy procedure and subsequentlyrequired a laminectomy when a large disc herniation at the index level was identifiedand excised.

In June 1974, a similar combined operation was performed on a 52-yr-old male withclinical and myographic evidence of a large disc herniation at L5-S1 and a smaller protru-sion at L4-L5. The L5-S1 herniation was excised through the laminectomy site, and theL4-L5 intervertebral disc was decompressed via a mechanical nucleotomy techniquethrough the inserted cannula. Although in the ensuing years a number of patients under-went a simple mechanical nucleotomy via the inserted cannula and the combined proce-dure, we were unable to demonstrate an appreciable reduction in the size or shape of theherniation following a simple central nucleotomy. Therefore, our efforts were then directedtoward the development of instruments and surgical techniques that would provide betteraccess to posterolaterally dislodged disc fragments via a posterolateral approach. Newlydesigned instruments were developed that included a cannulated obturator (Fig. 8)for precise positioning of instruments and a 6.5-mm-od cannula that accommodated anupbiting forceps. This was followed by development of a flexible-tip forceps and adeflecting tube that permitted dorsal angulation of the inserted forceps and aided inevacuation of posterior nuclear tissue. In 1981, under the auspices of the Human Sub-jects Committee of The Graduate Hospital (Fig. 9), I initiated a series of preliminary

History of Lumbar Disc Surgery 11

Fig. 8. Kambin cannulated spinal obturators. Blunt-end cannulated obturators for precisepositioning of the instruments as shown.

investigations on the feasibility of the use of a 6.4-mm-od cannula using upbiting andflexible-tip forceps (Figs. 10 and 11) (34,35,40–43).

In 1975, Hijikata (Fig. 12) from the Toden Hospital in Japan independently experi-mented with mechanical nucleotomy via a 2.6-mm-od cannula that was inserted intothe center of the intervertebral disc via a posterolateral access. He reported a satisfac-tory postoperative outcome in 64% of patients (44). Following Hijikata’s experience,Schreiber and Suezawa developed a series of cannulas that were telescoped one overthe other and placed in the center of the intervertebral disc via a posterolateral access.The larger cannulas with a 7 to 8-mm internal diameter (id) permitted the insertion oflarger forceps and more rapid evacuation of nuclear tissue (45). In 1981, in the UnitedStates, Blum et al. (46) experimented with Hijikata’s nucleotomy technique andreported their findings before the International Society of the Lumbar Spine. In 1983,Hoppenfield (47) also used a posterolateral approach and manual instruments fornucleotomy. Friedman and Jacobson experimented with a far lateral approach to accessthe lumbar intervertebral disc. These investigators passed a no. 40 French chest tubethrough an incision over the iliac crest and directed it toward the intervertebral disc atthe index level. After annulotomy the disc fragments were evacuated with manual for-ceps (48). In 1985, Onick promoted the concept of central nucleotomy via a mechanicaltool called a nucleotomy (49). The small caliber of the instruments and the simplicity ofthe operative procedure contributed to the popularity of the operative technique in theensuing years (Fig. 13).

12 Kambin

Fig. 9. Permission from Human Subjects Committee of The Graduate Hospital for experimentaluse of 6.4-mm-od cannula, using upbiting and flexible-tip forceps in percutaneous disc surgery.

History of Lumbar Disc Surgery 13

Fig. 10. Original instruments developed in early 1980s for percutaneous discectomy underX-ray control.

Fig. 11. Deflecting tube and flexible-tip forceps for access and removal of posteriorly lodgeddisc herniation and entry to L5-S1 intervertebral disc.

14 Kambin

Fig. 12. Illustration from article by Hijikata showing the principle of central nucleotomyin 1975.

Fig. 13. Illustration demonstrating the use of nucleotome in the treatment of disc herniation.

The introduction of laser light into the surgical armamentarium opened another front inthe management of lumbar disc herniation (35–50). The small caliber and relative flexi-bility of the laser fibers was a source of encouragement and appeared to be suitable fornuclear vaporization. A variety of laser lights were introduced into the marketplace andthen used by many investigators. In January 1990, with the permission of the FederalDrug Administration and Internal Review Board of The Graduate Hospital, I initiated aclinical study of the feasibility of vaporizing disc fragments with laser light underarthroscopic illumination and magnification (51). It was found that the wide arc ofdeflection of the laser fibers and concern about injury to neural structures preventedadequate decompression and lysis of posterior herniated disc fragments.

Striving Toward Access and Retrieval of Posterior HerniatedDisc Fragments Via an Intradiscal Approach

While the advantages and safety of a small-caliber nucleotome and laser fibers fornuclear decompression were being promoted and debated in the late 1980s, my colleaguesand I continued to utilize the standard 6.4-mm-od cannula for discectomy. Although ourdeflecting tube and flexible-tip forceps (Fig. 11) permitted posterior nucleotomy, we wereunable to adequately access and retrieve subligamentous or nondisplaced extraligamentousherniations.

After a series of cadaver studies, it was determined that a high negative atmosphericpressure could be introduced into a contained intervertebral disc without any inadvertentcomplications. Subsequently, we introduced this technique into our clinical practicein an attempt to dislodge the herniated disc fragments and move them into the path of theinserted cannula (34,40,41). However, this technique was not always successful and wastherefore later abandoned. An articulating forceps (Fig. 14) was introduced whose tipdeflected far enough to access posterior and posterolateral herniated disc fragmentsintradiscally and to decompress directly the nerve roots (42,43,52).

IDENTIFICATION OF A SAFE ZONE ADJACENT TO NERVE ROOTS FORANCHORING OF INSTRUMENTS

Although posterolateral access to the intervertebral disc was used for biopsy ofvertebral bodies (53–55), discography, chemonucleolysis (38), and automatednucleotomy (49), the site of lodging of instruments and annular window on the annulushad not been clearly defined. The close proximity of major neurovascular structures tothe posterolaterally inserted instruments necessitated the identification of a safe zone onthe posterolateral surface of the annulus fibrosus for anchoring cannulas with largerdiameters. After a series of cadaver dissections at The Graduate Hospital and theAnatomy Laboratory of the Hospital of the University of Pennsylvania, a triangularsafe zone on the posterolateral annulus, between the traversing and exiting nerve roots,was identified. Subsequently, we positioned needles in and around the safe zone and

History of Lumbar Disc Surgery 15

Fig. 14. Articulating suction forceps used for intradiscal access to non-migrated sequestrateddisc herniation.

radiographic studies were conducted. These allowed us to identify the radiographiclandmarks of the safe zone in both the anteroposterior and lateral projections. There-fore, we began to emphasize the importance of localization of the tip of the insertedneedle on the annulus at the onset of the operative procedure rather than in the center ofthe intervertebral disc (Fig. 15).

The triangular working zone is bordered anteriorly by the exiting root, inferiorly bythe proximal plate of the lower lumbar segment, and medially by the traversing root andthe dural sac. The floor of the triangular working zone is occupied by the intervertebraldisc, the vertebral plate, and the posterior boundary of the adjacent vertebra (Fig. 16A,B) (42,43,56,57). This region is covered by loosely woven adipose tissue and, attimes, superficial veins, which are readily observed by arthroscopic or endoscopicexamination. Mirkovik and Schwartz (58) independently measured the dimensions ofthe triangular working zone and have confirmed that cannulas with larger diameters canbe safely inserted between the traversing and exiting roots in the course of arthroscopicor endoscopic spinal surgery.

The description of the radiographic landmarks of the triangular working zone madeit possible to lodge the instruments precisely and to monitor them fluoroscopicallyboth anteroposteriorly and laterally. It was stipulated that a midpedicular positioning ofthe instruments in the anteroposterior projection is suitable for intradiscal subligamen-tous or intracanalicular access to the contents of the spinal canal. Lateral pedicular line

16 Kambin

Fig. 15. Illustration demonstrating complications that may become associated with localiza-tion of needle in the center of the disc at the onset of percutaneous spine surgery. Note that theneedle may pass through the ligamentum flavum dora and enter the intervertebral disc with afinal satisfactory radiographic appearance in the anteropasterior and lateral projection.

positioning in the anteroposterior projection may be used for evacuation of anextraforaminal herniation (42,43).

History of Development of Larger-Diameter Cannulas

The oldest and most commonly used cannulas are the ones described by Ottolenghi(54) and Craig (53) that were commonly used for vertebral body biopsy. Hijikataoriginally suggested the use of a 2.6-mm-od cannula (44). However, he later modifiedhis technique and used larger-diameter cannulas.

Onik developed an automated nucleotome (Fig. 13) for mechanical resection ofnuclear tissue (49). The instrument was designed along the lines of Hijikata’s instru-ments. At this stage of development, emphasis was placed on access and retrieval ofnuclear tissue, rather than removal of herniated disc fragments and direct decompres-sion of the nerve roots. Introduction of a large-diameter cannula in the clinical setting

History of Lumbar Disc Surgery 17

Fig. 16. (A) Copy of photo of triangular working zone which was published in l988. (B)Illustration showing the boundaries of the triangular working zone: A, the exiting root; B, duralsac; C, intervertebral disc; D, traversing root.

18 Kambin

lead to further investigation and description of the triangular working zone on the pos-terolateral annulus.

My colleagues and I originally used a Craig cannula for mechanical nucleotomy.However, in the early 1980s, we began to use cannulas with a larger diameter (6.4-mmod) (40). These provided a 5-mm inner working space. In addition, we introduced theconcept of using a blunt-tipped cannulated obturator for precise positioning of theinstruments on the annulus (Fig. 8).

We later introduced the concept of the unilateral biportal approach and oval cannulas(5 × 8 and 5 × 10 mm id) (59–64), (Fig. 17 A,B) that were designed to fit within the trian-gular working zone. The height of the intervertebral disc in the triangular working zoneprevents the insertion of larger cylindrical-shaped cannulas into the intravertebral discwithout the need for undue resection of the vertebral plates and part of the vertebral bodiesof the adjacent segments. Schreiber et al. (45) and Shepperd (65) have continued to usegradually dilating, telescopic cannulas up to 10 mm in diameter to enter the intervertebraldisc via a posterolateral access. In our experience, overstretching of the nerve rootsby the larger cannulas was associated with postoperative dysesthesia, which led to thedevelopment of oval-shaped cannulas that proved safe in our clinical practice.

As early as 1991, we used 10- to 23-mm-id cannulas for the endoscopic interlaminarapproach and intracanalicular surgery (62,63,66) (Fig. 18A,B) and arthroscopic forami-nal decompression (60,79) (Fig. 18C). A modified version of this technology recentlyhas been marketed (67).

Arthroscopic and Endoscopic Visualization and Birth of the Term Minimally Invasive Spinal Surgery

Bozzini, an obstetrician from Frankfurt, is credited with introducing the concept ofvisualizing internal organs in 1807 (Fig. 19), (68). His work was originally introducedto a faculty in Vienna and was rejected. He was criticized and censored for havingunreasonable curiosity. However, Bozzini’s noble idea continued to flourish, and manyinvestigators further developed, enhanced, and successfully utilized endoscopes for thediagnosis and treatment of a variety of medical disorders (69).

Use of the scope for diagnosis of spinal abnormalities dates back to 1931, whenBurman from the Hospital for Joint Diseases in New York City described his experiencewith the use of an endoscope for visualization of intracanalicular pathologies of the caudaequina in cadaver specimens. However, owing to the size of the instruments, he wasunable to inspect the intrathecal structures (70).

In 1938, Pool from Columbia-Presbyterian Hospital in New York developed amyeloscope for intra thecal inspection of normal and abnormal structures (71,72). Inrecent years, other investigators have utilized rigid and flexible fiberoptics for visual-ization of the epidural and subarachnoid spaces (73,74). However, in our experience, itis difficult to advance flexible fiberoptics, particularly on the ventral surface of the dura.Invariably, close contact with and adhesions between the ventral dura and the posteriorlongitudinal ligament prevent clear visualization and advancement of the fibers andmay result in a dural tear.

Hausmann and Forst (75) used an arthroscope to inspect the contents of the interver-tebral disc following open laminectomy and discectomy. Schreiber et al. (45) used anarthroscope via a second portal that was inserted into the intervertebral disc dorsolater-

History of Lumbar Disc Surgery 19

Fig. 17. (A) From top: 5 × 10 mm id oval cannula; two cannulated obturators are passedthrough the appropriate jig in preparation of insertion of a 5 × 10 mm oval cannula; and a 5 × 8mm oval cannula, a cannulated obturator, and a half-moon cannula are passed through the lumenof the appropriate jig in preparation of insertion of a 5 × 8 mm oval cannula. (B) Illustrationdemonstrating cross section of two cannulated obturators. which permits their use together priorto insertion of an oval cannula.

ally from the opposite side in order to inspect and resect nuclear tissue under directvisualization.

A meaningful use of arthroscopes and endoscopes in the field of spinal surgery wasnot realized until 1988, when the anatomical and radiographic appearance of the pos-terolateral annulus was described for safe positioning of instruments adjacent to thespinal canal (42,56). Subsequently, the arthroscopic appearance of intradiscal, perian-

20 Kambin

Fig. 18. (A) Illustration of the 10-mm-id cannula with side window used for translaminaraccess to the spinal canal for the removal of sequestrated disc herniation under endoscopic con-trol; (B) demonstration of the use of scope through the side window and the forceps through thecannula for visualization and extraction of disc herniations from the spinal canal; (C) illustrationof short cannulas which were used for foraminal decompression under arthroscopic control.

History of Lumbar Disc Surgery 21

Fig. 19. Illustration of Bozzini’s endoscope for visualization of internal organs.

Fig. 20. Photo of our original working-channel arthroscope that permitted manipulation andangulation of inserted instruments within intervertebral disc.

nular, and neural structures was demonstrated (42,43). In the mid- and late 1980s, dur-ing our investigational phase, when our efforts were being directed toward intradiscalaccess and retrieval of herniated disc fragments, with the cooperation of industry(Dyonics, Andover, MA), we developed a working channel arthroscope that providedample space for manipulation and angulation of instruments inserted within the inter-vertebral disc space (Fig. 20). However, an inability to establish adequate fluid irriga-tion within the intravertebral disc, poor visualization, and an inability to maneuver theforceps within the disc space adequately led to discontinuation of the use of thisinstrument. By the mid-1990s, endoscopic examination and visualization of the con-tents of the spinal canal via a posterolateral, transforaminal approach became a com-mon practice (43,52,61,77,78). Our experience with the use of radiofrequencycoagulators for control of epidural bleeding further facilitated access and retrieval ofsequestered disc fragments from the spinal canal via a transforaminal approach (62,64,79).Endoscopic laminatomy and foraminotomy has been reported (60,66,67,80,81). How-

22 Kambin

Fig. 21. Photograph taken during the First International Symposium held at the GraduateHospital in Philadelphia in 1983. From left to right: Dr. Hijikata, Professor Adam Schreiber,Dr. Parviz Kambin.

Fig. 22. Scientific exhibits at the American Academy of Orthopaedic Surgeons. From left toright: Parviz Kambin, John Sheppherd, Hal Matthew.

ever, the advantages of this technology in compared with conventional microscopicdiscectomy is currently debated.

Although endoscopes had been widely used for laparoscopic and thoracoscopicsurgery, in recent years Obenchain and other investigators have used this technology inthe treatment of a variety of spinal disorders (82–89).

History of Lumbar Disc Surgery 23

Fig. 23. Photograph from hands-on sessions held at the Graduate, Hospital and MCP Hahne-mann University Hospital annually. The orthopedic and neurological surgeons had the opportunityto learn and practice the principle of arthroscopic and endoscopic surgery on cadaver specimens.

Fig. 24. Photograph of faculty during panel discussions, question-and-answer period followinga closed-circuit live surgical demonstration.

In April 1990, the term minimally invasive spinal surgery was coined when a surgi-cal approach for the treatment of a variety of spinal disorders under arthroscopic orendoscopic magnification and illumination became a reality, and the International Societyfor Minimal Intervention in Spinal Surgery was established. The society has heldannual symposiums and workshops in both the United States and Europe, where thefaculty and surgeons interested in the field of minimally invasive spinal surgery had theopportunity to communicate and interact. (Figs. 21–24).

ACKNOWLEDGMENT

I would like to extend my appreciation to Dr. Oscar Sugar, Emeritus Professor ofSurgery, University of Illinois for his contribution and assistance in translating the articlesby Alajouanine and Crouzon et al.

REFERENCES

1. Hippocrates in Garrison’s History of neurology; Charles C Thomas, Springfield, IL, 1969, 7–8.2. Aristotle In medicine: an illustrated history, Harry N. Abrams, New York, 1978, p. 219.3. Erasistratos Medicine: an illustrated history, Harry N. Abrams, New York, 1978, pp. 310–312.4. Avicenna in Medicine an illustrated history, Harry N. Abrams New York, 1978, p. 313.5. Avicenna in The illustrated history of surgery, A B Nordbok, Gothenburg, Sweden, 1988, p. 73.6. Avicenna in Medicine an illustrated history, Harry N. Abrams, New York 1978, p. 310.7. Cotunio D. De Ischiade Nervosa Cammentarius. Simoncos Brothers, Naples, 1764.8 Putti V. The Lady Jones lecture: pathogenesis of sciatic pain. Lancet 1927;2:53.9. Valleix. Quoted from Reynolds F and Katz S: Herniated lumbar intervertebral disc, American

Academy of Orthopaedic Surgeons Symposium on the Spine, CV Mosby, St. Louis, 1969.10. Lasegue CH. Considerations sur la sciatique. Arch Gen Med 1864;2:558. 11. Brissard. Quoted from Reynolds F and Katz S: Herniated lumbar intervertebral disc, American

Academy of Orthopaedic Surgeons Symposium on the Spine Mosby CV, St. Louis,1969;84–96.

12. Virchow R. Untersuchungen uber die Entwickelung des Schadelgrundes, G Reimer, Berlin,1857.

13. von Luschka H. Die Gelenke des menschlichen Korpers, 1858; IV G Reimer, Berlin, 1858. 14. Kocher T. Die Verletzungen der wirbelsaule. zugleich als beitrag zur phsyiologic des men-

schlichen ruckenmarks. Mitt Grenzgeb Med Chir 1896;1:415.15. Schmorl G. Die pathologische anatomic der wirbelsaule. Verch Deutsch Orthop Gest

1926;21:3.16. Goldthwait JE. The lumbosacral articulation: an explanation of many cases of alumbago,

sciatica and paraplegia. Boston Med Surg J 1911;164:365.17. Elsberg CA. Experiences in spinal surgery: observations upon 60 laminectomies for spinal

disease. Surg Gynecol Obstet 1913;16:117.18. Elsberg CA. Extradural spinal tumors, primary, secondary, metastasis. Surg Gynecol Obstet

1928;46:1. 19. Stookey B. Compression of the spinal cord due to ventral extradural cervical chondromas.

Arch Neurol Psychiatry 1928;20:275. 20. Bucy PC. Chondroma of intervertebral disc. JAMA 1930;94:1552.21. Alajouanine TH. From the presidential address by Professor Jean Cauchoix before Annual

Meeting of the International Society for the Study of the Lumbar Spine, San Francisco,June 1978.

22. Alajouanine TH, Petit-Dutaillis, D. Le nodule fibrocartilageneux de la face face pos-terieure des disques inter-vertebraux, II. Etude clinique et therapeutique d’une varietenouvelle de compression radiculo-medullaire extra-durale. Presse Med1930;38:1749–1751.

23. Crouzon O, Petit-Dutaillis D, Christophe J. Sur un cas de compression de la queue decheval, d’origine traumatique, par un nodule fibro-cartilagineux du disque intervertebral:operation. Gurison Rev Neurol (session of 7 May 1931); l(5):612–617 (this is chronologi-cally vol. 55).

24. Dandy WE. Loose cartilage from the intervertebral disk simulating tumor of the spinal cord.Arch Surg 1929;29:660.

25. Mixter WJ, Barr JS. Rupture of the intervertebral disc with involvement of the spinal canal.N Eng J Med 1934;211:210.

24 Kambin

26. Yasargil MG. Die klinischen erfahrungen mit der mikrotechnik. Schweiz Arch Neuropsy-chiatry 1972;111:493–504.

27. Yasargil MG. Small approach to herniated disc. Adv Neurosurg 1997;4:81.28. Williams RW. Microsurgical lumbar discectomy: report to American Association of Neu-

rology and Surgery, 1975. Neurosurg 1979;4(2):140.29. Hult L. Retroperitoneal disc fenestration in low back pain and sciatica. Acta Orthop Scand

1956;20:342–348.30. Virgin WJ. Experimental investigation into the physical properties of the intervertebral

disc. J Bone Joint Surg 1951;33B:60.31. Brown T, Hansen RJ, Yorra AJ. Some mechanical tests on the lumbosacral spine with par-

ticular reference to the intervertebral disc: a preliminary report. J Bone Joint Surg1957;39A:1135.

32. Nachemson A. Lumbar intradiscal pressure. Acta Orthop Scand (Suppl) 1960;43:104.33. Nachemson A. In vivo discometry in lumbar discs with irregular myelogram. Acta Orthop

Scand 1965;36:418.34. Kambin P, Brager M. Percutaneous posterolateral discectomy: anatomy and mechanism.

Clin Orthop 1987;223:145–154.35. Kambin P: History of disc surgery, in Minimal Intervention in Spinal Surgery. (Kambin ed),

Urban & Schwarzenberg, Baltimore, 1991, pp. 3–8.36. Sakamoto T, Yamakawa H, Tajima T, Sawaumi A. A study of percutaneous lumbar

nucleotomy and lumbar intradiscal pressure. March 1989 International Symposium on Per-cutaneous Nucleotomy, Bruxelles, 1989.

37. Hampton D, Laros G, MacCarron R, Franks D: Healing potential of annulus fibrosis. Spine1989;14(4):398–401.

38. Smith L, Garvin PJ, Gesler RM, Jennings RB. Enzyme dissolution of the nucleus pulposus.Nature 1963;198:l311,1312.

39. Thomas L. Reversible collapse of rabbit ears after intravenous papain. J Exp Med 1956;104:245.40. Kambin P, Gellman H. Percutaneous lateral discectomy of the lumbar spine: a preliminary

report. Clin Orthop 1983:174:127–132.41. Kambin P, Sampson S. Posterolateral percutaneous suction—excision of herniated lumbar

intervertebral discs: report of interim results. Clin Orthop 1986;207:37.42. Kambin P. Arthroscopic microdiscectomy, in Minimal Intervention in Spinal Surgery

(Kambin pr ed.), Urban & Schwarzenberg, Baltimore, 1991, pp. 67–1100.43. Kambin P. Arthroscopic microdiscectomy. Arthroscopy 1992;8:287–295.44. Hijikata S. Percutaneous diskectomy: a new treatment method for lumbar disk herniation.

J Toden Hosp 1975;5:5–13.45. Schreiber A, Suezawa Y, Leu H. Does percutaneous nucleotomy with discoscopy replace

conventional discotomy? Clin Orthop 1989;238:35–42.46. Blum RS, Swartz A, Morris J. Percutaneous nucleotomy. May 1981;presented at the Inter-

national Society of Lumbar Spine: Paris, France. 47. Hoppenfield S. Percutaneous removal of herniated lumbar disc through the tube technique.

May 1983; Presented at the American Association of Bone and Joint Surgeons, Sea Island,Georgia.

48. Friedman WA. Percutaneous discectomy: an alternative to chemonucleolysis. J Neurosurg1983;13:542.

49. Onik G, Helms C, Ginsburg L, et al. Percutaneous lumbar discectomy using new aspirationprobe. AJR 1985;144:1137–1140.

50. Ascher PW, Holzer P. Laser denaturation of the nucleus pulposus of herniated interverte-bral discs, in Arthroscopic Microdiscectomy Mininal Intervention in Spinal Surgery(Kambin pr, ed.), Urban & Schwarzenberg, Baltimore, 1991, pp. 137–140.

51. Kambin P. Endoscopic laser discectomy. Jan 1990;FDA, IDE G890238–S1.52. Kambin P. Arthroscopic microdiscectomy: lumbar and thoracic spine in Spine Care

(White All, ed.), CV Mosby, St. Louis, 1995, pp. 1002–1016.

History of Lumbar Disc Surgery 25

53. Craig, FS. Vertebral-body biopsy. J Bone Joint Surg 1956;38-A:93–102.54. Ottolenghi, CE. Vertebral body biopsy, aspiration biopsy. J Bone Joint Surg 1955;37–A.55. Valls J, Ottolenghi CE, Schajowicz F. Aspiration biopsy in diagnosis of lesions of vertebral

bodies. JAMA 1948;136:376.56. Kambin P. Percutaneous lumbar discectomy: current practice. Surg Rounds Orthop 1988;

31–35.57. Kambin P, McCullen G, Parke W, et al. Minimally invasive arthroscopic spinal surgery.

Instruct Course Lect 1997;46:143–161.58. Mirkovik SR, Schwartz DG. Anatomic considerations in posterolateral procedures. Spine

1995;20:1965–1971.59. Kambin P. Unilateral biportal percutaneous surgical procedures and instrumentation. US

patent 1991;5,395,317. 60. Kambin P. Arthroscopic foraminal surgical procedures performed through a working channel

scope. Patent application 08/207831,1994.61. Kambin P. Diagnostic and therapeutic spinal arthroscopy. Neurosurg Clin N Am 1996;(1):

65–76.62. Kambin P. Arthroscopic microdiscectomy and lumbar intervertebral fusion, in The Adult

Spine: Principles and Practice (Frymoyer J W, Ducker T, Hadler N, et al. eds.), New York,1996, pp. 2037–2047.

63. Kambin P. Arthroscopic techniques for spinal surgery, in Operative Arthroscopy (McGinty JB,Casperi RB, Jackson RW, Poehling GG, eds.), Lippincott-Raven, Philadelphia, 1996, pp.1215–1225.

64. Kambin P, Gennarelli T, Hermantin F. Minimally invasive techniques in spinal surgery:current practice. Neurosurg Focus 1998;4(2):l–l0.

65. Shepperd JAN. Percutaneous and minimal intervention spinal fusion. in ArthroscopicMicrodiscectomy Minimal Intervention in Spinal Surgery (Kambin p, ed), Urban &Schwarzenberg, Baltimore, 1991, pp. 127–129.

66. Kambin P. Principles and experiences with monoportal arthroscopic microdiscectomy. Videopresentation of endoscopic laminatomy and foraminatomy. Bulgrist Orthopaedic Hospital,University of Zurich School of Medicine, Zurich, Switzerland, December 2, 1993.

67. Foley KT, Smith MM. Microendoscopic discectomy. Techniques Neurosurg 1997;3(4):301–307.

68. Bozzini PD. Lichleiteroder beschreibank einer einfachen vorrichtung und inhern anwendsungsurerlechtung innerer hohlen und zwischenraume des lebenden animalischen korpers.Weimer,1807.

69. Jacobaeus HC. Possibility of the use of the cystoscope for investigations of serious cavi-ties. Munch Med Wochenschrift 1910;57:2090–2092.

70. Burman MS. Myeloscopy or the direct visualization of the spinal canal and its contents.J Bone Joint Surg 1931;13:695–696.

71. Pool JL. Direct visualization of dorsal nerve roots of the cauda equina by means of themyeloscope. Arch Neurol Psychol 1938;39:1308–1312.

72. Pool JL. Myeloscopy: diagnostic inspection of the cauda equina by means of the endo-scope. Bull Neurol Inst NY 1938;7:178–189.

73. Ooi Y, Morisaki N. Intrathecal lumbar endoscope [1st report in Japanese]. Clin OrthopSurg 1969;4:295.

74. Ooi Y, Satoh Y, Mita F, Schaffer JL. Myeloscopy and endoscopic nucleotomy, in SpineCare (White AH, Schofferman J, eds.), CV Mosby, St. Louis, 1995, pp. 1009–1073,

75. Hausmann B, Forst R. Nucleoscope: instrumentation for endoscopy of the intervertebraldisc space. Arch Orthop Trauma Surg 1983;102:37–59.

76. Kambin P, Zhou L. History and current status of percutaneous arthroscopic disc surgery.Spine 1996;21(24).

26 Kambin

77. Kambin P, O’Brien E, Zhou L, Schaffer JL. Arthroscopic microdiscectomy and selectivefragmentectomy. CORR 1998;(347):150–167.

78. Mathews HH, Kyles M , Lang BH, Fiore SM, Gordon CL. Spinal endoscopy: indications,approaches and applications. Orthop Trans 1995;19:219.

79. Kambin P, Casey K, O’Brien E, et al. Transforaminal arthroscopic decompression of lateralrecess stenosis. J Neurosurg 1996;84:462–467.

80. Destandau J. A special device for endoscopic surgery of the lumbar spine. Neurol Res1999;21:39–42.

81. Roh SW, Kim D, Cardoso A, Fessler RG. Endoscopic foraminotomy using Spine Med Sys-tem in cadaveric specimens. Spine 2000;25(2):260–264.

82. Mathews HH, Evans MT, Molligan HJ, Long BH. Laparoscopic diskectomy with anteriorlumbar interbody fusion: a preliminary review. Spine 1995;20:1797–1802.

83. Connolly PJ, Yuan HA, Kolata RJ, Clem MF. Endoscopic approach to the lumbar spineusing the insufflation technique. in Atlas of Endoscopic Spine Surgery (Regan, McAffee,Mack, eds.), Quality Medical Publishing, St. Louis, 1995.

84. Horowitz MB, Moossy JJ, Julian T, Ferson PF, Huneke K. Thoracic discectomy usingvideo assisted thoracoscopy. Spine 1994;19:1082–1086.

85. Mack MJ, Regan JJ, Bobechko WP, Acuff TE. Application of thoracoscopy for disease ofthe spine. Ann Thorac Surg 1993;56:736–738.

86. Obenchain TG. Laparoscopic lumbar diskectomy. a case report. J Laparoendosc Surg1991;1:145–149.

87. Obenchain TG, Cloyd D, Slavin M. Outpatient laparoscopic lumbar discectomy. Descrip-tion of technique and review of first twenty-one cases. Surg Tech Int 1993;2:415–418.

88. Regan JJ. Technique and approach for laparoscopic discectomy and fusion, Atlas of Endo-scopic Spine Surgery (Regan, Mcaffee, Mack, eds). Quality Medical Publishing st. Louis,pp. 130–135.

89. Regan JJ, Mack M J, Picetti GDI, A technical report on video-assisted thoracoscopy inthoracic spinal surgery: preliminary description. Spine 1995;20: 831–837.

History of Lumbar Disc Surgery 27

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

2Arthroscopic and Endoscopic Anatomy

of the Lumbar Spine

Parviz Kambin, MD

INTRODUCTION

The success of arthroscopic and endoscopic spinal surgery hinges by and large on clearvisualization and identification of various anatomical structures. During intracanalicularsurgery, venous bleeding may obstruct the surgeon’s visual field. This must be controlledwith topical anticoagulants, radiofrequency probes, cold irrigation fluid, and a simpleincrease in its inflow.

THORACOLUMBAR FASCIA

The thoracolumbar fascia appears as a heavy band of interwoven, whitish fibers thatlacks a blood supply (Fig. 1).

PARASPINAL MUSCLES

The sacrospinalis, quadratus lumbrorum, and psoas major muscles appear as reddishbundles that are readily visualized under arthroscopic magnification and illumination(Fig. 2). The muscle fibers are moderately vascular. At the end of arthroscopic spinesurgery, it is not unusual to see extravasation of blood through the small incision usedfor insertion of instruments. This bleeding can be controlled by applying compressionto the skin and the paraspinal muscles.

TRIANGULAR WORKING ZONE

The triangular working zone is a safe zone on the posterolateral surface of theannulus adjacent to the spinal canal. It is suitable for safe lodging of instruments dur-ing posterolateral access to the intervertebral disc and the spinal canal (Fig. 3) (seeChapters 3 and 4). The annular surface in the triangular working zone is borderedanteriorly by the exiting root, inferiorly by the proximal plate of the inferior lumbarsegment, medially by the dural sac and the traversing root, and posteriorly by thearticular processes of the adjacent segment (1,2). The annular surface in the triangu-lar working zone is covered with loosely woven globules of adipose tissue (Fig.4A,B). It should be noted that the fatty tissue is relatively stationary and does not

30 Kambin

Fig. 1. Arthroscopic view showing the thoracolumbar fascia as an avascular heavy band ofinterwoven fibers.

Fig. 2. The paravertebral muscles of the lumbar spine are seen as a moderately vascularmuscle bundle.

Anatomy of Lumbar Spine 31

Fig. 3. Illustration of triangular working zone: A, exiting root that forms anterior boundaryof triangular working zone; B, triangular working zone; C, traversing root.

move in and out of the cannula as the patient breathes. Small-caliber veins that mayhave to be coagulated may be observed on the surface of the triangular working zone(Fig. 5A,B). Coarse fibrous bands of the annulus are observed after removal of theadipose tissue (Fig. 6A,B). At times, a thin layer of fibers of the psoas muscle areobserved on the dorsolateral surface of the annulus in the triangular working zone(Fig. 7).

NERVE ROOTS AND ROOT GANGLIA

Both exiting and traversing roots that form the lateral and medial boundaries of the tri-angular working zone are in the path of dorsolaterally inserted instruments and may besubject to insult during intradiscal or extraannular approaches to the lumbar spine. At theonset of arthroscopic disc surgery, the final and proper positioning of the instruments is bestdetermined by accurate placement and documentation of the tip of the inserted instruments(18-gage needle or guide wire) on the annular surface in the triangular working zone. Thefibers of the posterior longitudinal ligament extend laterally into the triangular workingzone and extraforaminal region (4–8) (Fig. 8). These fibers are innervated by branches ofsinu-vertebral nerve and are highly sensitive to palpation by the inserted instruments. Thesuperficial layer of the annulus and the expansion of the posterior longitudinal ligamentmust be adequately anesthetized during the operative procedure and annular fenestration.The exiting root is well protected in the triangular working zone, where it lays under thepedicle in the pedicular notch and is accompanied by the radicular artery, radicular vein,and branches of the sinu-vertebral nerve. The posterior sensory nerve fibers and the ante-rior motor fibers usually join one another prior to their departure from the dural sac. How-ever, at times they part individually from the dura. The posterior sensory root is larger thanthe anterior motor root, and it continues into the fusiform root ganglia and then joins theanterior motor root. Usually the L4 and L5 dorsal root ganglia are intraforaminal and mustbe protected during foraminal access to the intervertebral disc or the spinal canal. The S1root ganglion is likely to be intraspinal.

32 Kambin

Fig. 4. (A,B) Arthroscopic view of annular surface in triangular working zone. Note theloose/woven adipose tissue on the surface of annulus.

The mobility of the traversing and exiting roots at the lower lumbar region allowsthe operative surgeon to retract these structures during both arthroscopic and openspinal surgery. However, limited mobility of the nerve roots in the upper lumbar spineand at the thoracolumbar junction, combined with the bulk of the cauda equina andconus medullaris in the subarachnoid space, demands protection of neural tissue byaccurate positioning of the instruments, clear visualization, and careful handling of theresecting instruments.

Anatomy of Lumbar Spine 33

Fig. 5. Interoperative photo demonstrating presence of superficial veins on surface of annulus.

In a series of T1 imaging studies of the thoracolumbar junction, the end of the conuswas identified at the L1 level in 50% of individuals but at the L2 segment in 20% ofpatients. Therefore, great care must be exercised while performing arthroscopic orendoscopic surgery adjacent to these two segments. The nerve roots appear as pale yellow,multifiber structures under arthroscopic illumination and magnification. Fine vessels arepresent on the surfaces of the nerve root and root ganglia (Fig. 9A–D). The inflamednerve roots are highly sensitive to palpation and compression. The traversing root may be

34 Kambin

Fig. 6. Interoperative photos showing (A) annular surface following removal of adiposetissue and (B) magnified view of avascular fibers of annulus following extraction of superficialadipose tissue.

Anatomy of Lumbar Spine 35

Fig. 7. A thin layer of psoas muscle that covers the triangular working zone is observed andis being removed with the aid of laser light.

Fig. 8. Cadaveric dissection showing fibers of posterior longitudinal ligamentum, which extendsto dorsolateral aspect of annulus, foraminal, and extraforaminal region.

observed adjacent to the dural sac in the medial pedicular line region and is usuallysurrounded by epidural veins and adipose tissue.

VASCULAR STRUCTURES

Vascular structures of epidural and neural tissue play an important role in the patho-physiology of pain that is commonly observed in individuals with symptomatic discherniation, spinal stenosis, and failed back surgery syndrome. It has been shown that

interference with the normal blood flow of delicate neural and epidural venous sys-tems (Fig. 10) may cause venous stasis edema of the nerve root, neural fibrosis, andchronic pain (4,8). This phenomenon is usually observed in patients presenting withsymptomatic disc herniation and spinal stenosis. Segmental arteries, branches from theaorta and the internal iliac artery, provide blood supply to the neural, osseous, andmuscular structures of the spinal column. Branches from the lumbar arteries departfrom the segmental arteries and enter the intervertebral foramen (radicular artery) withthe exiting root to supply the medullary arteries of the spinal cord and the nerve roots.

Segmental arteries distal to L4 originate from the hypogastric arteries that are branchesof the internal iliac arteries (10) (Fig. 11). When the posterolateral approach is used foreither discectomy or anterior column stabilization and the instruments are directed andproperly positioned in the triangular working zone, these segmental arteries are not sub-

36 Kambin

Fig. 9. Endoscopic view of content of (A,B) spinal canal and (C,D) herniated lumbar disc.(Reprinted from ref. 12, with permission [A].)

ject to insult and injury. However, when the instruments are inserted too far anteriorly(vertically), they may penetrate the iliac artery or vein, causing rapid blood loss requir-ing immediate emergency exploration and repair of the injured vessels. By contrast,during laparoscopic or open retroperitoneal or transabdominal spinal surgery, thehypogastric and infraaortic arteries as well as sympathetic ganglia are in the path of theinserted instruments and must be protected.

When the instruments are properly lodged in the triangular working zone, the radicu-lar arteries remain protected under the pedicular notch. However, during decompression

Anatomy of Lumbar Spine 37

Fig. 9. (Continued)

of the lateral recess in the treatment of lateral recess stenosis, particularly when a back-firing laser light is being used, great care must be exercised to protect the vascular struc-tures and their accompanying exiting root in the pedicular notch. Only a small portion ofthe exiting root that is situated between the lateral pedicular line and the superior borderof the transverse process is subject to insult during the posterolateral approach for theremoval of extraforaminal herniations (7). However, positioning the instruments medialto the lateral pedicular line at the onset of operative procedures will help avoid any com-plications (see Chapter 4).

EPIDURAL ADIPOSE TISSUE

Epidural adipose tissue is usually seen as rather large globules of fat that move inand out of the cannula when a patient inhales and exhales. This adipose tissue is sur-rounded by the epidural venous system, which must be controlled when transforaminalintracanalicular access is employed for the removal of a sequestered disc.

38 Kambin

Fig. 10. Photograph demonstrating arteries and venous system of nerve root and A.V.anastomosis (courtesy of Wesley Parke).

Fig. 11. Illustration showing origin of segmental arteries of lumbar spine.

DURAL SAC

The dural sac is seen as grayish tissue surrounded by epidural veins that readilybleed (Fig. 9A,B).

INTRACANALICULAR LIGAMENTS

In a virgin spine, fine and mobile vascular and ligamentous structures on the floor ofthe spinal canal may be seen under endoscopic magnification (Fig. 12A,B). Hoffman’sligament, which extends from the ventral and lateral dura to the posterior longitudinalligament, and the dural ligament, which extends from the dura and traversing root to theposterior longitudinal ligament and periosteum, may be identified following adequatehemostasis.

Anatomy of Lumbar Spine 39

Fig. 12. (A,B) Ligamentous structures of ventral dura and nerve root.

40 Kambin

Fig. 13. (A) Gross anatomy of annulus fibrosus and NP in young cadaver specimen. Thenucleus has been injected with methylene blue. Note, however, the thickness of the annulusfibrosus anteriorly and posterolaterally. (B) Cotton-ball appearance of nucleus in adolescence(arthroscopic view).

Anatomy of Lumbar Spine 41

Fig. 14. (A,B) Arthroscopic view of NP in a 45-yr-old male following interdiscal injection ofdiluted indigo carmine. Note the unstained partially collagenized nucleus mixed with stainedsoft nuclear tissue.

NUCLEUS PULPOSUS

In adolescence the nucleus pulposus (NP) is well contained within the annularring (Fig. 13A). It has a whitish, cotton-ball appearance (Fig. 13B). The nucleusbreaks off and separates easily. It is not liquid and does not flow. Nuclear tissueabsorbs a considerable amount of fluid and has a tendency to swell. By contrast, thenucleus will become dehydrated and partially collagenized in the fifth and sixthdecades of life. A mixture of collagenized tissue and whitish soft nucleus may beobserved within the intervertebral disc in this older group of patients (Fig. 14A,B).In addition, multiple collagenized free fragments floating within the disc space mayalso be seen (Fig. 15A,B).

42 Kambin

Fig. 15. (A,B) Interoperative arthroscopic view of intervertebral disc. Note multiple collage-nized free fragments floating within a degenerated intervertebral disc (intradiscal sequestration).

ANNULUS FIBROSUS

In elderly patients, the nucleus becomes dehydrated and collagenized and begins tomigrate posteriorly through the torn fibers of the annulus (Fig. 16). From a clinician’spoint of view, the dorsolateral migration of the nuclear tissue will present three distinct

Anatomy of Lumbar Spine 43

anatomical and imaging presentations. A partial tear of the annular fibers may cause anintraannular herniation with imaging evidence of a small and gradually developedbulge or protrusion of the annulus (Fig. 17A) (11). By contrast, a subligamentousmigration of the collagenized nuclear tissue is associated with a sudden change in theexternal geometry of the annulus, with an imaging appearance of a distinct herniationhaving smooth borders (Fig. 17B). When the integrity of the posterior longitudinal liga-mentum has been altered, the collagenized nuclear fragments are dislodged into thespinal canal. Imaging studies will reveal a large extradural sequestered fragment withirregular and uneven borders (Fig. 17C).

Fig. 16. Fresh cadaveric specimen of a degenerated disc following injection of methyleneblue. Note the disorganization of the nuclear tissue, annular tear, and migration of the nucleusinto the periphery.

Fig. 17. Schematic drawing demonstrating (A) interannular, (B) subligamentous, and (C)extraligamentous, herniation of nucleus pulposus (H.N.P.).

44 Kambin

Fig. 18. (A) Endoscopic view of capsular ligamentum flavum complex; (B,C) magnifiedview of capsular ligamentum flavum complex, which appears as an avascular structure.

Anatomy of Lumbar Spine 45

Arthroscopic differentiation between the torn annulus and nucleus may be diffi-cult. Intraoperative injection of diluted indigo carmine has a tendency to stain thenucleus and the torn fibers of annular tissue while sparing intact annular fibers.Schreiber and Leu originally introduced this technique for intraoperative tissue dif-ferentiation (12).

CAPSULAR, LIGAMENTOUS FLAVUM COMPLEX

During foraminal or transforaminal intracanalicular surgery (3,4,9), the capsule ofthe facet joints combined with a thickened inflamed ligamentous flavum may interferewith free passage of instruments for surgical removal, or vaporization of the content ofthe spinal canal.

In our experience, removal of the fibers of the posterior longitudinal ligamentuminvariably is adequate and will provide ample access to the ventral surface of thedural sac and traversing root (13) (see Chapter 4). In contrast to the traversing root,the capsule and the ligamentum flavum complex are observed as an avascular tissue(Fig. 18A–C).

POSTERIOR LONGITUDINAL LIGAMENTUM

The ventral surface of the posterior longitudinal ligamentum may be observed whensubligamentous access to a contained disc herniation has been attempted. Visualizationof the ventral surface of the posterior longitudinal ligamentum signifies that adequatedecompression of the nerve root and dural sac has been accomplished. Followingextraction of the torn fibers of the annulus, partial nucleotomy, and creation of a smallcavity ventral to the posterior longitudinal ligamentum, a 30 or 70° arthroscope may beused for visualization of the above structures. In contrast to the annulus, the fibers ofthe posterior longitudinal ligamentum run perpendicular to the vertebral plates and areseen as avascular structures (Fig. 19A–C).

Fig. 18. (Continued)

46 Kambin

Fig. 19. (A) Intraoperative photo demonstrating ventral surface of posterior longitudinalligamentum following evacuation of a contained disc herniation; (B,C) magnified view ofventral surface of posterior longitudinal ligamentum (PLL) shown in Fig. 17A.

Anatomy of Lumbar Spine 47

REFERENCES

1. Kambin P. Percutaneous lumbar discectomy: current practice. Surg Rounds Orthop1988;31–35.

2. Kambin P. Arthroscopic Microdiscectomy: Minimal Intervention in Spinal Surgery R(Kambin P, ed.), Urban & Schwarzenberg, Baltimore, 1991, pp. 67–100.

3. Kambin P. Gross and arthroscopic anatomy of the lumbar spine, in Operative Arthroscopy(McGinty JB, Casperi RB, Jackson RW, Poehling GG, eds.), Lippincott-Raven, Philadel-phia, 1996, pp. 1207–1214.

4. Kambin P, McCullen G, Parke W et al. Minimally invasive arthroscopic spinal surgery.Instruct Course Lect 1997;46:143–161.

5. Kambin P. Arthroscopic microdiscectomy lumbar and thoracic spine, in Spine Care, vol. 2(White A, Schoffeman JA, eds.), Mosby, St. Louis, 2002. pp. 1002–1016.

6. Mirkovik SR, Schwartz DG. Anatomic considerations in posterolateral procedures. Spine1995;20:1965–1971.

7. Kambin P, Brager M. Percutaneous posterolateral discectomy: anatomy and mechanism.Clin Orthop 1987;223:145–l54.

8. Park WW. Anatomy of spinal nerve and it’s surrounding structures. Semin Orthop1991;6(2):65.

9. Mathews HH, Kyles MK, Lang BH, Fiore SM, Gordon CL. Spinal endoscopy: indications,approaches and applications. Orthop Trans 1995;19:219.

10. Park WW, Whalen JL, VanDemark RE, Kambin P. The infra-aortic arteries of the spine:their variability and clinical significance. Spine 1994;19(1):1–5.

11. Schaffer JL, Kambin P. Minimally invasive spine surgery. Textbook of RheumatologyUpdate 1994;9:2–12.

12. Schreiber A, Leu HJ. Biportal percutaneous lumbar nucleotomy, in Spinal Surgery (KambinP, ed.), Urban & Schwarzenberg Baltimore, 1991, pp. 67–100.

13. Hermantin FU, Peters T, Quartaro L, Kambin P. A prospective, randomized study compar-ing the results of open disectomy with those of video-assisted arthroscopic microdisectomy.J Bone Joint Surg Am 1999;81:958–965.

Fig. 19. (Continued)

3Instruments and Surgical Approaches for Minimally

Invasive Spinal Surgery Via Posterolateral Access

Parviz Kambin, MD

INSTRUMENTS

A variety of instruments have been marketed by at least five industries. All of themshare the already published principles (1–4):

1. Precise positioning of needle or guide pin under fluoroscopic control in the triangularworking zone.

2. Use of a blunt-ended cannulated obturator or soft-tissue dilator over the previously positionedguide pin. The blunt end of the obturator will have a tendency to bypass the traversing orexiting root as it descends toward the triangular working zone.

3. Final positioning of a working cannula that is passed over the previously positioned soft-tissuedilator.

The instruments include the following:

1. An 18-gage needle, 15 cm (6 in.) in length (Fig. 1).2. A blunt-end cannulated obturator with a 4.9-mm outer diameter (od).3. Cannulas: A number of cannulas are available and have been used by various surgeons for

arthroscopic spinal surgery. The round universal cannula has a 6.4-mm od that provides aninner diameter (id) working area of 5 mm. Recently, a round, bevel-ended cannula has beenintroduced and utilized by some surgeons (T. Yeung, personal communication).

Two sizes of oval cannulas are available (Fig. 1): a cannula with a 6.4 × 10 mm od thatprovides a 5 × 8 mm id working area; and a larger oval cannula, primarily used for arthro-scopic anterior column stabilization, that provides a 10 × 5 mm id working area (4,5)(available from Stryker, Howmedica, Osteonics, Allendale, NJ).

I and others have used a series of telescoping oval cannulas (Fig. 2) in order to maxi-mize access to the intervertebral disc for the introduction of bone grafts. These cannulaswere designed to fit within the dimensions of the triangular working zone. Because theheight of the triangular working zone is somewhat limited by the height of the intervertebraldisc, oval cannulas will not exert undue traction to the neural structures when it lies withinthe triangular working zone. A 10- and a 12- mm oval cannular jig permits parallel insertion ofboth a cannulated obturator and a half- or full-moon auxiliary obturator in preparation ofinsertion of 10- or 12-mm cannulas (Fig. 3A,B).

4. Triphens: Two sizes of triphens are available that can be used for annular fenestration orcutting and removing osteophytes. Both triphens fit within the lumen of the 5-mm cannula,which was previously described (Fig. 4). Three working scopes are available in 0, 8, and

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

20° models. The working channel of these instruments accommodates small-caliber for-ceps, knives, curettes, a radiofrequency coagulator, and palpating instruments (Fig. 5).

5. Arthroscopes: A variety of 0, 30, and 70° arthroscopes are available for intradiscal andperiannular surgery. These scopes may be used with the appropriate irrigation sheaths inconjunction with round or oval-shaped cannulas (Fig. 6).

6. Straight-upbiting and flexible-tipped forceps (Fig. 7).7. Articulating suction forceps. 8. Deflecting tube that permits 40° dorsal angulation of the flexible-tipped forceps when it is

fully inserted into the deflecting tube. 9. Variety of powered trimmer blades (disc shavers).

10. Monopolar or bipolar radiofrequency coagulator.11. Video equipment that is available in most operating room settings.

50 Kambin

Fig. 1. Instruments for spinal surgery. Shown are from left to right a 5 × 10 mm id ovalcannula, a 5 × 8 mm id oval cannula, a 5 × 5 mm id cannula, a cannulated obturator, an 18-gageneedle, and guide wire.

Fig. 2. Schematic drawing of a series of telescopic cannulas.

Instruments and Surgical Approaches 51

Fig. 3. (A) Jig for insertion of two cannulated obturators in preparation of insertion of 5 × 10mm oval cannula. (B) A cannulated obturator and a half-moon obturator are inserted into theintervertebral disc in preparation of insertion of a 5 × 8 mm id oval cannula.

Fig. 4. Shown are from top to bottom a 3-mm triphen, a 5-mm triphen, and a universal 5-mm-idcannula.

52 Kambin

Fig. 5. Working-channel scope with a variety of instruments that can be used in conjunctionwith working scope.

Fig. 6. Three video discoscopes are available: 0, 30, and 70°. The scopes are inserted into thescope sheath, which is used for inflow of saline solution. The scope and the scope sheath assemblyfit within the lumen of the 5-mm-id cannula, as well as of the oval cannulas.

SURGICAL APPROACHES

A variety of approaches have been used to access the intervertebral discs and vertebralbodies in the lumbar region. Selection of the surgical approach is largely dictated by thesite of the pathology and the experience of the operating surgeon with the particularapproach.

Instruments and Surgical Approaches 53

Fig. 7. Additional instruments used for arthroscopic and endoscopic spinal surgery. From topto bottom, upbiting forceps, straight punch forceps, deflecting tube and flexible-tip forceps,punch forceps, deflecting suction forceps, and a variety of trimmer blades (disco shavers).

Uniportal Approach

A 5 × 5 mm or 5 × 8 mm id cannula is commonly used for uniportal access to the her-niated disc fragments in the lumbar region (1,2,5) (Fig. 8A,B). A larger-diameter cannulamay also be telescoped over the 5 × 5 mm id cannula to provide wider access to the inter-vertebral disc. A working channel scope is usually used with a 5-mm-id cannula. However,a 5 × 8 mm id cannula provides additional advantages by simultaneously accommodating a0 or 30° arthroscope, larger forceps, and resecting instruments (3–5) (Fig. 9A,B).

The uniportal approach is commonly utilized for retrieval of paramedial, small central,foraminal, and extraforaminal herniations (see Chapter 4). Uniportal use of a 5 mm idcannula also provides adequate access for a transforaminal approach to the spinal canal(see Chapter 4).

Bilateral Biportal Approach

In the bilateral biportal technique, two cannulas are inserted contralaterally from theright and left sides of the intervertebral disc (Fig. 10A,B). A small cavity is createdposteriorly adjacent to the spinal canal by resection of nuclear tissue and torn fibers ofthe annulus and the posterior longitudinal ligament. Communication between the rightand left portals is essential to establish free inflow of saline into one portal and outflowfrom the opposite portal. A radiofrequency probe has been most helpful in vaporizingnuclear tissue and establishing communication between the two portals (2,3,5). Adequate

54 Kambin

Fig. 8. (A) Properly positioned cannula at medial pedicular line for access to content of spinalcanal at L2-L3 level; (B) lateral interoperative fluoroscopic view of position of tip of cannulaadjacent to posterior boundary of L2-L3.

visualization is accomplished by insertion of a 70 or 90° arthroscope from one portaland introduction of articulating or upbiting forceps from the opposite portal. The bilat-eral biportal approach is suitable for removal of large central herniations and sequesterednonmigrated disc fragments. Following retrieval of a sequestered disc fragment, theventral surface of the dura is usually visualized with the arthroscope that is inserted intothe intervertebral disc. Retrieval of a sequestered fragment from the spinal canal is usually

Instruments and Surgical Approaches 55

Fig. 9. (A) Interoperative fluoroscopic examination showing position of 5 × 8 mm id oval cannulaat L5-S1 in anteroposterior projection. (B) Flexible-tip forceps and 30° arthroscopes are used inconjunction with a 5 × 8 mm id oval cannula for visualization and removal of larger disc fragments.

followed by a discharge of venous blood into the disc cavity. This blood has a tendencyto coagulate when the instruments are removed.

In addition, bilateral biportal access is necessary for anterior column stabilizationwhen a minimally invasive posterolateral approach is employed (see Chapter 5).

Unilateral Biportal Approach

In the unilateral biportal approach, two parallel or converging cannulas are positionedin the triangular working zone with the aid of specially designed jigs. The two parallel

56 Kambin

Fig. 10. (A) Interoperative lateral fluoroscopic examination of biportal access to L3-L4intervertebral disc. Note the posterior position of the instruments adjacent to the spinal canal.(B) Anteroposterior fluoroscopic examination of position of cannulas shown in (A). Note that a5 × 8 mm id oval cannula is used on the left and a 5-mm-id cannula on the right. The cannulasare positioned for removal of a centrally located sequestrated disc herniation.

Instruments and Surgical Approaches 57

Fig. 11. Schematic drawing of unilateral biportal access to foramen utilizing speciallydesigned jig.

Fig. 12. (A) Intraoperative endoscopic review demonstrating how high-speed diamond burris being used for laminatomy. (B) Intraoperative photograph following laminatomy and partialfacetectomy. Note the dural sac at the top of the photograph and the herniated disc in the axillaof the traversing nerve root. (C) Schematic drawing of (B).

58 Kambin

Fig. 12. (Continued)

obturators then may be replaced with an oval-shaped cannula that provides wider accessto the intervertebral disc (4,5) (Figs. 1 and 3). The 5 × 8 mm id cannula (Fig. 9A,B) issuitable for intradiscal or subligamentous access to a disc herniation. It also may beused for retrieval of extraforaminal and foraminal herniations. The 5 × 10 mm id cannula(Figs. 1 and 3A) is commonly utilized in the course of arthroscopic anterior columnstabilization. Two converging cannulas are useful for visualization and removal offoraminal and extraforaminal herniations (Fig. 11) (5,6).

Interlaminar Access

Endoscopic interlaminar access to the content of the spinal canal and the lateralrecess follows the principle of already established open operative procedures (see Fig. 18in Chapter 1). Although originally my colleagues and I used a working cannula with aside window (3–5) for insertion of a straight arthroscope, with the availability of 45 and90° angled arthroscopes (Figs. 3–5) in the pursuing years, we were able to introduce thearthroscope and resecting instruments directly from the proximal opening of the can-nula into the surgical field. We have used a high-speed diamond burr for laminatomy(Figs. 3–12A), and the retrieval of disc fragments has been similar to that of known tech-niques used during open procedures.

REFERENCES

1. Kambin P, Gellman H. Percutaneous lateral discectomy of the lumbar spine: a preliminaryreport. Clin Orthop 1983;174:127–132.

2. Kambin P. Arthroscopic microdiscectomy. Arthroscopy 1992;8:287–295.3. Kambin P. Arthroscopic techniques for spinal surgery, in Operative Arthroscopy (McGinty JB,

Casperi RB, Jackson RW, Poehling GG, eds.), Lippincott-Raven, Philadelphia, 1996, pp.1215–1225.

4. Kambin P, Gennarelli T, Hermantin F. Minimally invasive techniques in spinal surgery:current practice. Neurosurg Focus 1998;4(2):1–10.

5. Kambin P. Arthroscopic microdiscectomy, in The Adult Spine: Principles and Practice, 2nd ed.(Frymoyer JW, Ducker T, Hadler N, et al. eds.), Raven, New York, 1996, pp. 2023–2036.

6. Kambin P. Unilateral biportal percutaneous surgical procedures and instrumentation.US patent 5,395,317, October 1991.

Instruments and Surgical Approaches 59

4Herniated Lumbar Disc and Lumbar Radiculopathy

Parviz Kambin, MD

DIAGNOSIS OF LUMBAR DISC HERNIATIONS: INCLUSION AND EXCLUSION CRITERIA

The satisfactory outcome of arthroscopically and endoscopically assisted managementof herniated lumbar discs hinges on accurate diagnosis of the symptom-producing site.Although the mean age of onset of symptoms of disc herniation is around 35 yr, discherniations are occasionally observed in populations older than 60 or younger than 20 yrof age. In younger children, combined slippage of the intervertebral disc and the vertebralplates may be responsible for the presenting symptomatology.

A careful assessment of the patient’s complaints followed by a thorough physicalexamination is the essential and most reliable initial step in arriving at an accurate diag-nosis. A simple questionnaire that includes a pain analog scale will provide a wealth ofinformation to the operating surgeon. When the distribution of pain extends to a levelbelow the knee or involves the calf or lateral aspect of the leg, the surgeon’s attention isdirected toward L5 or S1 root involvement. By contrast, disc herniation in the middle orupper lumbar spine is associated with anterior thigh or groin pain. Similarly, expressionof numbness or “pins and needles” sensations involving the fifth toe or calf area suggestsS1 root compression most likely at the L5-S1 level. However, when similar complaintsare expressed on the dorsum of the great toe and lateral aspect of the leg, attention shouldbe drawn to L5 nerve root involvement.

The presence of midline localized tenderness at the index level, sciatic tilt, and ten-derness over the sciatic notch are objective confirmative findings. A thorough neuro-logical examination to assess reflex and sensory and motor abnormalities is essential.Absence of the achilles reflex suggests an S1 root compression. Unilateral absence ofthe tibialis posterior reflex should draw attention to L5 root involvement. The patellarreflex may be absent when the L3 or L4 nerve roots are compressed.

Positive tension signs (Fig. 1A), which include a positive straight-leg-raising,Lazarevic test (Fig. 1A) (1); positive Lasegue sign (Fig. 1B,C) (2,3); or positive bowsling sign (4), comprise one of the four essential criteria to consider in surgical manage-ment. A positive contralateral straight-leg-raising test suggests a midline or paramedialherniation. A positive reverse Lasegue sign or femoral stretching test indicates pressureon the L3 or L4 nerve roots.

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

62 Kambin

Fig. 1. (A) Straight-leg-raising test described by Lazarevic. While the leg is being raised andthe knee is kept in full extension, the ankle is dorsoflexed. The patient’s radicular pain is repro-duced. (B) The Lasègue test consist of two phases. The hip and the knee on the symptomaticside are first flexed. If the patient has hip disorder, the pain is reproduced. In phase two of thetest, the leg is raised while the hip and knee of the affected side are kept in full extension. (C)Bow-sling sign described by Cram. In the sitting position, the leg is raised while the knee is keptin flexion. Then the knee is gradually brought into full extension while the examiner’s handapplies pressure to the sciatic trunk in the popliteal region.

Nerve root tension may be tested with the patient standing erect while the examinerevaluates the presence or absence of points of tenderness, muscle rigidity, and range ofmotion of the lumbar spine. In the standing position, the patient is instructed to lean for-ward and slightly bend laterally toward the symptomatic side while holding the kneesin full extension. When a paramedial herniation is present, forward flexion with a slightlateral bending may reproduce the radicular symptoms (Fig. 2) (5,6).

Individuals with a lateral disc herniation usually complain of unilateral leg pain thatmay be associated with positive tension signs and with sensory and motor deficits. Bycontrast, patients with a central or slightly paramedial disc herniation may have bilat-eral symptoms or alternating symptomatology affecting one lower extremity or anotherfrom time to time. The latter group of patients is more likely to have positive ipsilateralor contralateral straight-leg-raising tests.

Compression of the exiting root in the foraminal canal may be related to a containedinterannular disc protrusion. This type of herniation invariably extends to the lateralboundary of the spinal canal, thus causing ipsilateral compression on both the travers-ing and exiting roots (Fig. 3). A sequestered foraminal herniation may migrate awayfrom the intervertebral disc, caudal or cephalad to the pedicular region.

At times a sequestered intracanalicular disc herniation migrates laterally into theforaminal canal and causes compression on the exiting root at the index level. Foraminalherniations usually have an acute onset with intense radicular symptoms. However,low-back pain may be absent or minimal.

Herniated Lumbar Disc 63

Fig. 1. (Continued)

64 Kambin

Fig. 2. As described by Kambin: in the standing position, the patient is instructed to leanforward and slightly bend toward the symptomatic leg while the examiner maintains thepatient’s knee in full extension. (Reprinted from ref. 43, with permission.)

Fig. 3. Schematic drawing demonstrating how a foraminal herniation can cause compressionon both exiting and traversing roots.

When the nerve root ganglion has been compressed in addition to the above, patientspresent signs and symptoms of sympathetic nerve involvement, such as severe skinhypersensitivity to touch or a sense of coldness or excess heat in the involved limb.

As with foraminal herniations, compression and tension on the exiting root in theextraforaminal region is usually associated with intense radicular symptomatology withminimal or no low-back pain.

The history and physical examination should also include a thorough medical andpsychological evaluation combined with a work history and history of drug use ordependency. Management of individuals with a history of long-term use of narcotic-basedmedication is best accomplished by obtaining a preoperative psychological and painmanagement consultation.

In summary, the inclusion criteria for arthroscopic and endoscopic management ofherniated lumbar discs to treat nerve root compression are similar to the acceptedcriteria for open laminotomy and discectomy. They include failure to respond to a well-designed and executed nonoperative management, correlative dermatomal distributionof pain in the lower extremity, positive tension signs, and positive correlative imagingstudies (Fig. 4).

At this state of the art, patients who present with signs and symptoms of caudaequina syndrome should be excluded from arthroscopically and endoscopically assistedmicrodiscectomy.

Transforaminal access to a large central herniation or sequestered disc at L5-S1 maybe difficult in individuals with high iliac crests. However, these herniations may beaccessed laterally through a circular fenestration that is made through the iliac crest.Instruments are passed through this fenestration into the intervertebral disc. Individualswith signs and symptoms of lateral recess stenosis may also successfully be treatedwith arthroscopic techniques. With the availability of bipolar radiofrequency probesand effective topical coagulators, adequate epidural hemostasis can be obtained and

Herniated Lumbar Disc 65

Fig. 4. The four prerequisites for surgical intervention in treatment of herniated lumbar discs.

sequestered disc herniations can be extracted via endoscopically assisted posterolateraltransforaminal access to the spinal canal.

ADVANTAGES OF ARTHROSCOPIC AND ENDOSCOPIC DISC SURGERY

Improvement of Visualization

The availability of small-caliber, rigid-rod endoscopes; three-dimensional cameras;and developing image modification technology has provided operating surgeons witha superior means of visualization and tissue differentiation than the naked eye andmicroscope. The “aquarium effect” of the technology provides clearer and larger imagesof anatomical structures.

Although the microscope permits visualization of the dorsolateral aspect of thecontents of the spinal canal, spinal endoscopy via a posterolateral approach makes itpossible to visualize the medial, lateral, and ventral surfaces of the nerve root and thedural sac without undue manipulation and retraction of neural tissue.

Lowered Incidence of Reherniation

Recurrent herniation following open laminotomy and discectomy is not uncommon.Atken and Bradford (7) reported an incidence of reherniation up to 24%. Balderstonet al. (8) reported recurrent herniation of 12% in two groups of patients who underwentopen discectomy with simple fragment extraction or fragment removal and curettage ofthe disc space. They reported similar outcomes in both groups. Recurrent disc hernia-tions at the site of previous surgery may be diagnosed with enhanced magnetic reso-nance imaging (MRI) studies. However, many of these herniations are asymptomaticand do not require surgical management. During open translaminar discectomy, annularfenestration is performed at the apex of the herniation. Therefore, the containing abilityof the annular ring is further weakened. This invites expulsion of nuclear tissue into theepidural space, particularly when the spine is exposed to flexion and rotational forcesduring the patient’s work and activities. The posterolateral annulus appears to be a moredesirable area for annulotomy. The natural axial width and intact fibers of the postero-lateral annulus (Fig. 13A in Chapter 2) combined with its inherent contractibility mayminimize the incidence of reherniation through the surgically induced annular fenestra-tion. The anatomical position of the facet joints also inhibits undue transmission ofexternal forces to the posterolateral boundary of the annulus fibrosis, therefore limitingexpulsion of nuclear tissue through the posterolateral annular defect.

In an animal model, Hampton et al. (9) reported on the healing potential of a surgi-cally induced defect in the annular fibers of 10 dogs. The dogs were sacrificed within3–12 wk postoperatively. Dissection of the surgical site demonstrated that the defectwas filled with a solid plug of fibrous structures. Postoperative imaging studies by my col-leagues and I on patients who had undergone percutaneous posterolateral discectomyconfirmed these findings. Markolf and Morris (10) reported a decrease in compressivestiffness and an increase in creep and the relaxation rate of the intervertebral disc incadaveric specimens that were exposed to annular fenestration and then followed byexposure of the spinal unit to compressive forces. In younger specimens, extrusion ofnuclear tissue had a tendency to seal off the annular defect and to restore normal functionof the spinal unit.

66 Kambin

Reduction in Incidence of Nerve Root Tethering and Formation of Epidural Scar Tissue

Although perineural and epidural scar tissue may not be pain-producing structures,tethering of the nerve root and dural sac has a tendency to inhibit smooth mobility andgliding of these structures in flexion and extension (Fig. 5A,B; see also Figs 9 and 12 inChapter 2). Nerve root tethering may be responsible for recurrence of sciatic pain when

Herniated Lumbar Disc 67

Fig. 5. (A) Interoperative endoscopic view of patient who had laminotomy and discectomy2 yr earlier. Note the massive epidural and perineural scar formation. (B) Endoscopic view ofvirgin spine showing clear visualization of traversing root and epidural space.

the patient has resumed normal physical activities following the surgical procedure(11). Epidural application of a fat graft, Gelfoam, and other artificial materials hasproven to be unsatisfactory in the prevention of epidural and perineural fibrosis afteropen spine surgery. I have emphasized avoiding manipulation of the content of the spinalcanal and suggested subligamentous access to contained and nonmigrated extraligamen-tous herniations (12).

Protection of Epidural Venous System and Minimization of Nerve Root Trauma

The delicate venous system of nerve roots may be further traumatized by undue intra-operative manipulation and retraction. Haaland et al. (13), Park (14,15), and other inves-tigators (16) have emphasized the importance of patency of neural venous systems in theprevention of venous stasis, which is invariably followed by neural edema, ischemia,and the development of pain (Fig. 6A) in a clinical setting. The presence of neural edemahas been objectively demonstrated with postoperative MRI studies following an exerciseprogram (Fig. 6B).

Maintenance of Integrity of Paraspinal Muscles

During arthroscopic or endoscopic spinal surgery, the paraspinal muscles, namely theerector spinalis, sacrospinalis, quadratus lumbrorum, and psoas major, are not severed,stripped, or retracted. A small soft-tissue dilator with a 4.9-mm outer diameter (od) has atendency to separate the muscle fibers and descend toward the annulus at the index level.This reduces the postoperative morbidity and eliminates potential denervation and mus-cle injury (17–20). The derangement of the muscle fibers and massive scar formation maybe readily observed in postoperative MRI studies of patients who have been exposed totraditional open spinal surgery (Fig. 7).

Maintenance of Spinal Stability

When posterolateral arthroscopically or endoscopically assisted disc extraction isattempted, the facet joints and bony structures are not disturbed. Therefore, the incidenceof postoperative instability, spondylolisthesis, and rapid collapse and narrowing of the discspaces is reduced (21).

Facilitation of Postoperative Imaging Studies

When a subligamentous approach for removal of a disc herniation is utilized, thecontents of the spinal canal are not manipulated nor disturbed. The absence of epiduraland perineural fibrosis (Fig. 8A–D) facilitates accurate postoperative imaging evalua-tion of the contents of the spinal canal if it becomes necessary (22). My clinicalobservation has been that even when transforaminal access for retrieval ofsequestered fragments under fluid medium was attempted, the incidence of postoper-ative scar formation appeared to be less prevalent than with open laminotomy anddiscectomy.

Facilitation of Accurate Intraoperative Documentation

The entire operative procedure may be documented for future reference or teachingpurposes via either intraoperative photography or videotape.

68 Kambin

Herniated Lumbar Disc 69

Fig. 6. (A) Schematic drawing demonstrating how venous obstruction of nerve root andneural edema can become symptom producing. (B) Postlaminotomy MRI study of surgical sitefollowing exercise program associated with recurrence of symptoms and MRI evidence of ipsi-lateral neural edema.

Cost-Effectiveness

The use of regional anesthetics during arthroscopic and endoscopic spinal surgerycombined with minimal postoperative morbidity has eliminated the need for hospital-ization and lengthy postoperative rehabilitation. Currently, most minimally invasiveoperative procedures are being performed in short-procedure units on an ambulatorybasis. This has contributed to the cost-effectiveness of minimally invasive spinalsurgery.

ABCs OF OPERATIVE TECHNIQUES

Arthroscopic or endoscopic spinal surgery via a posterolateral approach is performedthrough the triangular working zone located on the posterolateral annulus (5,23,24).Considering that the intervertebral disc is an amphiarthrosis when the surgery is beingperformed via an intradiscal approach, the term arthroscopic surgery is applicable.Some investigators have used the term endoscopic spinal surgery to describe the pos-terolateral approach to the contents of the spinal canal. However, note that the spinalcanal is not a cavity, so the term extra-articular or periannular arthroscopic discectomymay be more appropriate.

Choice of Operating Room Table

The operating room table for arthroscopic spinal surgery must be radiolucent andrelatively narrow so that the C-arm can be rotated from the anteroposterior (AP) to thelateral projection with minimal risk of contaminating the surgical field.

A pacemaker extension may be attached to the available operating room table andused for minimally invasive spinal surgery. A disadvantage of this arrangement is the

70 Kambin

Fig. 7. Post open surgery MRI study of lumbar spine demonstrating disruption andderangement of muscular tissue.

potential of inadvertently tilting the pacemaker extension and the table, particularlywhen an overweight patient is positioned for the surgery. In addition, the anesthesiolo-gist may have difficulty reaching and monitoring the patient’s status from the top ofthe table.

Herniated Lumbar Disc 71

Fig. 8. (A) Preoperative axial MRI demonstrating sequestrated disc herniation at L4-L5;(B) sagittal MRI findings shown in (A); (C) postoperative axial MRI study shown in (A) demon-strating annular defect at site of disc herniation and evidence of fibrovascular invasion at site ofextracted disc fragment; (D) sagittal view of findings shown in (C).

Fracture tables have been used for minimally invasive spinal surgery. The extensionof these tables used for positioning of the lower extremities is narrow and suitable forpositioning the radiofrequency frame and the patient. This provides ample space forrotation of the C-arm intraoperatively. The AMSCO 3080 table appears to be the tableof choice for arthroscopic spinal surgery. It is available in most operating rooms. Thistable has a long extension that is designed for positioning the lower extremities duringsurgery. During arthroscopic spinal surgery, we rotate the table and position the radiolucent

72 Kambin

Fig. 8. (Continued)

frame and the patient’s trunk and head on the distal end of the table. This allows freerotation of the C-arm around the patient. To further facilitate intraoperative movementof the C-arm, it is advisable to remove the foam mats from the top of the table and placethe frame directly on the table.

Radiolucent Frame

Most arthroscopic spinal procedures are performed when the patient is in a proneposition. This positioning becomes more critical when biportal access to the interverte-bral disc is utilized. The available bolsters (US Medical, Paoli, PA) are comfortable andwell padded. They provide ample room for the rib cage and adequate support for thepatient’s iliac crest and anterior superior iliac spine, thereby allowing reversal of lum-bar lordosis (Fig. 9A), slight flexion of the hip joints, and widening of the dimensionsof the foramen, so that the inserted instruments can be passed into the foramen andtriangular working zone.

Prior to positioning of the patient, the bolsters of the frame should be adjusted to thesize of the patient. The proximal ends of the bolster should be placed far enough apartto provide space for the patient’s rib cage. In addition, the distal ends of the bolstersshould be brought together so that they provide adequate support under the patient’siliac crest (Fig. 9B).

C-Arm

Although a number of C-arms are available in the marketplace, a C-arm with awide arch is most desirable for arthroscopic and endoscopic spinal surgery. This per-mits unrestricted rotation of the C-arm for both AP and lateral exposure. Whetherthe C-arm should be positioned next to the operating surgeon or on the opposite sideof the table is not important. I prefer positioning the C-arm on the symptomatic sideof the patient while the operating surgeon stands on the opposite side (Fig. 10). Ifthe patient presented with right sciatica, the C-arm is positioned on the right sideand the instruments are inserted from the right posterolateral access. This permits afavorable radiographic visualization and a better approach to the compressed andinflamed nerve root at the index level. It is advisable to adjust the height of the oper-ating room table and the C-arm at the onset of the procedure to make certain that theC-arm can be moved from the lateral to AP position without obstruction and potentialcontamination.

The C-arm should be covered with a sterile sheet or plastic and secured with a Klingbandage by wrapping it around the C-arm. The C-arm should always rotate under thetable, rather than on top of the operating table.

To have a reproducible AP and lateral image, the C-arm must be rotated 90° forlateral X-ray exposure. One should make certain that the sterile sheet or plastic doesnot prevent the full 90° rotation of the C-arm.

Positioning of Patient

Prone positioning of the patient on an adjustable radiolucent frame is most desirablefor arthroscopic or endoscopic spinal surgery. In this position, any inadvertent move-ment of the patient during the surgery can be readily corrected by repositioning thepatient, thus allowing repeated reproducible interoperative imaging studies (5,24–26).

Herniated Lumbar Disc 73

74 Kambin

Fig. 9. (A) Proper positioning of patient on radiolucent frame. Note the flattening of the lumbarlordosis. (B) Adjustment of bolsters to provide support under anterior, superior iliac spine ofpatient and room for expansion of chest wall. (Frame produced by USA Medical, a division ofUniversal Services Associates, Inc., Broomall, PA 19008.)

In addition, prone positioning of the patient is essential for bilateral biportal access to theintervertebral discs.

It is essential that adequate padding be placed under the kneecaps while the patientis lying in the prone position. If straps are placed on the dorsum of the thighs to securethe patient on the operating room table, adequate padding should be provided to makecertain that the peroneal nerve is well protected. Lateral positioning of the patient maybe attempted when uniportal access to the L5-S1 intervertebral disc is being attempted.Maintaining the symptomatic side of the patient up and wedging the table willwiden the disc space at the index level. This positioning allows the iliac crest to moveaway from the skin entry site (Fig. 11). In a clinical setting, the lateral positioning ofthe patient may be practical in younger and more flexible patients. When lateral posi-tioning is used for arthroscopic disc surgery, the patient must be secured on the operat-ing room table with bolsters and adhesive tape. This prevents inadvertent movementand rotation of the trunk during surgery.

Anesthesia

A majority of arthroscopic and endoscopic discectomies can be performed underlocal anesthesia with the assistance of an anesthesiologist using conscious sedation.

Herniated Lumbar Disc 75

Fig. 10. Drawing demonstrating position of patient and operating room setup for minimallyinvasive spinal surgery.

However, general anesthesia may be employed when ample experience with the opera-tive technique and visual differentiation of the anatomical structures are gained. At theonset of the operative procedure, a 20-gage needle is used and skin, subcutaneous tis-sue, and muscle layers are infiltrated with 1% Xylocaine® solution. This is then fol-lowed by insertion of an 18-gage needle 6 in. in length. The needle is directed towardthe foramen at the index level. Injection of a mixture of 1 cc of fentanyl (Elkins Sinn,Cherry Hill, NJ) with 3 cc of saline solution into the foramen as soon as the correctposition of the tip of the needle in the triangular working is established may minimizeinteroperative pain and reduce the incidence of postoperative development of dysesthe-sia, which has been reported in the literature (12–27). Considering that the expansion ofthe posterior longitudinal ligamentum in the foramen and the extraforaminal region ishighly innervated, prior to annulotomy the annular fibers should be anesthetized withXylocaine solution via a long 18-gage needle inserted through the previously posi-tioned cannula.

Prophylactic Antibiotic Therapy

In our practice, we have used 1000 mg of AncefTM (cefazolin sodium) intravenouslypreoperatively in most of our patients. This is usually administered in the waiting areabefore the patient is transferred to the operating room. Two additional doses of Ancef areadministered at 8 and 16 h postoperatively. Because most of the discectomies are per-formed on an ambulatory basis, patients are instructed to take two additional 1000-mgdoses of KeflexTM (cephalexin hydrochloride) orally every 8 h after their discharge.When a patient presents with a history of sensitivity to cefazolin, this antibiotic isreplaced with 500 mg of vancomycin injected preoperatively.

76 Kambin

Fig. 11. Lateral positioning of patient on operating room table for access to L5-S1 inter-vertebral disc.

In addition, we use a dilute solution of gentamicin (Elkin’s Sinn) (80 mg in 1000 ccof normal saline) for final irrigation of the surgical site prior to withdrawal of theinstruments and final closure.

Identification of Surgical Site

The surgical site must first be identified with an AP and a lateral X-ray study. Whilethe patient is positioned prone, a long needle or a Steinman pin is placed on the top ofthe skin over the disc space at the index level. The C-arm is then moved into place forAP radiographic exposure. The C-arm may be slightly tilted cephalad to direct the X-ray beams into the intervertebral disc at the surgical site. By moving the needle orSteinman pin cephalad or caudally, the surgical level is identified and the skin ismarked accordingly by drawing a transverse line (Fig. 12). One should make certainthat the patient is not tilted to the side and is symmetrically positioned on the frameand the table. The spinal processes should be seen in straight alignment on the midline(Fig. 13).

In addition, some surgeons prefer to place an opaque instrument at the patient’s side.They take an X-ray to determine the direction of the intervertebral disc in the lateralprojection. This helps to direct the needle toward the index level (Fig. 14).

Insertion of Needle

Proper positioning of the tip of the needle in the triangular working zone may betime-consuming. However, it is essential for successful evacuation of herniated discfragments and the final outcome of minimally invasive spinal surgery.

Herniated Lumbar Disc 77

Fig. 12. Interoperative fluoroscopic examination of surgical site. A needle is placed on theskin at the L4-L5 level. An Ap X-ray shows proper positioning of the needle. The skin is markedaccordingly prior to surgery.

78 Kambin

Fig. 14. Schematic drawing showing proper direction of inserted needle in lateral projection.The needle is held on the side of the patient and the lateral X-ray is exposed.

Fig. 13. Schematic drawing demonstrating that patient is symmetrically positioned on table.The spinal processes are well aligned on the midline. A–C represent the skin entry site. The siteof annular lodging of the instruments is shown at the midpedicular line.

Proper needle placement encompasses two essential steps: (1) selection of the skinentry site; and (2) proper positioning of the needle in the triangular working zone, thesite of annular fenestration (Table 1).

Skin Entry Site

Various techniques have been used for selection of the skin entry point. The size andweight of the patient certainly influence the choice. In addition, the operative techniqueand location of herniation affect selection of the skin window (Table 1). When transforam-inal access to the contents of the spinal canal is contemplated, it is desirable to positionthe skin entry site more laterally. By contrast, when the evacuation of an extraforaminalherniation is desired, a skin entry point 8 to 9 cm from the midline may be adequate.At times, degenerative and hypertrophic changes in the facet joints prevent proper posi-tioning of the needle in the triangular working zone, and it is therefore necessary to usea skin entry site more laterally to bypass the facet joints. However, far lateral position-ing of the skin window or vertical insertion of the needle may direct the tip of the nee-dle into the peritoneal cavity and cause contamination and complications. With thepatient in the prone position, a needle may be held at the patient’s side, and the tip ofthe needle placed at the center of the intervertebral disc at the index level under lateralfluoroscopic control. The distance from the center of the disc to the skin level then isselected to represent the required distance between the skin entry site and the midline(28). This technique exposes the patient and the operating surgeon to additional radia-tion. Although this technique may be used for central nucleotomy, note that duringarthroscopic or endoscopic fragmentectomy the instruments are positioned posteriorlyadjacent to the spinal canal and not in the center of the intervertebral disc. Therefore,the accuracy of such measurements has been questioned. Generally, in the lumbarspine, a skin entry site of about 10–12 cm from the midline is appropriate for insertionof the needle. If the needle cannot be properly positioned on the annulus, then the nee-dle may be withdrawn, a more lateral entry site selected, and the needle reinserted.

Herniated Lumbar Disc 79

Table 1Selection of Skin Entry Site

11–12 cm from midline 7–9 cm from midline 2–5 cm from midline

• Arthroscopic subligamentous • Discography • Arthroscopic facet discectomy arthrodesis

• Transforaminal endoscopic • Chemonucleolysis • Interlaminar microlaminatomydiscectomy (confirm skin and discectomyentry site via preoperativeprone CT)

• Arthroscopic foraminal • Vertebral body biopsy • Arthroscopic foraminotomydecompression (Ottolenghi-Craig) and partial facetectomy

• Diagnostic endoscopy • Arthroscopic anterior • Minimally invasive(confirm skin entry point column stabilization posterolateral interbody fusion with preoperative CT study and instrumentationof surgical site)

The most reliable method for selecting the skin entry site is a preoperative computedtomography (CT) study of the surgical site with the patient in a prone position (Fig. 15)(5,24,29). This study should always be performed when arthroscopic or endoscopicsurgery in the thoracic spine or at the thoracolumbar junction and transforaminal accessto the spinal canal are planned.

While inserting the needle and directing it toward the triangular working zone, it issafer to position the C-arm for the lateral exposure. To prevent deviation of the beveledtip of the 18-gage needle and to palpate tissue resistance, the needle should be advancedwith a slow rotary movement. Advancement of the needle should be stopped when itstip has reached the superficial layer of the annulus. The needle should not penetrate theannular fibers, and there is no reason to insert the needle into the center of the disc atthis stage of the operative procedure.

It is difficult to ascertain the required angle for insertion of the needle. The size of thepatient, the skin entry site, and the desired site for the annular window dictate the angleof insertion. Conducting a preoperative CT study in the prone position is the only wayto predetermine the required angle of insertion.

It is always safer to insert the needle in a more horizontal plane. This allows palpationof the facet joints with the tip of the needle. Then the needle may be partially withdrawnand reinserted in a more vertical direction, thus bypassing the facet joint and entering thetriangular working zone.

When transforaminal access to the spinal canal for retrieval of a sequestereddisc herniation is attempted and the needle is positioned medial to the midpedicularline, the stylet of the needle should be removed as the tip of the needle is advancedtoward the annular surface. If the spinal fluid begins to drip out, the needle should bewithdrawn and repositioned.

The operating surgeon should pay close attention to the length of the needle that isinserted into the paraspinal muscles. Standard needles used for arthroscopic or endoscopic

80 Kambin

Fig. 15. Preoperative prone CT scan study of surgical site demonstrating desired distance ofskin entry point from midline.

surgery are 6 in. (15 cm) long. If the skin entry point is 11 cm from the midline, at approx11 cm of penetration the tip of the needle should reach the posterior annulus in the triangu-lar working zone and annular resistance should be palpated. If the target is not reachedafter 11 cm of penetration, AP and lateral X-rays should be taken to confirm that the tip ofthe needle is equidistant from the target at the midpedicular line. Note that redirecting theneedle while it is embedded in the deep muscle layers is difficult. To reposition the tip ofthe needle, the needle should be pulled back into subcutaneous tissue and then redirectedand properly positioned.

Selection of Site of Annular Lodging of Instruments

Proper selection of the site where instruments are to be lodged on the annulus duringintracanalicular, foraminal, and extraforaminal surgeries is essential at the onset of theoperative procedure. To avoid inadvertent insertion of the tip of the needle and the subse-quent instruments into the spinal canal and penetration of the dural sac and the neuralstructures, a safe point of resistance adjacent to the spinal canal has been identified andits radiographic landmarks have been described (5,15,23,24,30) (see Fig. 3 in Chapter 2).

When attempting subligamentous access to a disc herniation, midpedicular position-ing of the tip of the needle as observed in the AP C-arm images (Fig. 16A–C) is desired(Table 2). However, for evacuation of an extraforaminal herniation, adequate access tothe herniation, site is achieved by positioning the instruments on the lateral pedicularline, as is observed in the AP X-ray projection.

Intraoperative Radiographic Landmarks

Proper positioning of the needle in the triangular working zone must be monitoredintraoperatively, by both AP and lateral fluoroscopic examinations. Although in the lat-eral projection the position of the tip of the needle may appear to be satisfactory, the APprojection may demonstrate an improper and potentially dangerous site for annulotomyand introduction of instruments (Fig. 17A).

If the needle is inserted too vertically in the AP fluoroscopic examination, the tipof the needle may appear to be satisfactory. However, the LA projection shows theunacceptable vertical insertion (Figs. 17B and 18A–D). By contrast, if the needle isinserted flat or close to the horizontal line in the lateral X-ray projection, the tip ofthe needle will be observed posterior to the vertebral bodies, and the AP projectionwill falsely show that the needle has been inserted into the intervertebral disc or thespinal cannal.

When the needle is properly positioned in the lateral X-ray projection, the tip of theneedle is aligned with the posterior vertebral bodies (Fig. 16C). In the AP projection,the tip of the needle is observed in alignment with the midpedicular or lateral pedicularline (Fig. 16B).

Positioning of Cannulated Obturator and 5 × 5 mm Inner Diameter Access Cannula

Prior to introduction of a blunt-end cannulated obturator (soft-tissue dilator), thestylet of an 18-gage needle is replaced with a fine guide wire and the needle is thenwithdrawn. The cannulated obturator is positioned over the guide wire with the operatinghand used to advance the obturator toward the triangular working zone with a slow rotary

Herniated Lumbar Disc 81

movement (Fig. 19) while the other hand maintains the proper direction of the insertedobturator. If the needle has been positioned at the midpedicular line or slightly medial toit, the guide wire should be withdrawn about 1 cm prior to its full insertion (Fig. 20D).This maneuver permits the blunt end of the obturator to bypass the traversing root andmove it away from the surgical site (Fig. 20C). While holding the obturator firmlyagainst the annular surface, a 5 × 5 mm inner diameter (id) universal cannula is placedover the obturator and directed toward the annular surface in the triangular working

82 Kambin

Fig. 16. (A) AP X-ray of lumbar segment for selection of position of instruments in triangularworking zone: A, medial pedicular line; B, midpedicular line; C, lateral pedicular line. (B)Proper positioning of needle on midpedicular line at L3-L4. (C) Lateral positioning of needleshown in (B), which is in alignment with posterior boundary of adjacent vertebrae.

zone. It is advisable to confirm and document the correct position of the cannula with anAP and a lateral fluoroscopic examination (Fig. 20E,F).

Arthroscopic and Endoscopic Examination of Triangular Working Zone

To establish adequate inflow and outflow of saline solution and clear visualization ofintracanalicular and extra-annular structures, a suction irrigation valve is first attachedto the proximal end of a universal cannula (Fig. 21). Practitioners have also used cannu-las with a permanently attached suction irrigation valve to achieve the same goal. A 0°arthroscope or a working channel scope may be used for inspection of these structures.If the cannula is not held firmly against the posterolateral annulus in the triangularworking zone, venous bleeding may obstruct visualization of these structures.

Herniated Lumbar Disc 83

Fig. 16. (Continued)

Table 2 Selection of Annular Window Position

Medial pedicular line Midpedicular line Lateral pedicular linein AP projection in AP projection in AP projection

Use C-arm and document annular lodging site in AP projection

• Endoscopic inspection • Subligamentous • Removal of extraforaminalof spinal canal’s contents arthroscopic herniation

microdiscectomy• Removal of sequestered • Removal of foraminal • Arthroscopic anterior

disc fragment herination column stabilization• Discography• Chemonucleolysis

84 Kambin

Fig. 17. (A) Schematic drawing of axial view of improper positioning of needle. This signi-fies the importance of having both AP and lateral intraoperative images in order to confirm theproper positioning of the needle. A lateral X-ray study will show a satisfactory positioning ofthe needle. However, an AP fluoroscopy will demonstrate that needle has not been properlypositioned. (B) Schematic drawing of axial positioning of a needle. In the AP fluoroscopicexamination, the tip of the needle may appear to be in a satisfactory position. However, in thelateral fluoroscopic images, improper positioning of needle is identified.

Herniated Lumbar Disc 85

Fig. 18. (A) Improper vertical position of needle. In the lateral projection, the position of theneedle appears to be satisfactory. (B) AP fluoroscopic examination shown in (A). The tip of the nee-dle remains away from the intervertebral disc. (C) Improper horizontal insertion of needle. In theAP projection, it appears that the needle has reached the midpedicular line. (D) Lateral fluoroscopicexamination shown in (C) demonstrating that needle is far away from L4-L5 intervertebral disc.

Loosely woven adipose tissue that covers the surface of the posterolateral annulus isusually observed in the triangular working zone (see Fig. 4A,B in Chapter 2). Thisadipose tissue may be wiped out with a cottonoid or removed with a radiofrequency(Fig. 22) probe for clear visualization of the annulotomy site (see Fig. 6A,B in

86 Kambin

Fig. 18. (Continued)

Chapter 2). The exiting roots accompanying radicular arteries and veins are locatedunder the pedicular notch and are not seen on the surface of the posterolateral annulus.If the cannula has been positioned medial to the midpedicular line, epidural adiposetissue, epidural veins, the traversing root, and the dural sac will be observed (Fig. 5B;see also Fig. 9A,B in Chapter 2). Note that globs of epidural adipose tissue are usuallylarger than periannular fat, and they have a tendency to move into the cannula when thepatient inhales. If the cannula is positioned in the midpedicular region, a slight tilt of themedial end of the cannula dorsally will permit inspection of the contents of the spinalcanal. However, prior to introduction of the trephine or resecting instrument, the medialend of the cannula should be turned away and brought into a more vertical position toprevent injury to the neural structures.

Annulotomy

Most annulotomies are performed in the triangular working zone (12,30,31). Subliga-mentous approach provides access to paramedial and medial herniations (Fig. 23). Ifaccess to the spinal canal is desired, lateral fibers of the posterior longitudinal ligamentmust be removed to expose the ventral surface of the traversing root and the dural sac(Fig. 23A,B).

Herniated Lumbar Disc 87

Fig. 19. Interoperative photograph showing advancement of cannulated obturator over guidewire with slow rotatory movement to prevent deviation of obturator and bending of guide wire.

88 Kambin

Fig. 20. (A) Proper positioning of needle at L5-S1 in lateral fluoroscopic examination; (B) properpositioning of needle at L5-S1 shown in (A). (C) Demonstrates that obturator has reached interver-tebral disc and guide wire has been removed; (D) Demonstrates that cannulated obturator isapproaching L5-S1 intervertebral disc and signifies withdrawal of guide wire. (E) final and properposition of cannula at L5-S1; (F) final position of instrument in AP fluoroscopic examination.

Annulotomy may be performed with a trephine or under direct visualization by usinga working scope. The superficial layers of the annulus should be anesthetized with topi-cal anesthetic and then with a local injection of lidocaine (Xylocaine) solution. The fourcorners of the annular fibers inside the cannula are perforated with an 18-gage needle.This is a useful test to ensure that the cannula is not seated on the vertebral plates and iswell centered on the disc space.

Positioning of Oval Cannulas

The proper positioning of the 5 × 8 or 5 × 10 mm id oval cannulas is accomplishedvia the following steps (26,30,32,33):

1. Reinsert the blunt cannulated obturator into the lumen of a 5 × 5 mm id universal cannula,and insert the distal end of the obturator into the intervertebral disc.

2. Withdraw the 5 × 5 mm universal cannula.3. Position a 10- or 12-mm od jig template over the proximal end of the inserted cannu-

lated obturator. This jig is provided with an oval-shaped bore that accommodates both

Herniated Lumbar Disc 89

Fig. 20. (Continued)

the cannulator obturator and a half- or full-moon auxiliary obturator (see Fig. 3A,B inChapter 3).

4. Insert a half-moon auxiliary obturator through the bore of the jig template, and direct ittoward the intervertebral disc at the index level.

5. Remove the jig template and insert an appropriate-size oval cannula by sliding it over thepreviously positioned cannulated obturator and half- or full-moon auxiliary obturators.Then gently tap the oval cannula into position and engage in the superficial layer of theannular fibrosis (Fig. 24A,B).

90 Kambin

Fig. 20. (Continued)

Retrieval of Herniated Disc Fragments

When subligamentous access to the herniation site is being used (6,12,22), the accesscannula is first engaged into the superficial annular fibers and the position of the accesscannula is maintained with one hand throughout the surgical procedure. A straight and anupbiting forceps are used to grasp and remove the herniated disc material that may be dis-lodged posterolaterally adjacent to the tip of the inserted cannula. The inserted trephine andforceps will have a tendency to sweep ventral to the posterior longitudinal ligamentum,traversing root and lateral dura, thus facilitating evacuation of the herniated fragments.

Herniated Lumbar Disc 91

Fig. 21. (A) A 5 × 5 mm id universal cannula; (B) a removable suction irrigation valve.

Fig. 22. Interoperative photograph demonstrating how radiofrequency probe is used forperiannular hemostatsis and removal of adipose tissue in triangular working zone.

The oval cannula permits insertion of a 0 of 30° scope and resecting instruments for visu-alization of the surgical site during retrieval of the herniated disc fragments (Fig. 25;see also Fig. 9A,B in Chapter 3).

Fluid Management

Because most arthroscopic spinal surgeries are currently performed with the use of fluidmedia, proper intraoperative fluid management makes it possible for the operating surgeonto better differentiate between anatomical and pathological structures. Most of the

92 Kambin

Fig. 23. (A,B) Intraoperative arthroscopic view demonstrating subligamentous access to discherniation.

arthroscopes and endoscopes used for minimally invasive spine surgery, including theworking-channel scope, are provided with at least two channels for the inflow and outflowof saline solution. However, fluid management may have to be adapted to the surgicalaccess selected and the technique utilized in the management of a given patient. Most sur-geons who have employed our original technique of uniportal intradiscal access to posteri-orly lodged disc fragments have experienced extreme difficulty in establishing the inflow

Herniated Lumbar Disc 93

Fig. 24. (A) Lateral fluoroscopic examination demonstrating that cannulated obturator andauxiliary obturator are inserted into intervertebral disc and oval cannula has reached surfaceof annulus; (B) lateral fluoroscopic examination shown in (A) demonstrating withdrawal ofobturators and anchorage of oval cannula in superficial layer of posterolateral annulus.

and outflow of saline solution. Nuclear debris invariably blocks the outflow channel of thearthroscope and the free flow of saline solution in the cavity that is created in the interver-tebral disc. To obtain a quick view of the surgical site, we previously advocated the use of a50-cc syringe for a bulk delivery of saline solution into the disc cavity (24,34). However,this technique has been abandoned and is no longer used.

Currently, most of the arthroscopic procedures that were previously performed via anintradiscal access are now performed by using a subligamentous (12,22) approach, atransforaminal approach (29,31,35,36), or bilateral biportal access (22).

In our experience, allowing free gravitation of outflow saline solution has been mosteffective and has reduced the incidence of intraoperative bleeding and formation of airbubbles in the surgical field. During both subligamentous and transforaminal intra-canalicular access, a large volume of saline solution may be introduced into the surgicalfield and allowed to gravitate out from the end of the cannula and collect in a sterileplastic bag that is attached to the surgical drapes. The volume of inflow saline solutionmay be increased by elevating the inflow bag or applying a pressure cuff around the inflowbag. The gravitation of outflow solution may be used in conjunction with a 5 × 5 mm id.Universal cannula or an oval 5 × 8 mm id cannula.

When bilateral biportal access to the intervertebral disc is employed (see Fig.10A,B in Chapter 3), inflow of saline solution is connected to the inflow valve of thearthroscope. This tends to remove nuclear debris and blood away from the arthro-scope, thus permitting visualization of the surgical site. The saline solution is thenretrieved through the cannula that was introduced into the intervertebral disc fromthe opposite portal.

94 Kambin

Fig. 25. Schematic drawing demonstrating how scope and upbiting forceps are used inconjunction with oval cannula for visualization and extraction of herniated disc fragments.

During transforaminal access to the contents of the spinal canal (29,31,35–36),epidural bleeding invariably obstructs clear visualization of the anatomical and patho-logical structures. The following steps have proven useful in controlling epiduralbleeding:

1. If the bleeding site is visible and can be located, use a radiofrequency probe or bipolar elec-trocoagulator to control the bleeding points.

2. Close the outflow valve of saline solution and increase the volume and pressure through theinflow valve.

3. Use cold saline solution for irrigation.4. Introduce a small cottonoid saturated with topical anticoagulants into the surgical site by

holding it with the tip of a forceps.5. Some surgeons have reported adequate epidural hemostasis with injection of diluted

epinephrine solution (1:500,000) into the surgical site. After injecting the solution, use aclear saline solution for further irrigation and completion of the operative procedure.

6. Use hypotensive anesthesia.

Retrieval of Foraminal and Extraforaminal Herniations

With the introduction and availability of MRI and CT scans, orthopedic and neuro-logical surgeons have been able to establish an early definitive diagnosis of disc hernia-tions in the lateral zone and render appropriate surgical management when deemednecessary.

The hidden zone, which was well described by MacNab, now is readily diagnosed byhigh-resolution CT and MRI evaluations. The lateral zone has been divided into threesections: subarticular, foraminal, and extraforaminal regions (38) (Fig. 26A). Wiltsedescribed the paraspinal sacrospinalis splitting approach for the removal of far-outextraforaminal disc herniations (39).

The midline approach has also been used to access the lateral zone. Both of theseapproaches, in particular the midline approach, require extensive soft-tissue dissection,partial facetectomy, and bone removal, which may become complicated by segmentalinstability. Therefore, the midline approach should be reserved for management ofpatients with severe lateral recess stenosis and facet hypertrophy. In our experience,arthroscopic microdiscectomy is the procedure of choice for the management of themajority of foraminal and extraforaminal disc herniations.

In contrast to intracanalicular disc herniations, foraminal and extraforaminal hernia-tions are associated with signs and symptoms of compression and tension on the exitingroot. Therefore, a patient with a disc herniation at L4-L5 most likely will exhibit signsand symptoms of L4 root involvement, whereas a patient with intracanalicular parame-dial disc herniation at the same level will present objective and subjective evidence ofL5 root involvement (Fig. 26B). A sequestered disc fragment in the lateral zone maymigrate distally and cause pressure on the traversing root and the dural sac.

A patient with an acute ruptured disc in the foraminal or extraforaminal region(Fig. 27A) will present all signs and symptoms of disc herniations, including positivetension signs, with or without neurological deficit. By contrast, an individual with acontained bulging and protruding disc associated with small marginal posterolateralosteophytes may present signs and symptoms of lateral stenosis, having radicular painbut lacking positive tension signs.

Herniated Lumbar Disc 95

96 Kambin

Fig. 26. (A) Schematic drawing showing subarticular foraminal and extraforaminal zone inAP and axial projection; (B) schematic drawing demonstrating that a paramedial herniation atL4-L5 produces compression at L5 root whereas a foraminal and an extraforaminal herniationat same level can produce compression of L4 root.

The positioning of the patient, draping of the C-arm, and locating of the surgical siteare similar to those described previously (see Chapters 3 and 4). The skin entry site isusually about 9 or 10 cm from the midline. For removal of an extraforaminal herniation,the tip of an 18-gage needle is positioned over the lateral pedicular line, as is observed inthe AP radiographic evaluation. However, one should be aware of the fact that in the lat-eral X-ray images the tip of the needle may not appear to be aligned with the posteriorboundary of the adjacent vertebral bodies, because the bulk of the herniation preventsalignment of the tip of the needle with the posterior vertebral line.

Following initial annular fenestration and subligamentous evacuation of the discmaterial (Fig. 27B,C), the cannulated obturator is reinserted into the lumen of the can-nula and the cannula and obturator assembly are gradually moved laterally into a morevertical position. This maneuver permits further access and evacuation of the remainder ofdisc fragments. In addition, the cannula has a tendency to move laterally and protect the

exiting root during the evacuation stage. If marginal posterolateral osteophytes are deemedto be responsible for the development of lateral recess stenosis, a trephine or poweredinstruments may be used for annulectomy and evacuation of the osteophytes (27,40). Inour hands, the oval cannula with an id of 5 × 8 mm has been particularly helpful for accessand retrieval of extraforaminal herniations. At times a sequestrated herniated disc migratesto the foraminal and subpedicular region of the proximal segment (Fig. 28A). The retrievalof herniated fragments requires positioning of the 5 mm id cannula in the subpedicularregion for extraction of sequestrated disc fragments (Fig. 28B–F).

Herniated Lumbar Disc 97

Fig. 27. (A) Preoperative axial MRI study of foraminal and extraforaminal herniation at L3-L4. (B) Interoperative arthroscopic view of herniation site. Note that the probe is being used toisolate the herniation from the surrounding tissue. (Reprinted from ref. 43, with permission.) (C)Interoperative arthroscopic view demonstrating that herniation has been evacuated and subliga-mentous region is being probed in search of additional disc fragments.

98 Kambin

Retrieval of Paramedial or Central Disc Herniations

Subligamentous access to disc herniations requires that the skin entry site be morelateral ([12,22]; Table 1). The needle and instruments are positioned at the mid-pedicular line. Following inspection of the annulotomy site, the medial end of thecannula is tilted toward the patient’s dorsum, and the contents of the spinal canal areinspected (Fig. 29A–C). The medial end of the cannula is then tilted ventrally so thatthe open end of the cannula is not in the spinal canal (Fig. 30). Annular fenestration isperformed with a working channel scope in accordance with the principles that weredescribed previously. The cannulated obturator is reinserted into the lumen of thecannula and is driven into the intervertebral disc. This maneuver makes it possible toadvance the cannula over the obturator and to engage it with the superficial layer ofthe annulus fibrosus. A cavity is then created ventral to the posterior longitudinal lig-ament by partial nucleotomy and evacuation of the torn fibers of the posterior annu-lus (Fig. 31A,B). The intradiscal tail of nonmigrated sequestered fragments or asubligamentous herniation is then identified under direct visualization and removed.

The subligamentous approach is particularly desirable for evacuation of a containeddisc herniation, thus making it unnecessary to enter the spinal canal.

Retrieval of Sequestered Disc Herniation

Transforaminal access (22,29,31,36) requires that the skin entry site be more lateraland that the needle be positioned slightly medial to the midpedicular line (Fig. 32A,B)(Tables 1 and 2). It is advisable to confirm the skin entry site via a prone preoperative CTstudy. The principle of the surgical approach is similar to that described for subligamentousaccess. Following the creation of a cavity under the posterior longitudinal ligament, lateralfibers of the posterior longitudinal ligament adjacent to the tip of the inserted cannula arealso removed with a forceps or a radiofrequency probe (Fig. 33A–C).

Fig. 27. (Continued)

Herniated Lumbar Disc 99

Fig. 28. (A) Axial view of migrated sequestrated disc herniation. Note that at disc level (left)the herniation is not seen. However, at the subpedicular region (right), a large free fragment isdemonstrated. (B) Proper positioning of needle in subpedicular region in AP projection. (C) Lateralview of needle positioning illustrated in (B). (D) Positioning of cannula in subpedicular region.(E) Interoperative photograph demonstrating retrieval of sequestrated disc fragments withstraight forceps. (F) Interoperative photograph showing partial removal of sequestrated fragmentsand exposure of nerve root ganglia.

Epidural bleeding and adipose tissue may obstruct clear visualization of intra-canalicular structures. Epidural bleeding may be controlled by temporarily increasingthe pressure of the saline solution through the inflow valve and closing the outflowvalve. The use of cold saline solution has also been useful in controlling epiduralbleeding. We have used cottonoids saturated with topical anticoagulants and radiofre-quency probes to stop bleeding. Excess epidural adipose tissue may also be evacuatedwith the aid of forceps and easily with the use of a radiofrequency probe. Some surgeons

have used a cold dilute epinephrine solution (1 mg/1000 cc of saline solution) for con-trol of multiple, minute epidural bleeders.

Once epidural bleeding has been controlled and excess epidural adipose tissueremoved, a working channel scope is introduced into the lumen of the cannula and thedisc fragments are visualized and evacuated. The cannula may be tilted cephalad orcaudad for further access to the migrated disc fragments within the spinal canal.

Bilateral biportal intradiscal (Fig. 34A–D) access may be used for evacuation of anonmigrated sequestered disc herniation. This approach is most effective for retrievalof central herniations. It prevents undue intraoperative manipulation of the neural

100 Kambin

Fig. 28. (Continued)

Herniated Lumbar Disc 101

Fig. 28. (Continued)

structures. A cavity is created ventral to the posterior longitudinal ligamentum. Com-munication between the right and left portal is developed via resection of the nucleuswith forceps or with the aid of a radiofrequency probe. Inflow and outflow of salinesolution is established between the two portals. The intradiscal segment of the sequestrated

102 Kambin

Fig. 29. (A) Schematic drawing demonstrating how medial end of cannula is tilted towarddorsum of patient by pushing lateral end of cannula ventrally, therefore allowing visualization ofcontents of spinal canal; (B,C) interoperative photograph demonstrating visualization of con-tents of spinal canal following dorsal tilt of medial end of cannula.

Herniated Lumbar Disc 103

Fig. 30. Schematic drawing showing ventral tilt of medial end of cannula in preparation ofannulotomy.

Fig. 29. (Continued)

disc herniation is then identified and grasped with forceps and removed (Fig. 35A–C).Although this technique is more time-consuming than the uniportal approach, it sparesthe neural and vascular structures from undue manipulation and potential injury (Fig.36; see also Fig. 10 in Chapter 2). At the completion of the operative procedure, theventral dura and torn fibers of the posterior longitudinal ligamentum are observed via a

104 Kambin

Fig. 31. (A) Interoperative photograph demonstrating subligamentous annulotomy. Thetraversing root is seen at the top of photo above. PLL, posterior longitudinal ligamentum.

Herniated Lumbar Disc 105

Fig. 32. (A) AP fluoroscopic examination demonstrating proper positioning of needle onmedial pedicular line for intracanalicular access; (B) lateral positioning of needle as shown in(A); (C) proper positioning of cannula for intracanalicular surgery.

30 or 70° arthroscope (Fig. 37A–C). To ensure adequate decompression of the travers-ing nerve root, the spinal canal may be examined prior to withdrawal of instruments(Fig. 38A,B).

A larger-diameter cannula may be positioned at the interlaminar space under fluoro-scopic control from a distance of 2 cm from the midline for mini-laminatomy and retrievalof sequestered disc fragments under endoscopic visualization. In the past, we have used

cannulas with side windows for introduction of straight endoscopes (Fig. 18A,B inChapter 1). This allowed the insertion of a variety of instruments through the proximalopening of the cannula (6,26,37). With the later availability of angled endoscopes, we wereable to insert both endoscopes and instruments through the proximal opening of a larger-diameter cannula (40).

106 Kambin

Fig. 33. (A) Intraoperative photograph demonstrating how a cavity is created under posteriorlongitudinal ligamentum (PLL) and part of ligamentum has been removed for exposure of massiveherniation around traversing root; (C) schematic drawing showing how a radiofrequency probeis used for vaporization of PLL and torn fibers of annulus and exposure of herniated fragmentsand nerve root, under endoscopic visualization.

Retrieval of Thoracic or Thoracolumbar Herniation

Considering the limited muscle mass in the thoracolumbar area, thorough preop-erative planning becomes an essential part of retrieving a thoracic or thoracolumbarherniation.

The skin entry site should be determined and selected by using a prone CT study ofthe index level before surgery (Fig. 15) (5,29). Accurate placement of the skin entrysite will prevent inadvertent violation of the thoracic or abdominal cavity by theinserted instruments. Because many extradural defects in the thoracic region do not rep-resent a soft disc herniation, a preoperative CT scan study of the surgical site is helpfulin differentiating between soft herniations and osteophytes. Owing to anatomical limi-tations, which are imposed by the rib cage and the limited muscle mass, it is invariablydifficult to place the skin entry site far enough laterally and to use a subligamentousapproach to the herniation site. Bilateral biportal access to the intervertebral disc maybe necessary for visualization and decompression of the ventral surface of the dural sacand the nerve roots. Great care should be exercised not to use trimmer blades or vibratinginstruments adjacent to the spinal cord during an intradiscal approach to the thoracicor thoracolumbar region.

Retrieval of Recurrent Disc Herniations

Recent technological advances in the field of minimally invasive spinal surgery haveopened a window of opportunity to access, isolate, and retrieve recurrent herniated discfragments. Neurolysis of tethered nerve root under endoscopic control is the essentialfirst step (Fig. 39). The surgical approach for retrieval of recurrent disc herniations issimilar to that previously described for subligamentous access to the intervertebral disc.A small cavity is first created under the posterior longitudinal ligament by removingpart of the nucleus and torn annular tissue. This is then followed by partial resection of

Herniated Lumbar Disc 107

Fig. 33. (Continued)

lateral fibers of the posterior longitudinal ligament adjacent to the tip of the insertedcannula (Fig. 33A,B). This permits exposure of the ventral surface of the traversingroot and the dural sac. The recurrent disc herniation then may be isolated and retrievedvia a working channel scope. If it becomes necessary, a bipolar radiofrequency probemay be utilized for hemostasis and vaporization of epidural and perineural scar tissue.Prior to retrieval of the cannula at the end of the operative procedure, a steroid com-pound may be used to infiltrate epidural and perineural structures.

108 Kambin

Fig. 34. (A) Proper positioning of two needles in AP projection for bilateral biportal accessto a large nonmigrated central disc herniation. (B) Proper positioning of needles shown in (A) inlateral radiographic examination. (C) Insertion of two cannulas from right and left side ofpatient. This accommodates ipsilateral insertion of forceps and introduction of the arthroscopefrom the opposite portal for intradiscal triangulation. (D) Position of inserted instruments inlateral radiographic examination.

HOW TO PREVENT COMPLICATIONS

Hazards Associated With Needle Insertion

A far lateral or vertical insertion of needles at the onset of a minimally invasiveposterolateral approach to the intervertebral disc should be avoided. When perform-ing surgery in the upper lumbar spine or at the thoracolumbar junction, the surgeonshould request a preoperative prone CT study of the surgical site in order to determineaccurately the distance of the skin entry site from the midline and the appropriateangle of needle insertion. Inadvertent vertical insertion of the needle may cause injuryto the iliac arteries and veins or may penetrate the contents of the abdominal cavity.

Herniated Lumbar Disc 109

Fig. 34. (Continued)

Horizontal insertion of the needle at the onset of the operative procedure is always asafer practice.

Hazards Associated With Placement of Cannulated Obturator

The cannulated obturator should smoothly descend toward the triangular workingzone as it advances over the previously positioned guide wire. The surgeon shouldmaintain the direction of the insertion of the obturator with one hand while using the

110 Kambin

Fig. 35. (A) Interoperative photograph demonstrating how a cavity is created within inter-vertebral disc and intradiscal extension of sequestrated fragment is identified (arrows).(Reprinted from ref. 43, with permission.) (B) Interoperative photograph showing use of deflectingsuction forceps to access herniated disc fragments. (C) Intraoperative photograph demonstratinguse of radiofrequency probe to develop communication between right and left portals.

Herniated Lumbar Disc 111

Fig. 35. (Continued)

Fig. 36. Endoscopic view of undisturbed traversing root with intact accompanying finevascular structures.

opposite hand to advance the obturator with a slow rotary movement. If the obturator isnot aligned with the guide wire, forceful insertion of the obturator will tend to bend theinner end of the guide wire that is anchored into the annular fibers; misalignment alsomakes withdrawal of the guide wire difficult. If in doubt, intraoperative fluoroscopicexamination should be conducted to confirm that the direction of the obturator and theguide wire has remained satisfactory (Fig. 19).

112 Kambin

Fig. 37. (A) Interoperative photograph following intradiscal evacuation of large centralherniation. Note the exposure of the ventral dura (arrows) and magnified view of a few fibers ofthe posterior longitudinal ligamentum shown at the top. (B) Intradiscal view of ventral dura fol-lowing evacuation of sequestrated disc herniation. (C) Note torn fibers of the posterior longitu-dinal ligamenum and exposure of ventral dura (arrow) following intradiscal biportal evacuationof sequestrated disc herniation.

Hazards Associated With Migration of Cannula

The design of most instruments does not permit them to penetrate more than 2 cmbeyond the distal extremity of the inserted cannula. Inadvertent intradiscal migration ofthe cannula will allow deep penetration of the tip of the forceps and other instrumentsinto the intervertebral disc. This will cause complications if the instruments enter theabdominal cavity. When an intradiscal or subligamentous approach to the intervertebral

Herniated Lumbar Disc 113

Fig. 37. (Continued)

disc is attempted, the distal end of the cannula should be engaged in the superficial layersof the annulus and held in this position by the operating surgeon throughout the surgicalprocedure.

Hazards Associated With Periannular or Intracannicular Bleeding

Periannular bleeding usually occurs when the cannula is not firmly held against theannular fibers in the triangular working zone. If excessive bleeding is encountered, thecannulated obturator should be reintroduced into the lumen of the cannula and firmlyheld against the annular surface. The cannula should then be advanced over the obturatorand held in position. If the bleeding remains uncontrolled, AP and lateral X-rays shouldbe taken to establish that the cannula is on the midpedicular line in the AP projectionand has not been introduced into the spinal canal. Superficial veins on the surface of theannulus in the triangular working zone are invariably observed; they should be coagu-lated with a radiofrequency probe or by introducing a cottonoid saturated with topicalanticoagulants. Epidural bleeding is common during transforaminal access to the spinalcanal and should be meticulously controlled to achieve a clear view prior to insertion ofgrasping instruments.

Potential Hazards Associated With Use of Power-Driven Suction Nuclear Resectors

Great care must be exercised when a subligamentous approach is utilized for retrieval ofherniated disc fragments. The use of suction and trimmer blades directly under the traversingroot and dura is extremely dangerous and will cause serious complications. During a bipor-tal approach to an intervertebral disc, a trimmer blade is invariably used to create a cavityand communication between the two portals. When the trimmer blade is retracted into thecannula to free nuclear debris, the open end of the trimmer blade should be turned ventrallyand away from the neural structures and dural sac. Recently, we successfully used radiofre-quency probes to resect nuclear tissue with no undue complications.

Reducing Incidence of Postoperative Dysesthesia

Development of skin hypersensitivity following arthroscopic or endoscopic spinesurgery via posterolateral access has been reported (22,27). Symptoms usually occur 4to 5 d following surgery and involve the preoperative symptomatic extremity. The pre-senting symptoms include complaints of a burning sensation, a causalgic type of pain,and an inability to wear a stocking or keep the limb under covers. Intraoperative traumato the nerve root ganglia of the exiting or the traversing root may be responsible for thepresenting symptoms. In our experience, this complication was more prevalent whentransforaminal access to the spinal canal or decompression of the lateral recess stenosiswas attempted (27). The use of power-driven disc shavers or laser light in the foramenmay also contribute to development of this complication.

114 Kambin

Fig. 38. (A,B) Intraoperative photograph demonstrating final examination of traversing root fol-lowing intradiscal retrieval of sequestrated disc herniation. Note the site of extracted disc fragment.

Intraoperative injection of a mixture of fentanyl and normal saline solution aroundthe nerve root ganglia at the onset of the operative procedure following proper position-ing of an 18-gage needle in the triangular working zone has proven to be helpful in pre-vention of these complications.

Although most patients respond to oral administration of nonsteroidal or short-termsteroid therapy, in our clinic we have had success with the intraforaminal steroid injec-tion particularly in the first few days following the onset of the symptomatology. Undera sterile environment, an 18-gage needle is inserted into the foramen at the site of theprevious surgery. Following radiographic confirmation of the position of the tip of the

Herniated Lumbar Disc 115

Fig. 39. (A,B) Intraoperative photograph showing neurolysis of traversing root surroundedby scar tissue and isolation of recurrent disc herniation.

needle, a diluted mixture of a steroid compound is injected into the foramen aroundboth the traversing and exiting nerve roots.

Preventing Postoperative Discitis and Infection

The incidence of postoperative bacterial discitis in the literature has been <0.5%(22,41,42). When surgery is performed in an operating room setting and there is adher-ence to the principle of surgical technique including draping of the patient and C-arm,the chance of developing a postoperative infection will be greatly reduced. The use ofprophylactic antibiotics is essential in the prevention of this serious complication.

REFERENCES

1. Lazarevic LK. Ischias postica Cotunni. Sypski Arch 1980;7:23.2. Forst JJ. Contribution a l’etude clinique la sciatique. Paris, These 33:1881.3. Lasegue EC. Consideration aure la sciatique. Arch Gen Med 1864;6:558–580.4. Cram RH. A sign of nerve root pressure. J Bone Joint Surg 1953;35-B:192–195.5. Kambin P. Arthroscopic microdiscectomy. Arthroscopy 1992;8:287–295. 6. Kambin P. Arthroscopic techniques for spine surgery, in Operative Arthroscopy (McGinty

JB, Casperi RB, Jackson RW, Poehling GG, eds.), Lippincott-Raven, Philadelphia, 1996,pp. 1215–1225.

7. Atken AP, Bradford CH. End result of ruptured intervertebral disc in industry. Am J Surg1947;73:365.

8. Balderston RA, Gilyard GG, Jones AM, et al. The treatment of lumbar disc herniation:simple fragment and excision versus disc space curettage. J Spinal Disord 1991;4:22–25.

9. Hampton D, Laros G, MacCarron R, Franks D. Healing potential of annulus fibrosis. Spine1989;14(4):398–401.

10. Markolf KL, Morris JM. The structural components of the intervertebral disc: a study oftheir contributions to the ability of disc to withstand compressive forces. J Bone Joint Surg1974;56-A:675–687.

11. Ross JS, Robertson JT, Frederickson RC, Petric JL, Obuchowski N, Modic MT, deTriboletN. Association between peridural scar and recurrent radicular pain after lumbar discectomy:magnetic resonance evaluation. ADCON-L European Study Group. Neurosurgery1996;38:855–863.

12. Hermantin F, Quartararo L, Peters T, Kambin P. A prospective, randomized study compar-ing the results of open discectomy versus video-assisted microdiscectomy. J Bone JointSurg 1999;81-A(7):958–965.

13. Haaland AK, Graver V, Ljunggren AE, Loeb M, Lie H, Magnaes B, Godal HC. Fibrinolyticactivity as a predictor of the outcome of prolapsed intervertebral disc surgery with referenceto background variables: results of a prospective cohort study. Spine 1992;17:1022–1027.

14. Park WW. The significance of venous return impairment in ischemic radiculopathy andmyelopathy. Orthop Clin North Am 1991;22:213–221.

15. Park WW. Clinical anatomy of the lower lumbar spine, in Arthroscopic Microdiscectomy:Minimal Intervention in Spinal Surgery (Kambin P, ed.), Urban & Schwarzenberg, Baltimore,1991, pp. 11–29.

16. Hoyland JA, Freeman AJ, Jayson ML. Intravertebral foramen venous obstruction: a causeof periradicular pain fibrosis. Spine 1989;14:558–568.

17. Dullerud R, Graver V, Haakonsen M, Haalaud AK, Loeb M, Magnaes B. Influence of fibri-nolytic factors on scar formation after lumbar discectomy: a magnetic resonance imagingfollow-up study with clinical correlation performed 7 years after surgery. Spine1998;23:1464–1469.

116 Kambin

18. Pountain GD, Keegan AL, Jayson ML. Impaired fibrinolytic activity in defined chronicback pain syndromes. Spine 1987;12:83–86.

19. Rantanen J, Hurme M, Falck B, Alaranta H, Nykvist F, Lehto M, Einola S,Kalimo H. Thelumbar multifidus muscle five years after surgery for a lumbar intervertebral disc herniation.Spine 1993;18:568–574.

20. Weber BR, Grob D, Dvorak J, Muntener M. Posterior surgical approach to the lumbarspine and its effect on the multifidus muscle. Spine 1997;22:1765–1772.

21. Kambin P, Cohen L, Brooks ML, Schaffer JL. Development of degenerative spondylosis ofthe lumbar spine after partial discectomy: comparison of laminatomy, discectomy and pos-terolateral discectomy. Spine 1995;20:599–607.

22. Kambin P, O’Brien E, Zhou L. Arthroscopic microdiscectomy and selective fragmentectomy.Clin Orthop 1998;347:150–167.

23. Kambin P. Percutaneous lumbar discectomy: current practice. Surg Rounds Orthop1988;31–35.

24. Kambin P. in Arthroscopic Microdiscectomy: Minimal Intervention in Spinal Surgery(Kambin P, ed.), Urban & Schwarzenberg, Baltimore, 1991, pp. 67–100.

25. Kambin P. The role of minimally invasive surgery in spinal disorders. Adv Oper Orthop1995;3:147–171.

26. Kambin P. Arthroscopic microdiscectomy, in The Adult Spine: Principles and Practice Edi-tion, (Frymoyer JW, Ducker T, Hadler N, et al. eds.), Raven, New York, 1996, pp.2023–2036.

27. Kambin P, Casey K, O’Brien E, et al. Transforaminal arthroscopic decompression of lateralrecess stenosis. J Neurosurg 1996;84:462–467.

28. Yeung A, Tsou PM. Posterolateral endoscopic excision for lumbar disc herniation, surgicaltechnique: outcome and complications in 307 consecutive cases. Spine 2002;27:722–731.

29. Kambin P. Arthroscopic microdiscectomy: lumbar and thoracic in Spine Care, vol. 2(White A, Schofferman JA, eds.), Mosby, St. Louis, 1995, pp. 1002–1016.

30. Kambin P, McCullen G, Park W, et al. Minimally invasive arthroscopic spinal surgery.Instruct Course Lect 1997;46:143–161.

31. Kambin P. Diagnostic and therapeutic spinal arthroscopy. Neurosurg Clin North Am1996;7(1):65–76.

32. Kambin P. Unilateral biportal percutaneous surgical procedures and instrumentation. USpatent 5,395,317, 1991.

33. Kambin P, Gennarelli T, Hermantin F. Minimally invasive techniques in spinal surgery:current practice. Neurosurg Focus 1998;4(2):1–10.

34. Kambin P, Sampson S. Laminectomy versus percutaneous lateral discectomy—a comparativestudy. Paper presented at 51st AAOS Meeting, February 1984, Atlanta, GA.

35. Kambin P, Casey K. Posterolateral arthroscopic and endoscopic microdiscectomy of lum-bar spine. In Spinal Cord Surgery (Zileli M, ed.) Vol. 1 2002; pp 705–718.

36. Mathews HH, Kyles MK, Lang BH, Fiore SM, Gordon CL. Spinal endoscopy: indications,approaches and applications. Orthop Trans 1995;19:219.

37. Kambin P. Principles and experiences with monoportal arthroscopic microdiscectomy.Video presentation of endoscopic laminectomy and foraminatomy. Bulgrist OrthopaedicHospital University of Zurich School of Medicine, Zurich, Switzerland, December 2, 1993.

38. Wiltse LL, Rothman SLG, Chafetz N. Stenosis in spondylolisthesis of the lumbar spine(Anderson GBJ, McNeill TW, eds.), Mosby Yearbook 1992, pp. 179–207.

39. Wiltse LL, Bateman JG, Hutchinson RH, et al. The paraspinal sacrospinalis-splittingapproach to the lumbar spine. J Bone Joint Surg (Am) 1968;50:919–926.

40. Kambin P. Arthroscopic foraminal surgical procedures performed through a working channelscope. US patent application 08/207831 March 1994.

Herniated Lumbar Disc 117

41. Hadjipavlou AG, Kambin P, Lander PH, Crow WN, Simmons JW. Imaging-guided minimallyinvasive surgery for low back pain, sciatica and spinal infection. J Intervent Radiol1999;14:1–22.

42. Schaffer JL, Kambin P. Percutaneous posterolateral lumbar diskectomy and decompressionwith a 6.9 millimeter cannula: analysis of operative failures and complications. J BoneJoint Surg (Am) 1991;73:822–831.

118 Kambin

5Management of Discogenic Pain

and Spinal Instability Using Minimally Invasive Surgical Techniques

Parviz Kambin, MD

INTRODUCTION

Although a causal relationship between lumbar disc herniation and radiculopathy hasbeen well established by many clinical studies and scientific tests, the role of other paingenerators of the lumbar spine in clinical manifestations of pain and disability is stillbeing debated. The purpose of this chapter is to identify these pain generators and offeran alternative treatment modality for their management using minimally invasive surgicaltechniques.

PAIN GENERATORS AND THEIR MANAGEMENT

Degenerated and Bulging Intervertebral Discs

Fibers of the posterior longitudinal ligament, as well as superficial layers of theannulus fibrosis, are innervated by the sinovertebral nerve. The fibers of the posteriorlongitudinal ligamentum extend laterally to the foramena and into the extraforaminalregion (1) (see Chapter 2). These fibers are extremely sensitive and cause pain whensubjected to compression and tension. As early as 1948, Falconer et al. (2) reportedreproduction of back pain and a referral type of leg pain when the intervertebral discwas compressed during surgery under local anesthesia. With the advent of minimallyinvasive spinal surgery, the sensitivity of the posterior longitudinal ligament and its lat-eral expansion has been well documented in recent literature. Kambin et al. (3) studiedthe pathophysiology of the bulging annulus that is associated with a degenerated inter-vertebral disc; partial tear of annular fibers; and peripheral migration of degenerated,collagenized nuclear tissue. Parke (1) demonstrated the role of the bulging intervertebraldisc in producing tension on the fibers of the posterior longitudinal ligament and in theclinical presentation of low-back pain.

Peripheral Tear of the Annulus

Although discography has been the subject of controversy, the usefulness of com-puted tomography (CT) discography in identification of pain-producing peripheral

119

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

annular tears has been documented (4–6). In addition, the successful relief of disablingpain following anterior column stabilization further confirms the importance of carefuldiagnosis of this pathological condition. Detachment of the annular rim that may beassociated with ingrowth of innervated granulation tissue may be a source of disabling

120 Kambin

Fig. 1. (A) MRI study showing area of high intensity at L5-S1 level; (B,C) interoperativephotograph demonstrating detachment of annulus from proximal vertebral plate and partial tearof annular fibers associated with ingrowth of granulation tissue that has been evacuated.

low-back pain (7). In magnetic resonance imaging (MRI) studies, the presence of ahigh-intensity zone in the posterior annulus appears to be owing to ingrowth offibrovascular tissue into the torn annular fibers (Fig. 1A–C). At times, the high-intensityzone does not communicate with the nucleus pulposus and therefore makes the diagno-sis of this condition with CT discography extremely difficult. In endoscopic diagnosisstudies by my colleagues and I, we have found that the above structures are highly sen-sitive to palpation and compression via the instruments that are inserted through thepreviously positioned cannula and have been a useful diagnostic tool in our surgicalarmamentarium (8).

Degenerative Facet Joint

Facet joints are weight-bearing synovial joints and, like joints of the extremities, arecapable of producing pain. They may be treated with minimally invasive techniques.Posterolateral osteophytes arising from the vertebral plates and hypertrophic changes inthe facet joints may be responsible for symptoms of lateral recess stenosis that requirespecific surgical management (9).

Inflammatory Agents

The role of inflammatory agents in the development of pain has been extensivelyreviewed in the literature. It has been suggested that pain produced by discography isrelated to manipulation of the intervertebral disc that indirectly affects the vasoactiveintestinal peptide (VIP) and substance P found in dorsal root ganglia (6). A variety of neu-ropeptides, including calcitonin, gene-related peptide, and substance P, have been foundin dorsal root ganglia. A relationship has been reported between low-frequency vibrationand the amount of substance P and VIP in the dorsal root ganglia. A high level of phos-pholipase A2 activity has been found in lumbar disc herniation (10). It has been speculatedthat loss of annular integrity and expulsion of nuclear material are associated with localinflammatory response in the epidural space and the nerve roots. The addition of mechani-cal compression on the already inflamed nerve root by the herniated disc is then associatedwith the onset of symptomatology and the clinical development of pain.

Discogenic Pain and Spinal Instability 121

Fig. 1. (Continued)

Diagnostic Modalities

In the absence of radiculopathy, the diagnosis and localization of a pain-producingintervertebral disc is unfortunately not a simple task. Although MRI studies invariablydemonstrate the presence of annular tears, it is extremely difficult to establish that theMRI findings are responsible for the presenting symptomatology. However, MRI stud-ies remain an excellent modality for diagnosis of the extent of hydration and degenera-tion of the intervertebral disc, the presence or absence of bulging or annular protrusion,and the identification of a high-intensity zone within the annular fibers (Fig. 1A,B). CTdiscography and provocative and anesthetic testings are reliable diagnostic modalitiesin identifying a pain-producing intervertebral disc. The presence of posterior and pos-terolateral subligamentous tears and fissures that reproduce the patient’s symptomatol-ogy is a reliable guide for diagnosis of an abnormal and symptomatic intervertebraldisc. The diagnostic task becomes simpler when degeneration is limited to a singlespinal unit. However, in our experience, the identification of a single, symptom-produc-ing intervertebral disc in the presence of multilevel degenerative changes has been dif-ficult and somewhat unreliable.

Diagnostic arthroscopic probing that was described by Shepperd (8) may be a usefultest for identifying a painful high-intensity zone that has been diagnosed with an MRIstudy. Under local anesthesia, a 5-mm-inner diameter (id) cannula is positioned in thetriangular working zone adjacent to the spinal canal (see Chapter 4). A working scope isused, and a blunt-end probe is employed to palpate the posterior annulus and the site ofthe high-intensity zone. Severe pain simulating the patient’s symptomatology is invari-ably experienced by the patient on palpation of the granulation tissue. In addition, thecessation of symptoms following local infiltration of xylocaine solution into the annularfibers further confirms the presence of annular pathology.

TREATMENT MODALITIES

Transforaminal Epidural and Perineural Steroid Therapy

The technique of transforaminal steroid therapy is an offspring of foraminal nee-dle placement, which was developed in the early stages of percutaneous lumbar dis-cectomy (11–13) (see Chapter 4). Invariably, owing to the presence of epiduraladhesions in an already operated spine, the injected steroid in the middle or upperlumbar spine will not descend to the site of inflamed structures in the lower lumbarregion. Therefore, transforaminal access allows delivery of the steroid compounddirectly to the epidural space, traversing and exiting the roots. In our experience,transforaminal steroid therapy has been helpful in the management of patients whoare suffering from chronic radiculopathy or those who present with localized low-back pain owing to degenerated bulging intervertebral discs. The procedure is per-formed in a sterile environment. Needle positioning is similar to what was describedunder arthroscopic and endoscopic microdiscectomy in previous chapters. Underfluoroscopic control, the tip of an 18-gage needle is positioned at the midpedicularline in anteropasterior (AP) fluoroscopic examination. Then a mixture of appropriatesteroid compound and anesthetics is injected into the foramen adjacent to the epiduralspace and the nerve roots.

122 Kambin

Arthroscopic Annular Debridement

Subligamentous access to the intervertebral disc at the index level is first established(see Chapter 2). Then a posterior quadrant nucleotomy is performed, and the torn annu-lar fibers are excised with a forceps and vaporized with a radiofrequency probe underfluid medium. Sometimes it is difficult to differentiate visually the nucleus and tornannular tissue. However, intradiscal injection of diluted indigen carmine will stain thenucleus and torn annular fibers, thus assisting the visual diagnosis of torn annular fibers.When torn annular tissue has been extracted, the fibers of the posterior longitudinalligament that run perpendicular to the vertebral plate are identified. This indicates thatadequate debridement has been accomplished (see Fig. 19A–C in Chapter 2).

Endoscopic Neurolysis of Nerve Roots

Recurrence of sciatic pain owing to perineural scar and tethering of the nerve rootsfollowing a successful laminatomy and discectomy is not uncommon. In our limitedexperience, we have had impressive success with endoscopic neurolysis and excision ofepidural and perineural scar formation (see Fig. 39 in Chapter 4).

Reduction of Nuclear Mass and Partial Arthroscopic Decortication of Vertebral Plates

Using a uniportal approach, the nucleus is resected and the vertebral plates are partiallydecorticated. Pain relief following this procedure may be attributed to a reduction inintradiscal pressure; decreased tension on the posterior longitudinal ligament; and a grad-ual reduction in the height of the intervertebral discs, fibrous union, and autostabilization.

My limited experience with this technique has proven satisfactory in approx 65% ofindividuals with a single degenerated disc. None of the patients in this group had clini-cal or imaging evidence of lateral stenosis preoperatively. The outcome was more grati-fying when free fragments of collagenized nucleus were found within the intervertebraldisc and extracted (see Fig. 15A,B in Chapter 2).

Expandable Cages

An expandable cage may be introduced into the intervertebral disc through a cannulapositioned posterolaterally. The cage is then filled with cancellous bone (14).

Shepperd (8) experimented with a rigid metallic disc spacer that was introduced intothe intervertebral disc using a posterolateral approach. However this procedure remainsinvestigational.

Disc Prosthesis

In the last 10 yr, many investigators have attempted to design and implant a discprosthesis that provides stability and allows physiological motion between the lumbarmotion segments (15–19). Long-term clinical studies dealing with this technique arenot yet available. Various endoprosthetic designs have been proposed: low-friction ball-and-socket surfaces (20), an artificial disc with contained fluid chambers (21), a springand hinge system (22), and an artificial disc made of rubber and various elastomers. Theongoing investigation in our institution for replacement of the nucleus pulposus alonevia a minimally invasive technique appears promising (23). In this technique, the

Discogenic Pain and Spinal Instability 123

nucleus is replaced with a dehydrated biocompatible hydrogel polymer that is passedthrough a cannula and positioned in the center of the intervertebral disc by using a pos-terolateral approach. The hydrogel is then hydrated and allowed to expand and reach itspredetermined shape and size within the confines of the annulus.

Thermal Intradiscal Therapy

Saal and Saal (24) recently reported on their experience with intradiscal thermal therapyon 25 patients presented with chronic discogenic low-back pain. A thermal catheterwas introduced into the intervertebral disc via a posterolateral approach. Thecatheter’s temperature was gradually raised to 90°C over a period of 13 min and wasmaintained for 4 min (24). I have had no experience with this technology.

SEGMENTAL INSTABILITY AND ITS MANAGEMENT

The term segmental instability has not been well defined in the literature. However, itis commonly used as an indication for surgical fusion of a spinal unit. White and Panjabi(9) defined clinical instability as “loss of the ability of the spine under physiologic loadsto maintain relationships between vertebrae in such a way that there is neither damagenor subsequent irritation to the spinal cord or nerve roots.”

There are various kinds of segmental instabilities. Frymoyer and Selby (25) classifiedthe various instabilities as axial, rotational, translational, retrolisthetic, and postsurgical.

There has been no uniform consensus among surgeons on the indications for spinalfusion. No generally accepted criteria have been developed or published on this subject.The following criteria described in the literature appear to be arbitrary and subjective:“prolapsed intervertebral disc in a young patient who wishes to return to the same typeof manual work,” “prolapsed intervertebral disc with disc-space narrowing,” “primarycentral disc herniation,” “disc herniation with a long-standing history of back pain,”and “back pain being greater than leg pain.”

Although these criteria are useful in deciding whether or not to fuse a given motionsegment, they do not provide an objective assessment of the source of pain and disabil-ity, nor does they ensure that the patient will indeed benefit from the arthrodesis.

Clinically, the surgical stabilization of a motion segment is justified under the fol-lowing set of circ*mstances:

1. When the integrity of the stabilizing structures of the motion segment has been compro-mised: This includes developmental instability associated with a defect in the pars interar-ticularis; surgically induced instability; and posttraumatic fracture or ligamentous injuries,infection, and osteomyelitis.

2. When the ability of the intervertebral disc to contain and transmit the external forces hasbeen altered: This includes localized degenerative disc pathology as well as degenerativespondylolisthesis and retrolisthesis associated with radiographic evidence of hypermobilityon lateral flexion and extension or AP side-bending films. In addition, symptom-producingadult scoliotic curves fall in this category.

In a clinical setting, generally there are three requisites for surgical stabilization ofspinal units: disabling back pain, positive provocative and anesthetic testing, and abnor-mal dynamic studies (Fig. 2).

Fusion of the vertebrae is the oldest and still the most common and acceptablemethod of treating disabling low-back pain. Posterior stabilization, advocated by Albee(26) in 1911 and Hibbs (27) in 1912 for the treatment of Pott’s disease, has also been

124 Kambin

used by clinicians for stabilization of painful degenerative conditions of the lumbarspine. The contributions of Watkins (28) and Wiltse et al. (29) to the concept of postero-lateral fusion also deserve recognition.

Cloward (30) should be credited for the description and popularization of anteriorcolumn stabilization. This remains the cornerstone of the current method for treatingpainful and unstable discogenic lumbar spine pain. The broad contact surface of thevertebral plates in the lumbar region, its adequate blood supply, and a natural exposureof interbody grafts to compression forces while the patient is ambulatory (12–31) havecontributed to the high success rate and wide acceptance of anterior column stabilizationwhen fusion of lumbar segments is deemed necessary. The addition of internal skeletalfixators by Harrington and other investigators further revolutionized the art of stabilizingspinal segments without the need to use casts or bulky braces (32,33).

The following advances in the field of minimally invasive spinal surgery may beresponsible for the renewed interest in arthroscopic interbody fusion:

1. Identification of radiographic and arthroscopic landmarks of the triangular working zone onthe dorsolateral corner of the annulus fibrosis for safe positioning of cannulas (11,12,34).

2. Development of technology for percutaneous insertion of pedicular fixators (13,33,35).3. Availability of manual and power-driven tools that may be passed through cannulas to

access and decorticate the concave surfaces of the vertebral plates.4. Availability of osteoinductive bone proteins, gene therapy, and tissue engineering to

enhance osteogenesis at the fusion site.

The concept of percutaneous interbody fusion that was initiated in the mid-1980sevolved from earlier experience by my colleagues and I with percutaneous discectomy.This was followed by positioning of pedicular screws through inserted cannulas andsubcutaneous (sc) placement of the plates in the early 1990s (36), therefore adding anew dimension to the final outcome of the operative procedure.

Although Magerl (33) utilized percutaneously inserted screws and external fixatorsfor stabilization of thoracolumbar spine fractures, we thought there was a need for morecontrol and more accurate placement of pedicular screws. Therefore, we positioned aguide pin in the medullary canal of the pedicle under fluoroscopic control. This was

Discogenic Pain and Spinal Instability 125

Fig. 2. The three prerequisites for surgical stabilization of the spinal unit.

then followed by introduction of a 10-mm-id cannula adjacent to the site where the guidepin entered the pedicle. This permitted tapping of the medullary canal in the pedicleand probing of its cortices prior to final insertion of the pedicular screws. In addition,sc placement of the plates eliminated the incidence of pin tract infection and complica-tions associated with the use of bulky external fixators.

Leu and Schreiber (37) have published on their experience with percutaneous lumbarfusion and the use of AO external fixators using the bull’s-eye technique.

Arthroscopic Anterior Column Stabilization Augmented With Percutaneously Inserted Pedicular Bolts and sc Plates

In addition to arthroscopic microdiscectomy instruments, special tools have beendeveloped to facilitate successful performance of anterior column stabilization. In ourinstitution, we have discontinued the use of telescopic cannulas that allowed the inser-tion of a 9 × 9 mm id cannula into the intervertebral disc through which we proceededwith decortication of the vertebral plates and insertion of bone grafts (31). Undueexpansion of the height of the intervertebral disc was felt to be responsible for stretch-ing of the nerve root and postoperative development of pain and dysesthesia. A 5 × 10mm id oval cannula (Fig. 3) allows insertion of larger pituitary and cup forceps forrapid removal of nuclear tissue. In addition, this larger cannula provides passage ofinserted decorticators to access the concave surfaces of the vertebral plates.

A cannulated soft-tissue dilator with a 9-mm outer diameter provides access to thepedicles without damaging the surrounding soft tissues. It also allows the proper posi-tioning of a 10-mm-id pedicular cannula over the four pedicles adjacent to the fusionsite (Fig. 4). Cannulated pedicular bolts and extension bars of various lengths (Fig. 5)permit precise positioning of the bolts within the medullary canal of the pedicles overthe previously inserted guide pins.

126 Kambin

Fig. 3. Tools to facilitate successful performance of anterior column stabilization: from topto bottom, two cannulated obturators inserted into obturator jig, 5 × 10 mm id oval cannula, andtwo sizes of vertebral plate decorticators.

Discogenic Pain and Spinal Instability 127

Fig. 4. Additional tools for anterior column stabilization: from left to right, Steinman pin,cannulated obturator, four 10-mm-id pedicular cannulas.

Fig. 5. Various sizes of pedicular bolts with straight and step-off extension bars.

Principles of Operative Technique

The patient is positioned prone on a radiolucent frame and table. General or spinalanesthesia is usually required. A biportal access to the intervertebral disc adjacent to thefusion site is established (see Chapter 3) (7,13,31,34,37–39) (Fig. 6A,B). The ipsilateralinsertion of a 5 × 10 mm id oval cannula facilitates ample access to the intervertebraldisc for extraction of the nuclear tissue, decortication, and insertion of the bone grafts(Fig. 7A–D). Meticulous nucleotomy under arthroscopic control is necessary prior todecortication of the vertebral plates. We have used both laser light and radiofrequencyprobes for nuclear debulking (13–34). Chemopapain solution may be used for a rapid

128 Kambin

Fig. 6. (A) Interoperative AP view of biportal access to intervertebral disc. Note the insertionof a 5 × 10 mm id oval cannula on the left and a 5 mm id cannula on the right. A Steinman pinhas been inserted into the pedicle of L4. (Printed with permission from ref. 13.) (B) LateralX-ray view of position of cannula as shown in (A).

and satisfactory nucleolysis. This step is then followed by ample irrigation of the inter-vertebral disc space. Decortication of the vertebral plates is then accomplished withappropriate decorticators and a curved curette. Bone grafts are harvested from the iliumand packed between the decorticated vertebral plates.

The use of pedicular fixation has been essential to the success of arthroscopicanterior column stabilization. To ensure accurate placement of the pedicular bolts, a

Discogenic Pain and Spinal Instability 129

Fig. 7. (A) A gooseneck curette has been passed through the oval cannula for manual decor-tication of the vertebral plates. (Reprinted with permission from ref. 13.) (B) A plate decorticatoris being used for removal of the cartilaginous surface of the vertebral plates via a biportalapproach. (C) A laser light is being utilized for removal of nuclear tissue. (D) Decorticated ver-tebral plates are shown.

Steinman pin is first inserted into the medullary canal of the pedicle of segmentslocated above and below the fusion site. The skin entry site is determined by review-ing and measuring the preoperative axial CT study of the vertebra above and belowthe fusion site. Although the bull’s-eye technique has been used to ensure the properpositioning of the pedicular screws (33–37), we have found that positioning of theguide pin under fluoroscopic control in both the AP and lateral projections providesmore information and results in better final placement of the pedicular bolts (13–39).

The inherent inaccuracy of the bull’s-eye technique is related to three factors:1. The procedure requires tilting of the C-arm away from the pedicle that is being probed or,

alternatively, tilting the patient and the operating table away from the C-arm that is positioned

130 Kambin

Fig. 7. (Continued)

for AP fluoroscopic examination. The tilt of the C-arm directly affects the direction andfinal position of the inserted guide pins into the pedicle. Peripheral X-ray beams may dis-tort and alter positioning of the guide pin and, ultimately, the position of the pedicularbolt. An arbitrary 20−30° tilt of the C-arm may not ensure proper positioning of the guidepins in every individual and in various locations of the spinal column. An accurate preop-erative determination of the anatomical angle of the pedicle in relation to the vertebralbody of the same segment provides more specific information for insertion of the guidepins.

2. The skin entry points of the guide pins also play an important role in the final positioning ofthe guide pins in the pedicles. For example, excessive tilting of the X-ray tube will requirethat the skin entry site be placed more laterally, which might not be desirable.

3. The distance between the X-ray tube and the patient may affect the final position of theguide pins.

In our institution, prior to the surgical procedure, we perform a prone axial CTstudy through the pedicles of vertebrae adjacent to the fusion site. It is desirable torequest an axial view that accurately reflects the exact size of the individual’s vertebralbodies (Fig. 8).

A longitudinal line (AB) that bisects the vertebral bodies and extends to the center ofthe spinal process of the same segment is drawn on the axial CT scan film. Another line(BC) that represents the desirable position of the guide pin is drawn through the pedicle.Line AC represents the distance between the midline and the skin entry point. The anglebetween lines AB and BC represents the desired angle of insertion of the guide pin (orthe tilt of the C-arm if the use of the bull’s-eye technique is contemplated).

Discogenic Pain and Spinal Instability 131

Fig. 8. Axial CT scan of segment adjacent to fusion site. Line AB passes through the spinal processand bisects the vertebra. Line BC represents the desired direction of the pedicular bolt. Line AC rep-resents the distance from the midline and the desired site of skin incision for positioning of thepedicular cannula and insertion of the pedicular bolts. (Reprinted with permission from ref. 13.)

At the onset of the operative procedure, while the patient is in a prone position underfluoroscopic control, a guide pin is positioned over the spinal processes of the lumbarsegments. The midline of the spinal column is identified, and the skin surface is markedaccordingly (Fig. 9).

Two guide pins are then placed over the pedicles of the segments that require stabi-lization. Accurate position of the guide wire is confirmed radiographically. A traverseline is then drawn on the skin surface perpendicular to the previous line drawn over thespinal processes. Information obtained from measurement of the axial CT scan is thentransferred to the surgical site. The skin entry point (Fig. 8, lines A and C) is measuredform the midline, and the guide pin is introduced manually and directed toward thepedicle with a rotary movement. AP and lateral fluoroscopic images confirm the posi-tioning of the guide pin on the pedicle. In the AP projection the tip of the insertedguide wire should be seen in the lateral boundary of the pedicle (Fig. 10A). This step isthen followed by full insertion of the guide pin into the medullary canal of the pedicleand the vertebral body of the same segment. The proper direction of the inserted guidepin should be monitored in the lateral fluoroscopic examination (Fig. 10B).

Following satisfactory positioning of the Steinman pin, a skin incision about l cm inlength is made adjacent to the guide pin, and a blunt-end soft-tissue dilator and pedicularcannula are introduced until they reach the facet and pedicles of the vertebra (Fig. 11A).While the pedicular cannulas are firmly held against the pedicles, a cannulated pediculartap is used and passed over the guide pin, and the medullary canal of the pedicle is tappedin preparation for final insertion of the pedicular bolts. Prior to insertion of the pedicular

132 Kambin132 Kambin

Fig. 9. Schematic drawing. Line A represents a midline drawn over the spinal processes.Lines B and C represent lines drawn through the center of the pedicles adjacent to the fusionsite. Line D represents the distance from the midline and the site of the skin incision for insertionof the pedicular cannula.

bolts, the medullary canal of the pedicle is examined and probed to make certain that thecortices of the pedicle have not been violated (Fig. 11B). Subcutaneous placement of theplates requires that the pedicular bolts be lengthened to reach the sc region. Extensionbars (Fig. 5) of appropriate length are then selected and attached to the proximal end ofthe pedicular bolts. Stabilization is completed when the bolts are linked and firmly held toone another by a plate that is placed in the sc region of the lumbar spine (Fig. 12A–D).

Discogenic Pain and Spinal Instability 133

Fig. 10. (A) Interoperative AP fluoroscopic examination shows the proper positioning of theguide pin adjacent to the lateral boundary of the pedicle. (B) The lateral view of the fluoroscopicexamination shown in (A) shows the appropriate direction of the guide pin in the center of thepedicle.

Screw Fixation of First Sacral Segment

Percutaneous anterior column stabilization between the fifth lumbar vertebra and thesacrum has been challenging. The height of the iliac crest, particularly in male individuals,will interfere with establishing biportal access to the intervertebral disc for nucleotomy,decortication, and bone grafting. However, partial resection of the cephalad vertebral

134 Kambin

Fig. 11 (A) Interoperative photograph demonstrating that cannulas A and B have beeninserted into intervertebral disc at index level and four pedicular cannulas are positioned overpedicles in preparation for insertion of pedicular bolts. The needle in the center was used forspinal anesthesia. (B) Interoperative photograph demonstrating how medullary canal of pedicleis being palpated to make certain that cortices of pedicle have not been disturbed.

Discogenic Pain and Spinal Instability 135

Fig. 12. (A) Interoperative lateral X-ray view of surgical site. Note the position of the pedicularcannulas and insertion of pedicular bolts over the previously positioned guide pins. (B) Interoper-ative AP fluoroscopy shows that four pedicular bolts are properly positioned. An angled-ringcurette is used for further decortication of the vertebral plates in preparation of introduction of thebone grafts. (C) Postoperative X-ray study shows the proper position of the pedicular bolts andextension bars, and sc position of the plates. (D) Lateral fluoroscopic examination shown in (C).Note the inserted bone grafts at L4-L5 and sc position of the vertebral plates.

plates (34) adjacent to the fusion site through the previously positioned cannula willprovide better access to the intervertebral disc at L5-S1.

Percutaneous insertion of pedicular screws to the first sacral segment also is difficult andtime-consuming. Our experience with paramedial miniexposure of L5-S1 intervertebral

136 Kambin

Fig. 12. (Continued)

disc and the facet joints for partial facetectomy and exposure of the triangular workingzone has provided ample access to the intervertebral disc and articular process of the firstsacral segment for insertion of the pedicular screw. Following extensive disc resection viaa posterolateral approach and decortication of the vertebral plates, tricortical and cancel-lous bone is packed at the fusion site. Insertion of the pedicular bolt into the fifth lumbarvertebra and the sacrum is then accomplished under direct vision or via endoscopic illumi-nation and magnification. Various approaches and techniques have been used for distal fix-ation of the construct to the sacrum or iliac crest. Most of the difficulties are related toinadequacy of bone quality.

Discogenic Pain and Spinal Instability 137

Fig. 13. (A) AP and (B) lateral postoperative X-ray study of a 45-yr-old laborer presentedwith Grade II spondylolisthesis of L5-S1 associated with L5 radiculopathy. The patient also haddegenerative disc disease with positive CT discography and anesthetic testing at L4-L5. Percuta-neous insertion of the pedicular bolts at L4-L5 with percutaneous fusion at the same level wasperformed. Insertion of the sacral screw and interbody fusion at L5-S1 were accomplished via amini-incision. Note the medial and distal direction of the sacral screw. Fibular cortical bone andautogenous cancellous bone were used for fusion at L5-S1.

138 Kambin

Fig. 14. (A) AP and (B) lateral postoperative X-ray study of posterolateral arthroscopicfusion at L5-S1 and sc positioning of plates.

In our clinic, we have obtained satisfactory sacral fixation by inserting the screwdorsomedially (Fig. 13A,B). The entry point is at the distal end of the articular pro-cess of the first sacral segment. A guide wire is inserted and directed anteromediallyinto the vertebral body of the first sacral segment. The guide wire bypasses thespinal canal and the S1 and S2 nerve root. The cortical bone of the articular processof the first sacral segment provides adequate cortical support and anchorage for theinserted screw.

The proper position of the guide wire is then evaluated via AP and lateral fluoro-scopic examination. This step is followed by insertion of a cannulated pedicular tap and

Discogenic Pain and Spinal Instability 139

Fig. 15. (A) Preoperative lateral X-ray of a 40-yr-old male patient who presented with radio-graphic evidence of retrolisthesis at L3-L4 and intractable back and anterolateral right thighpain. (B) Postoperative lateral X-ray following percutaneously inserted pedicular bolts, reduc-tion of retrolisthesis, and sc fixation. The patient reported relief following stabilization of theabove motion segment and use of a brace for 4 wk. Subsequently he underwent arthroscopicposterolateral interbody fusion with a satisfactory outcome.

of a sacral screw of appropriate length and diameter. Owing the angle of insertion of thescrew into the sacrum, crowding of the proximal end of the inserted screws may occur,therefore interfering with placement of connecting plates or rods.

140 Kambin

Fig. 16. (A) Preoperative lateral X-ray of a 54-yr-old male patient who presented withspondylolisthesis of L4-L5 and related clinical symptoms. (B) Postoperative lateral X-ray ofsame patient shown in (A) demonstrating percutaneously inserted pedicular bolts into pediclesof L4-L5 and sc placement of connecting plates. This patient had a satisfactory outcome follow-ing initial use of fixators and reduction of spondylolisthesis. He subsequently underwent percu-taneous arthroscopic interbody fusion via a posterolateral approach.

Discogenic Pain and Spinal Instability 141

Fig. 17. (A) Postoperative CT of fusion site. Although part of the bone grafts have beenabsorbed, a solid fusion and extension of the bony trabeculae across the fusion site is demonstrated.(Reprinted with permission from ref. 13.) (B) Reconstruction CT of fusion site shown in (A).

Attachment of an offset extension bar to the sacral screw will provide ample separa-tion between the L5-S1 screws for insertion of the plates or rods (Fig 14A,B).

Diagnostic Stabilization of Motion Segments

Unstable lumbar segments may be stabilized percutaneously via pedicular boltsand sc plates for a period of 3 to 4 wk. If the patient’s symptoms are resolved andthere is evidence of objective improvement of the presenting symptomatology, arthro-scopic interbody fusion is then attempted using the criteria described previously(Figs.15A,B and 16A,B).

142 Kambin

Fig. 18. Showing dislodged pedicular bolt into spinal canal. The hardware was later surgicallyremoved.

RESULTS

In our institution, we have achieved 95% solid fusion and stabilization using thedescribed surgical technique (Fig. 17A,B). None of our patients encountered infectionor neurovascular complications. There were no instrument failures. All of the hardwarewas retrieved on an average of 7 mo postoperatively. In one overweight patient, a frac-ture of the pedicle and clinical evidence of paresthesia of the involved extremity wereassociated with displacement of the pedicular bolt into the spinal canal. Surgical removalof the hardware within 24 h was associated with recovery from the neurological deficitand fibrous union of the spinal unit (Fig. 18).

THORACOSCOPIC AND LAPAROSCOPIC DECOMPRESSION AND STABILIZATION

A thoracoscopic approach to thoracic and laparoscopic access to the lumbar spinehas been used for stabilization of motion segments (40). Although in skilled handsaccess to and decompression of L5-S1 segments should not present a serious difficulty,the presence of major vessels in the middle and upper lumbar spine may interfere withproper midline access, adequate decompression, fusion, or placement of spacers. Ateam of surgeons is usually required to perform this operative procedure. At present,this approach is being used in only a few select centers.

REFERENCES

1. Parke WW. Clinical anatomy of the lower lumbar spine, in Arthroscopic Microdiscectomy,Minimal Intervention in Spinal Surgery (Kambin P, ed.), Urban & Schwarzenberg, Baltimore,1991, pp. 11–29.

2. Falconer MA, McGeorge M, Begg AC. Observations on the cause and mechanism of symp-tom-production in sciatica and low-back pain. J Neurol Neurosurg Psychiatry 1948;11,13–26.

Discogenic Pain and Spinal Instability 143

3. Kambin P, Nixon J, Chait A, et al. Annular protrusion: pathophysiology and roentgeno-graphic appearance. Spine 1988;13:671–675.

4. Sachs B, Vanharanta H, Spivey M, et al. Dallas discogram description: a new classificationof CT/discography in low back disorders. Spine 1987;12:287–294.

5. Vanharanta H, Sachs B, Spivey M, et al. The relationship of pain provocation to lumbardisc deterioration as seen by CT/discogram. Spine 1987;12:295–298.

6. Weinstein J, Claverie W, Gibson S. The pain of discography. Spine 1988;13(12):1344–1348.7. Kambin P. The role of minimally invasive surgery in spinal disorders. Adv Operat Orthop

1995;3:147–171.8. Shepperd JAN. Percutaneous and minimal intervention spinal fusion, in Arthroscopic Microdis-

cectomy, Minimal Intervention in Spinal Surgery (Kambin P, ed.), Urban & Schwarzenberg,Baltimore, 1991, pp. 127–129.

9. White A, Panjabi MM. Clinical Biomechanics of the Spine, JB Lippincott, Philadelphia, 1978.10. Saal, JS, Franson RC, Dobrow R, Saal JA, White AH, Goldthwaite N. High levels of inflam-

matory phospholipase A2 activity in lumbar disc herniations. Spine 1990;15(7): 674–678.11. Kambin P. Percutaneous lumbar discectomy: current practice. Surg Rounds Orthop

1988;31–35.12. Kambin P. Arthroscopic microdiscectomy. Arthroscopy 1992;8(3):287–295.13. Kambin P, Gennarrelli T, Hermantin F. Minimally invasive techniques in spinal surgery:

current practice. Neurosurgical Focus 1998;4(2):Article 8.14. Kambin P. Expandable intervertebral cage and surgical method. US patent 5,665,122, 1997.15. David T. Lumbar disc prosthesis, surgical technique, indications and clinical results in 22

patients with a minimum of 12 months follow up. Eur Spine J 1993;1:254–259.16. Enker P, Stefee A, McMillin C, Keppler L, Biscup DO, Miller DO. Artificial disc replace-

ment. Spine 1993;18:1061–1070.17. Griffith SL, Shelokow AP, Buettner-Janz K, LeMaire JP, Zeegers WS. A multicenter retro-

spective study of the clinical results of the link SB Charite intervertebral disc prosthesis.Spine 1994;19:1842–1849.

18. Lemaire JP, Skalli W, Lavaste F, et al. Intervertebral disc prosthesis. Clin Orthop1997;7:337:64–76.

19. Scott AH, Harrison OJ. Increasing age does not affect outcome after lumbar disc replace-ment. Int Orthop Sicot 2000;24:50–53.

20. Sali RM, Pettine KA. Intervertebral disk arthroplasty. US patent 5,258,031,2, 1993.21. Monson GL. Synthetic intervertebral disc prosthesis. US patent 4,083,477,5, 1969.22. Hedman TP, Kostuik JP, Fernie GR, Maki BE. An artificial spinal disc. US patent

4,759,769,12, 1988.23. Marcolongo MS, Kambin P, Lowman A, Karduna A. Experience with minimally invasive

nuclear replacement. Fourteenth International Symposium on Arthroscopic and EndoscopicSpinal Surgery May 5–6, 2000, The Graduate Hospital, Philadelphia.

24. Saal JS, Saal, JA. Management of chronic discogenic low back pain with thermal intradis-cal catheter. Spine 2000;25(3):382–388.

25. Frymoyer JW, Selby DK. Segmental instability rationale for treatment. Spine1985;10:280–286.

26. Albee FH. Transplantation of a portion of the tibia into the spine for Pott’s Disease: a pre-liminary report. JAMA 1911;57:885–888.

27. Hibbs RA. An operation for Pott’s disease on the spine. JAMA 1912;59:422–436.28. Watkins MB. Posterolateral fusion of the lumbar and lumbosacral spine. J Bone Joint Surg

Am 1953;34:1014–1018.29. Wiltse LL, Bateman JG, Hutchinson RH, et al. The paraspinal sacrospinalis-splitting

approach to the lumbar spine. J Bone Joint Surg Am 1968;50: 919–926.30. Cloward RB. The treatment of a ruptured lumbar disc by vertebral fusion: indications,

operative technique, after care. J Neursurg 1953;10:154–168.

31. Kambin P. Posterolateral lumbar interbody fusion, in Arthroscopic Microdiscectomy, Min-imal Intervention in Spinal Surgery (Kambin P, ed.), Urban & Schwarzenberg, Baltimore,1991, pp. 117–121.

32. Harrington PR, Tullos HS. Reduction of severe spondylolisthesis in children. South Med J1969;62:1–7.

33. Magerl FP. Stabilization of the lower thoracic and lumbar spine with external skeletalfixation. Clin Orthop 1984;189:125–141.

34. Kambin P. Arthroscopic lumbar intervertebral fusion, in The Adult Spine: Principlesand Practice (Frymoyer JW, Ducker T, Hadler N, et al. eds.), Raven, New York, 1996, pp.2037–2047.

35. Kambin P. Percutaneous fixation of the vertebrae. US patent 5,242,443, 1930.36. Kambin P. Arthroscopic lumbar fusion with pedicular bolts and subcutaneous plates.

ISMISS Scientific Exhibit, Feb 20–25, 1992, AAOS Meeting, Washington, DC.37. Leu HJ, Schreiber A. Percutaneous fusion of the lumbar spine: a promising technique.

Spine State Arts Rev 1992;6:593–604.38. Kambin P. Arthroscopic lumbar interbody fusion, in Spine Care (White AH, Schofferman

JA, eds.), Mosby, St. Louis, 1995, pp. 1055–1066.39. Kambin P. Method for percutaneous arthroscopic disc removal bone biopsy and fixation of

vertebrae. US patent 6,175,758,3, 2001.40. Regan JJ, Yuan H, McCullenG. Thoroscopic and endoscopic access to the spine. Instruc-

tional Course Lectures. AAOS 1997;46:127–141.

144 Kambin

6Lateral Recess Stenosis of Lumbar

Spine Foraminoplasty

Parviz Kambin, MD

INTRODUCTION

In 1900, Sachs and Fraenkel (1) described the diagnosis and treatment of lateralrecess stenosis as an entity. Epstein et al. (2) further clarified it as a distinct clinicalentity. The availability of computed tomography (CT) and magnetic resonance imagingin recent years has facilitated visualization of the content of the lateral recess and diagnosisof this pathological condition (Fig. 1).

The nerve root canal begins from the nerve root sheath and terminates when the exitingroot emerges from the foramina. The superior facet, capsular ligamentous complex (Fig. 18in Chapter 2) forms the posterior boundary or roof of the lateral recess. Expansion of theposterior longitudinal ligamentum to the foramen, the intervertebral disc, and the poste-rior surface of the adjacent vertebral bodies forms the ventral or anterior surface of theforamen. The exiting root occupies the pedicular notch superiorly.

Degenerative changes in the facet joints associated with synovial hypertrophy, thick-ened and fibrotic facet capsules, and ligamentum flavum complex (Fig. 2) contribute tothe narrowing and stenosis of the lateral recess. In addition, marginal osteophytes arisingfrom the vertebral bodies, combined with posterior bulging and protrusion of the interver-tebral disc, cause further restriction, thus adding tension and compression on the exitingnerve root and its vascular structures. It has been shown that interference with thevenous return of the nerve root causes chronic edema of the root, which may becomeassociated with intra- and perineural fibrosis (3–5). The pathophysiology of the bulgingannulus or protrusion has also been described (6–8).

With the advancement of aging, dehydration and collagenization of the nucleus pulposus,combined with tear and disorganization of the annular fibers, plays an important role inthe development of abnormal protrusion of the intervertebral disc (Fig. 17A in Chapter 2).

CLINICAL PRESENTATION

Patients with spinal stenosis are usually seen in the physician’s office with signs andsymptoms of neurogenic claudication and, at times, complaining of numbness or a feelingof pins and needles in the lower extremities (9,10).

145

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

The symptoms are diminished when the patient sits or reclines. This is in contrast tovascular claudication, for which the symptoms subside when the patient stops walking.Individuals with lateral recess stenosis have a tendency to bend forward while walking.Extension of the lumbar spine invariably is associated with pain. Neurological examinationusually is not revealing; no reflex abnormality, sensory deficit, or positive tension signsare found.

146 Kambin

Fig. 1. Preoperative axial CT scan study of a 60-yr-old male presented with signs and symptomsof bilateral lateral recess stenosis. Note degenerative changes of the facet joints, narrowing ofthe foramen, bulging of the annulus.

Fig. 2. Schematic drawing demonstrating how posterior marginal osteophytes from vertebralbodies combined with inflamed and hypertrophic facet capsules contribute to stenosis of nerveroot canal.

SURGICAL MANAGEMENTThe evolution of minimally invasive spinal surgery and the availability of microbipolar

electrocoagulators, radiofrequency probes, and flexible-tip microinstruments have permittedaccess to both the ventral and posterior boundaries of the neural canal. Reshaping of thedimensions of the lateral recess via resection of the compressive elements under arthro-scopic illumination and magnification has become a standard operative procedureamong minimally invasive spine surgeons.

As early as 1988, my colleagues and I used mechanical tools successfully for removal ofposterior osteophytes and resection of fibrotic and bulging annulus for the treatment of lat-eral recess stenosis (11,12). Subsequently, we were able to utilize a radiofrequency probefor vaporization of the inflamed facet capsules and the ligamentum flavum that were con-tributing to the clinical manifestation of lateral recess stenosis. In recent years, laserlights via a flexible-tip working scope have been used for ambulatory treatment of spinalstenosis (13,14).

Arthroscopic access to the lateral recess requires further lateralization of the skinentry site. This allows insertion of the cannula in the foramen and provides access to thecompressive elements on both the ventral and dorsum of the nerve root foramen. Whenin doubt, particularly when surgery is being attempted in the upper lumbar spine, it isadvisable to secure a preoperative prone CT scan study from the surgical site. This willensure safety of the content of the abdominal cavity and its vital structures.

The needle is positioned at the midpedicular line as observed in the anteroposteriorfluoroscopic examination. This step is followed by introduction of the cannulated obtu-rator and positioning of the working cannula (see Chapter 3). Under arthroscopiccontrol, mechanical tools may be used for removal of annular protrusion and marginalosteophytes that are arising from the vertebral plates adjacent to the intervertebral disc(Fig. 3). We have used a prebent radiofrequency probe for vaporization of the articularcapsule and inflamed synovial tissue.

RESULTS

The outcome of a prospective study of 40 consecutive patients who underwentarthroscopic foraminal decompression of the lateral recess stenosis was published in1996 (11). The reported outcome of patients who underwent arthroscopic decompressionof lateral recess has been compatible or better than the reported result following extensiveopen operative procedures (15,16).

DEVELOPMENT OF DYSESTHESIA

Approximately 4 to 5 d following the surgical procedure, patients began to experiencea burning sensation or hypersensitivity of skin to touch affecting the involved extremity(17). At times, patients are unable to use covers on their legs after this type of surgery.This subjective complaint of hypersensitivity usually is not associated with objectiveneurological deficit. Reflex abnormality, weakness, atrophy, or sensory deficit is notusually found. The dermatomal distribution of the dysesthesia at times is not clear. How-ever, it may follow the pattern of sensory nerve supply of a given nerve root at the site ofthe surgical procedure. The etiology of this disturbing complication is attributed tomanipulation, excess heat, or trauma to the nerve root ganglia.

Lateral Recess Stenosis 147

MANAGEMENT OF DYSESTHESIA

The operating surgeon has the responsibility to prepare and warn the patient ofpotential development of dysesthesia following surgery. This reduces or prevents theundue anxiety that invariably accompanies this organic disorder. In our experience,

148 Kambin

Fig. 3. (A) Intraoperative arthroscopic view of lateral recess of patient shown in Fig. 1. Notehow the osteophytes that are arising from the vertebral body of the proximal segment contributeto the development of root canal stenosis. (B) The fibers of the bulging annulus are seen distal tothe above osteophytes.

intraoperative injection of diluted fentanyl solution around the nerve root ganglia atthe onset of the operative procedure reduces the incidence of development of postop-erative dysesthesia.

Following proper positioning of the patient and insertion of an 18-gage needleinto the foramen, a mixture of 1 cc of fentanyl and 3 cc of saline solution is injectedinto the foramen. Within a few minutes following the injection, the surgeon may pro-ceed with positioning of the cannulated obturator and working cannula, and comple-tion of the operative procedure. Bathing of the root ganglia in the fentanyl solutionwill have a tendency to alter the sensitivity of the root ganglia to external stimulation.Fentanyl-induced antinociceptive effect is supraspinally mediated. It interacts withopioid receptors that are present in the dorsal ganglia and dura and the central ner-vous system (18).

In addition, it is advisable to inject a diluted solution of a steroid compound into theforamen prior to withdrawal of the instruments. Postoperatively the majority of patientsrespond favorably to the use of oral nonsteroidal anti-inflammatory medications andanalgesics within 4–6 wk. However, when the presenting symptoms are severe, under astrict sterile environment, the patient is positioned prone on the operating room tableand the 18-gage needle is reinserted into the foramen according to the previouslydescribed steps. Injection of the steroid compound around the nerve root ganglia invari-ably provides relief and enhances recovery time.

REFERENCES

1. Sachs B, Fraenkel J. Progressive ankylotic rigidity of the spine (spondylose rhizomelique).J Nerv Ment Dis 1900;27:1–15.

2. Epstein JA, Epstein BS, Rosenthal AD, et al. Sciatica caused by nerve root entrapment inthe lateral recess: the superior facet syndrome. J Neurosurg 1972;36:584–589.

3. Hoyland JA, Freemont JA, Jayson MIV. Intervertebral foramen venous obstruction: a causeof periradicular fibrosis? Spine 1989;14:538–568.

4. Olmarker K, Rydevik B, Holm S. Edema formation in spinal nerve roots induced by exper-imental graded compression: an experimental study on the pig cauda equina with specialreference to differences in effects between rapid and low onset of compression. Spine1989;14:569–573.

5. Parke WW. The significance of venous return impairment in ischemic radiculopathy andmyelopathy. Orthop Clin North Am 1991;22:213–222.

6. Kambin P, Nixon JE, Chait A, et al. Annular protrusion: pathophysiology and roentgeno-graphic appearance. Spine 1988;13:671–675.

7. Kambin P, McCullen G, Park W, et al. Minimally invasive arthroscopic spinal surgery.Instruct Course Lect 1997;46:143–161.

8. Schaffer JL, Kambin P. Minimally invasive spine surgery. Textbook Rheumatol Update1994;9:2–12.

9. Kirkaldy-Willis WH, Paine KW, Cauchois J, McIvor G. Lumbar spinal stenosis. Clin Orthop1974;99:30–50.

10. Yamada H, Oya M, Okada T, Shiozawa Z. Intermittent cauda equina compression due tonarrow spinal canal. J Neurosurg 1972;37:83–88.

11. Kambin P, Casey K, O’Brien E, Zhou L. Transforaminal arthroscopic decompression oflateral recess stenosis. J Neurosurg 1996;84:462–467.

12. Kambin P. Arthroscopic treatment of spinal pathology, in Operative Arthroscopy, 2nd ed.(McGinty JB, Casperi RB, Jackson RW, Poehling GG, eds.), Lippincott-Raven, Philadelphia,1996, pp. 1227–1233.

Lateral Recess Stenosis 149

13. Chiu J. Transforaminal endoscopic micro decompression of herniated lumbar discs andspinal stenosis. 22nd International Course for Percutaneous and Endoscopic Spinal Surgery,Jan 2004; ISMISS/SICOT, Zurich, Switzerland.

14. Knight MTN, Goswami A, Patko JT. Endoscopic foraminoplasty : a prospective study on250 consecutive patients with independent evaluation. J Clin Laser Med Surg2001;19:73–81.

15. Burton CV, Kirkaldy-Willis WH, Yong-Hing K, et al. Causes of failure of surgery on thelumbar spine. Clin Orthop 1981;157:191–199.

16. Ray CT. Transfacet decompression with dowel fixation: a new technique for lumbar lateralspinal stenosis. Acta Neurochir Suppl 1988;43:48–54.

17. Kambin P, O’Brien E, Zhou L. Arthroscopic microdiscectomy and selective fragmentectomy.Clin Orthop 1998;347:150–167.

18. Jaffe RA, Rose MA. A comparison of local anesthetic effect of Meperidine Fentanyl andSufentanil on dorsal root axons. Anesth Analg 1996;83:776–781.

150 Kambin

7Role of Epidural and Radicular Veins

in Chronic Back Pain and Radiculopathy

Wesley W. Parke, PhD

INTRODUCTION

Although the pattern and major components of the human medullary arterial supplywere widely known by the end of the nineteenth century, the details of the associatedvenous circulation were generally ignored. Despite Brechet’s (1) quite accurate descrip-tion and later depiction of the complexity of the spinal venous system, published in1835 (Fig. 1), the ubiquity and variability of its ramifications evidently discouragedfurther consideration beyond the comprehension that the veins collectively served ascollateral channels to the caval and azygos systems. Another factor that may have con-tributed to the general disregard of the anatomical particulars of these channels mayhave been related to their structural delicacy. The thinness and transparency of theirwalls render them almost invisible during conventional cadaver dissections unless theyshow postmortem evidence of congestion or are specifically filled with a contrastmedium. Clemens (2) noted that these vessels were quite pliable, which permitted con-siderable distension under collateral load. Thus, the Queckenstedt maneuver, whichtests the patency of the spinal subarachnoid space by compression of the jugular orintra-abdominal veins, causes an increase in cerebrospinal fluid (CSF) pressure by anexternal compression from the expansion of the collaterally loaded epidural plexus.Clemens also postulated that a passive congestion of the spinal cord was prevented byminute valves in the proximal sources of the radicular veins that drain the spinal cord, aunique situation considering that none of the other sinus veins possess valves. AfterBatson (3), in the mid–20th century, emphasized this fact by demonstrating the exten-sive multidirectional flow allowed in these vessels and its significance as a route ofmetastatic transport of neoplastic cells, interest in the form and function of the spinalvein dramatically increased. Thus, a closer inspection of the initially assumed randomentanglement of these vessels eventually yielded the following fairly consistent under-lying pattern. This pattern shows several (usually four or more) major longitudinalepidural channels that connect a circumferentially coursing series of vessels that are seg-mentally arranged in relation to the intervertebral foramina through which they commu-nicate with a segmentally equivalent perivertebral plexus (Fig. 2). The intervertebralconnecting branches usually consist of a superior set that exits the foramen embracing

151

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

the dorsal root ganglion (DRG) and the exiting nerve roots (Fig. 3), and an inferior setclosely related to the pedicle of the inferior foraminal margin.

The contempory experimentally and clinically derived consensus maintains that muchof the spinal neurogenic pain associated with degenerative changes in the vertebralosseous and soft-tissue components involves a compressive radiculomedullary ischemia.It was understandable that earlier investigations of the mechanical factors thought to beresponsible for this would be approached through studies of the more readily injectablearterial components almost exclusively. The term neuroischemia was then primarilyregarded as an impairment of the arterial flow because the extent and pervasive distribu-tion of venous channels appeared to guarantee an unlimited, trouble-free access to effer-ent vascular drainage.

Shortly after publications by Parke and associates (4,5) established the basic patternsof the radicular blood supply, clinical observations indicated that the thin-walled low-pressure side of the radiculomedullary circulation may play an unsuspected predomi-nant role in the etiology of intradural spinal ischemias. It then became evident that theefferent vascular channels may be directly affected by the spatial encroachments ofdegenerative vertebral tissues and/or be compromised by extravertebral changes in thevenous resistance consequent to general circulatory problems. Regarding the former,Magnaes (6) quantitated some of the compressive effects of spinal stenosis on the

152 Parke

Fig. 1. (A) Posterior and (B) lateralviews of spinal epidural venous plexus taken from a hand-colored copy of Brechet’s original work (published ca. 1835, courtesy of Scott MemorialLibrary, Jefferson Medical College).

lumbosacral roots. In noting that the typical L4–L5 stenosis, which often involves aspondylolithesis compounded by facet joint and ligamentum flavum hypertrophies, mayrestrict the flow of the CSF distal to the level of the lesion, he measured the resulting“spinal block” pressure. This he defined as the amount of elevation in CSF pressure conse-quent to an infusion caudal to the stenosis that was required to force the fluid past the block(Fig. 4).

In a series of 42 patients with clinical and myelographic indications of a central lumbarstenosis, Magnaes (6) was able to determine a pathological degree of pressure on the caudaequina in 67% of cases. He concluded that the block pressure was the main mechanicalfactor responsible for paresis and pain because it was highest during the spinal extension ofstanding and walking and exceeded the mean arterial pressure in several cases. Magnaesalso noted a frequent spontaneous elevation in CSF pressure in the caudal dural sac duringspinal extension and walking but considered this a subordinate factor. Unfortunately, Mag-naes’s work was too early to appreciate that it was the venous side of the radicular circula-tion that was most labile to compressive factors, and his publication appeared 2 yr beforeRydevik et al. (7) demonstrated the nutritional aspects of CSF circulation. In retrospect, itis now apparent that pressures well below those approaching the mean arterial pressurecould have a great effect on the radicular circulation by producing a resistance to venous

Role of Veins in Chronic Back Pain 153

Fig. 2. Schema showing various venous relations of a lumbar vertebra: 1, dorsal externalvertebral plexus; 2, dorsal epidural plexus; 3, ascending lumbar veins; 4, basivertebral vein; 5, ventralexternal vertebral plexus; 6, lumbar segmental vein; 7, muscular vein from posterior abdominal wall;8, circumferential channels of epidural plexus; 9, ventral longitudinal vein; 10, segmental cir-cumferential epidural.

drainage. In addition, the spinal block impedes the free replenishment of the CSF in thecaudal dural sac and further deprives the lower root sections of this source of metabolicturnover through a stagnation of the confined Venous Hypertension lower sac fluids.

VENOUS HYPERTENSION

Clinical investigators have recorded the exacerbation of neurogenic pain in cases inwhich spinal stenosis has been associated with venous hypertension. LaBan (8) andLaBan and Wesolowski (9) noted that patients with diminished right heart complianceand spinal stenosis may eventually show neurogenic pain even in static recumbent situa-tions. They attributed this to an increased external pressure on the already sensitized rootsby the distension of the epidural sinuses. Kaiser et al. (10) visualized through computedtomography (CT) the obstructive production of this epidural venous stasis by a spinalstenosis and its venous engorgement. Because the spinal canal, like the cranial cavity, is arelatively nondistensible container, there is a limited allowance for spatial encroachmentbefore neurological elements are compressed. In a spinal stenosis, the roots are alreadychronically restricted by the circumferential degenerative hypertrophies, but because theepidural venous sinuses also share this compromised space, it is not difficult to conceptu-alize that their distension, owing to increased cardiopulmonary resistance in the cavalvenous return, would provide an additional insult to a chronic situation. Because this painis not immediately manifest when the patient reclines but gradually increases with pro-longed recumbency, often to be initiated after the onset of sleep, LaBan (8) postulatedthe existence of a “venous creep.” This implies that the thin epidural sinus wallsrespond to the increased pressure by a delayed gradual distension that amplifies theepidural expansive pressure. More recently, Madsen and Heros (11) and Parke (12)

154 Parke

Fig. 3. Vertebra of human fetus (38-mm crown-rump length) showing centers of ossification(CC and VC) and, by virtue of obliquity of the cut, left intervertebral foramen. Terminal congestionfills part of the epidural and perivertebral plexus, showing foraminal veins embracing the DRG.Perirenal fascia (f) is indicated.

have shown that “arterialization” of spinal veins by anomalous arteriovenous shunts inthe region of the conus medullaris also exacerbates the neurogenic pain in patients withspinal stenosis. They hypothesized that a variable combination of increased mechanicalconstriction by dilated epidural veins and the direct increased resistance to the radicularcirculation by venous hypertension contributed to initiation of the pain. Aboulker et al. (13)recorded cases of caval anomalies in which the impeded venous return produced symp-toms of cord and root ischemias, and they concluded that epidural venous hypertensionalone may produce radicular and/or cord symptoms without adjunct stenotic compressions.

A graphic depiction of the venous compromise by degenerative encroachments isprovided in Fig. 5, in which the left side of the schema shows the consequences ofsegmental constrictions (spinal stenoses), and the right side indicates how such a situa-tion may be further exacerbated by extravertebral hypertensions similar to those dualsupply of Intradural roots reported by LaBan and colleagues (8,9).

DUAL SUPPLY OF INTRADURAL ROOTS

Because the very long intradural roots of the lumbosacral spinal nerves were withoutaccess to the frequent collateral support characteristic of peripheral nerves and were ini-tially believed to be supplied only from their distal ends, they were subjected to a seriesof studies in an attempt to provide a basic knowledge of the pathophysiology of nerve

Role of Veins in Chronic Back Pain 155

Fig. 4. Schema derived from data provided by Magnaes (6) showing how he determined“spinal block” (SB) pressure through infusion into lower dural sac. The degree of pressurerequired to force the fluid past the spinal block determined the SB pressure.

root nutrition. Subsequently, the injection studies of Parke et al. (4) and Parke andWatanabe (5), using a series of graduated pressure injections, showed that the rootsreceive an arterial supply from both ends through equivalent groups of longitudinallycoursing peri- and intraradicular arteries that maintain a rather consistant caliber untilthey anastomose in the midsections of the radicular fascicles. An additional importantfinding was the occurrence of numerous and relatively large normal arteriovenous anas-tomoses throughout the length of each root (Fig. 6A,B). These vascular shunts appar-ently serve to protect the functional integrity of the radicular circulation in the event offocal compressions. Of particular significance to the knowledge of root metabolismwere the contemporary investigations of Rydevik et al. (7), who, using isotopicallylabeled methyl glucose, demonstrated that approx 50% of the root nutrition is derivedfrom ambient CSF, a fact that necessitates the gauzelike architecture of the pia-arach-noid sheath.

A study of chronically compressed roots by Watanabe and Parke (14) indicated thatthe involved site of compression was most likely metabolically deprived. It had beenstrongly indicated that the radicular pain associated with nerve distortions is somehowrelated to the resulting ischemia because the reduction in oxygen intake in patients withneurogenic claudication exacerbates their symptoms and shortens their “claudication

156 Parke

Fig. 5. Schema showing confined spatial relationships of epidural venous sinuses in stenoticsection of lumbar spinal canal. The left side indicates how the compromised space may restrictthe venous flow at that level. The right side shows how extraspinal venous hypertension mayengorge sinuses and provide additional pressure on dura and contained roots.

time” when walking (15). However, Watanabe and Parke (14) found that the arterialphase of the vasa radiculorum appears to be well compensated and maintains some con-tinuity despite a severe compression provided that the pressure had developed slowlyover several years. Further study then indicated that it is the venous side of the radicularcirculation that is more vulnerable (Fig. 7).

VASCULAR PATTERN OF INTRADURAL ROOTS

Contrary to some preexisting concepts, vascular and neuronal analyses now indicate thatthe intradural roots are part of the central nervous system (CNS), and that the relationshipof the arteries to the veins more resembles that found in the brain than in the peripheralnerves. This is shown by the fact that the radicular veins do not accompany the arterial

Role of Veins in Chronic Back Pain 157

Fig. 6. (A) Low-power (×20) transillumination photomicrograph of L4 nerve root showinginjected arteries and gas-filled veins demonstrating course of these intraradicular vessels and frequentarteriovenous anastomotic shunts. (B) Graphic compilation of structure and vasculature of typical lum-bosacral nerve root. (From ref. 5.) The numbers in both (A) and (B) are common to equivalent struc-tures: 1, fascicular pia; 2, inter- and intrafascicular arteries; 3, longitudinal radicular artery; 4, largeradicular vein (does not course with artery); 5, arteriovenous anastomosis; 6, collateral radicularartery; 7, gauzelike pia-arachnoid that permits free percolation of CSF to assist root metabolism.

pattern of distribution but are fewer and run a separate and usually deeper more centralcourse. Being thin walled, they are more susceptible to the spatial distortions imposed bydegenerative changes in the dimensions of the spinal canal and intervertebral foraminaand show a complete interruption in the chronically compressed nerve root.

Nerve root distortion is a consistent finding in neurogenic claudication and sciatica, butstudies of both nerve roots and peripheral nerve have shown that although compressionalone may disrupt most nerve functions, it usually does not cause pain (16–18). There is cur-rently a consensus that some degree of intrinsic irritation or inflammation must developbefore nerve distortion, by itself, may elicit pain. Studies of graded and intermittent pres-sures on nerve roots in the pig cauda equina have revealed that the resulting nutritional dis-ruption in the affected nerve, by both impedance of CSF percolation and interruption of thevascular channels, leads to visible edema and venous congestion (19). It is this irritated andinflammed segment that becomes hypersensitive and elicits pain with additional or pro-longed insults. In 1998, Takata et al. (20) noted that compressed roots were evident in CTmyelograms owing to their edematous enlargement. Subsequently, a detailed study byKobayashi et al. (21) showed that compression disrupts a type of blood-nerve barrier uniqueto the radicular microvessels and permits abnormal extravasations that produce edema.

Asingle-level stenosis may exist without marked claudication, which has led to the assumptionthat a double level of nerve root compression is more likely to be symptomatic (21,22). The pre-viously mentioned shunt function concept indicates that the radicular vascular pattern with itsfrequent and effective series of arteriovenous anastmoses allows for some circulatory compen-sation for single-point compressions (Fig. 8). However, a two-point compression would vascu-larly isolate the intermediate root segment and lead to changes consequent to obviousmetabolic deprivation (Fig. 9). A two-point compression relative to a spinal stenosis may onlyinvolve a single nerve root wherein the superior level of compression results from the stenoticconstriction (of it and other roots) and a herniated nucleus pulposis provides a second lowerforaminal compression with the development of the irritable edematous condition in the indi-vidual root between these points, or there can be two levels of stenotic constriction that wouldaffect all the roots between the levels. In the latter situation, the intervening isolated dural saccould provide conditions analogous to a compartment syndrome (23).

158 Parke

Fig. 7. Schema illustrating role of arteriovenous anastomoses in maintaining vascular functionin compressed root. Note that the anatomoses ensure the afferent-efferent cycle of the bloodflow up to both sides of the compression point.

ANIMAL EXPERIMENTAL STUDIES

The best demonstrations of the venous lability of the radiculomedullary circulationhave been provided by studies involving experimental compressions of the lumbosacralroots in the dog (22) and in the pig (16,19,24). In these studies, it was noted that anobvious venous congestion was a major consistent feature associated with the site of theexperimental compression. Both research groups used graduated stages of constriction,and the dog model was reported to show little alteration in neurologic function until a50% constriction had been attained. In the pig model, however, Olmarker and colleagues(16,19,24) recorded the graduations of pressure in millimeters of mercury injected into

Role of Veins in Chronic Back Pain 159

Fig. 8. Illustration depicting advantage of both a proximal and distal source to radicularcirculation that allow flow from either end to supply any root segment from above or below asingle point of pressure.

Fig. 9. (A) Illustration depicting problem of two-point pressures on long nerve root. The interven-ing longer segment between the compressions may become almost completely nutritionally isolatedand dependent only on the CSF metabolic exchange. (B) Illustration showing that vascular deprivationlikely leads to changes in vascular permeability with its consequent edema and root sheath fibrosis.

compressing balloons, and they remarked on how little pressure was required to inducea venous congestion that they believed could eventually lead to alterations in nervefunction. Both groups noted that the venous congestion soon induced an intraneurialedema. This radicular edema cannot be readily equated to the lymphedema found in theperipheral systems because the spinal roots, like the rest of the CNS, possess no lym-phatics, since they are able to discharge their extracellular proteins directly into theCSF. Olmarker and colleagues (16,19,24) noted that the edema formed more rapidlyfollowing rapid compression than in slow compression and concluded that it resultsfrom alterations in the permeability of the microvessels. This may cause the microves-sels to leak macromolecules into the endoneurial spaces and alters the nutritional accessto the axons, impairing function of the nerves (25).

SLOW VS RAPID COMPRESSION

Both the findings of Watanabe and Parke (14) and Delmarter et al. (22) showed amarked loss in the number of neuronal fibers in severely constricted roots. However,the rapidly induced experimental compressions evidently produced a more immediateand drastic neurological deficit than did the 5- to 10-yr periods required to producethe 75% constrictions noted in the Parke and Watanabe case. It was also recorded thatbeyond the 50% constriction in the roots of dogs, the arteries were interrupted,whereas in the observed 75% constriction of the gradual, long-term compression ofthe human roots, the arteries remained intact (14). Thus, there is a strong indicationthat very slowly developing degenerative stenotic compressions may allow sufficienttime for arterial and neuronal compensations and substitutions to preserve the moreessential functions and may thus mask the severity of the condition, despite the lossof a great amount of larger, well-myelinated fibers and the irreversible loss of muscu-lar strength. This was well illustrated by the fact that the patient in the Parke andWatanabe case, despite a neurogenic claudication, was able to walk into the hospitalwithout assistance a few days before his death from cardiovascular failure, and withonly about 25% of the original nerve fibers supplying his leg musculature.

VASCULARITY OF DORSAL ROOT GANGLION

Contemporary accounts of radiculomedullary circulation provide little or no mention ofthe vascular distribution to the DRG (26–28), despite the fact that Bergmann andAlexander (29) published a well-detailed treatment of the subject several decadesbefore most investigations of intrinsic radicular circulation were conducted. The lack ofa general awareness of this work may be attributed to the fact that its publicationpreceded the development of sufficient interest and background (an informationalmatrix that historians conveniently label with the German word Fragestellung) neces-sary to emphasize its clinical significance. Unfortunately, Bergmann and Alexander(29) relied only on pen-and-ink drawings to corroborate their verbal descriptions of themajor vessels, and a mental reconstruction from their injected microscopic series ofcross-sections to determine the finer intraganglionic vasculature.

More recent studies have concentrated on the functional aspects of DRG circula-tion with a greater regard for the measurable physiological reactions than for the mor-phological intricacies of the vascular pattern. This apparent neglect of the probable

160 Parke

roles of the vascular peculiarites of the DRG may originate partly in the recognitionthat much of the vessel architectonics of the entire human radiculomedullary circula-tion is unique to the large primate species, and that the functional generalizationsderived from various experimental animals are demonstrably more reliable than themorphological ones (30). Recently, Parke and Whalen (31) were able to provide tran-silluminated photomicrographs of the finely injected dorsal root ganglia of humanperinatal cadavers and venous injections of adult human specimens. From these itwas determined that the vessel distributions from a few lower-level ganglia describedby Bergmann and Alexander (29) could be generalized to derive a pattern of distribu-tion that is characteristic of the DRG in all levels of the various vertebral regions, andthe arterial ramifications could be summarized in a single description (Fig. 10A). Inessence, it was shown that the nuclear parenchyma is highly vascularized primarilyfrom vessels entering the proximal and distal poles of each ganglion. These, throughpredominantly longitudinal derivatives that course parallel to the long axis of the gan-glion, supply the cordlike arrangements of the neuronal nuclei with a dense capillarybed. A secondary, finer network of arteries that is also mostly derived from the proxi-mal and distal polar sources forms a fine reticular system over the surface of the gan-glion. This periganglionic plexus communicates with the deeper vessels by a networkof centripetally coursing fine channels. The distal and proximal polar arteries arederived from the epidural branches of the intersegmental vertebral arteries, as indi-cated in the generalized illustration of the “generic” DRG arterial vasculature shownin Fig. 10A.

Injections of the cadaver segmental veins was a limited success, because the numer-ous open channels of the dissected segments failed to provide the closed system neces-sary to develop adequate pressures. Nevertheless, the results indicated the existence ofa consistent periganglionic venous plexus that receives the major efferent blood flowfrom the interior of the DRG parenchyma. This supports the observations thatBergmann and Alexander (29) derived from their histological sections.

In 1977, Howe et al. (32) published a physiologically based study showing that,unlike the lack of ectopic impulses generated by the compression of normal nerveroots, prolonged periods of repetitive nerve firings followed a brief acute compressionof the normal DRG. In an excellent article, Yoshizawa et al. (33) provided evidencethat the blood flow volume in the dorsal root of dogs is less than that of the gray matterof the cord and the peripheral nerves, whereas the DRG blood flow volume is approxi-mately twice that of the nerve root and similar to the blood flow in the gray matter ofthe cord. By using the hydrogen clearance method, they were able to show that theDRG blood flow was reduced 40–45% by a compression of 60 g on the distal side ofthe ganglion, and 10–15% by equivalent pressure on the proximal side.

Rydevik et al. (23) recorded the endoneurial pressures within the normal rat dorsalnerve root and its DRG and observed it to be higher than the internal fluid pressurewithin the animal’s sciatic nerve, with a reading of 3.7 cm of water. They then notedthat it rose to as high as 9.6 cm of water subsequent to a mechanical deformation.Because miniature compartment syndromes have been shown to exist after compressionof a peripheral nerve (34), it was concluded that such a high-pressure elevation in theDRG would likely affect the nutrition of the neuron cells, and a long-standing edemacould result in DRG endoneurial fibrosis. Although Rydevik et al. (23) recognized that

Role of Veins in Chronic Back Pain 161

162 Parke

Fig. 10. (A) Labeled schema of “generic” DRG with its regional vascular and dural rela-tions: 1, segmental artery; 2, lumbar, intercostal, or cervical artery; 3, spinal nerve; 4, radicu-lomedullary branch of segmental artery; 5, dorsal branch of segmental artery; 6, internal arteryto lamina; 7, paravertebral venous plexus; 8, distal polar DRG arteries; 9, proximal polar DRGarteries; 10, periganglionic arterial plexus; 11, meningeal segmental artery; 12, medullary arteryto dorsolateral spinal artery; 13, dorsal nerve root; 14, ventral nerve root; 15, smaller inconsis-tent meningeal intersegmental artery; 16, spinal dura; 17, epidural venous plexus; 18, dorsalartery to vertebral bodies and their ligaments. (B) This schema duplicates the anatomical rela-tions labeled in (A), but the periganglionic venous plexus has been reconstructed to emphasizeits vulnerable position adjacent to the fibrous dura where external pressures (large arrow) or ede-matous internal pressures (small arrows) could readily create a venostasis leading to a DRGcompartment syndrome. (Adapted from ref. 31.)

the increase in pressure within the closed fibrous dural compartment of the DRG wouldimpede normal vascular flow, they did not relate it to any predisposing characteristic ofthe vascular pattern.

The demonstrable fact that when a mechanical pressure is applied to an endoneurialand/or perineurial vascular system a venostasis is achieved at a significantly lowerpressure than that required to stop the arterial flow indicates that this greater lability of

the efferent side of the system may be more responsible for ischemic neuropathies thanwas previously supposed.

Regarding the intrinsic vascularity of the human DRG as described and illustratedhere, it is obvious that the structural arrangement of the vessels presents an inherentvulnerability to compressive factors. Weinstein (35) made a functional allusion thatthe DRG could be regarded as the “brain” of the motion segment unit. Thismetaphor may be taken one step further with respect to the DRG’s circulation. Likethe cerebral vascularity, the main arterial supply of the DRG cellular masses tends toimmediately run deep and central to reach their functional position in theparenchyma. The efferent veins, as in the cerebral cortical tissues, generally run aseparate centrifugal course and rise to the surface to collect in a periganglionicplexus that is distributed countercurrent to the much finer, but equivalently located,periganglionic arterial plexus.

An analysis of the unique structural arrangement of the vascularity of the humanDRG in relation to its relatively unyielding fibrous adnexa, as graphically rendered inFig. 10B, shows that by the structure and position of the major venous drainage, thehuman DRG is remarkably predisposed to the development of a classic compartmentsyndrome. The total effect of an external mechanical compression may be complex. Inan acute phase, the compression would force the dural capsule to impede directly theimmediately underlying efferent flow in the periganglionic plexus and thus resist thenormal afferent nutrition to the deeper parenchymal constituents. By subsequent alter-ations in the vascular permeability, a resulting endoneurial edema would reflexly pushthe periganglionic venous plexus against its unyielding fibrous container and create amore chronic venostasis (36). A prolonged maintenance of this condition could lead toendoneurial fibrosis, increased sensitivity, and repetitive ectopic firing of the containedneural elements.

CONCLUSION

The information discussed herein should be instructive when considering surgicalapproaches to the spine. The judicious necessity of a surgical intervention automaticallyadmits that the involved segmental region has been subject to a spatial encroachmentthat may have already affected the efferent vascular blood flow to the foraminal venouschannels. Therefore, the extensive use of electrocautery just to achieve a clearer fieldshould obviously be avoided. The operative field reductions inherent in the more minimalapproaches, especially those using percutaneous instrumentation, may provide somemarked advantages. The most prominent of these would be a mitigation of both theextent of required venostasis and the undesirable consequences resulting from the inad-vertent manipulation of the DRG.

REFERENCES

1. Brechet G. Essai sur les Veines der Rachis, Mequigon-Morvith, Paris, 1819.2. Clemens HJ. Die Venesysteme der menschlichen Wirbelsaule, Walter de Gruyter, Berlin, 1961.3. Batson OV. The function of the vertebral veins and their role in the spread of metastasis.

Am Surg 1940;112:138–145.4. Parke WW, Gammell K, Rothman RH. Arterial vascularization of the cauda equina. J Bone

Joint Surg 1981;63(A):53–62.

Role of Veins in Chronic Back Pain 163

5. Parke WW, Watanabe R. The intrinsic vasculature of the lumbosacral spinal nerve roots.Spine 1985;10:508–515.

6. Magnaes B. Clinical recording of pressure on the spinal cord and cauda equina: parts 1 and 2.J Neurosurg 1982;57:48–63.

7. Rydevik B, Holm S, Brown MD, et al. Diffusion from the cerebrospinal fluid as a nutritionalpathway for spinal nerve roots. Acta Physiol Scand 1990;138:247, 248.

8. LaBan MM. “Vespers curse” night pain—the bane of hypnos. Arch Phys Med Rehabil1984;65:501–504.

9. LaBan MM, Wesolowski DF. Night pain associated with diminished cardiopulmonarycompliance. Am J Phys Med Rehabil 1988;67:155–160.

10. Kaiser MC, Capesius P, Roilgen A, et al. Epidural venous stasis. Neuroradiology 1984;26:435–438.

11. Madsen JR, Heros RC. Spinal arteriovenous malformations and neurogenic claudication. JNeurosurg 1988;57:793–797.

12. Parke WW. The significance of venous return impairment in ischemic radiculopathy andmyelopathy. Orthop Clin North Am 1991;22:213–221.

13. Aboulker J, Bar D, Marsault C, et al. L’hypertension vieneuse intra-rachidienne paranomalies multiples du system cave: une cause majeure de souffrance medullaire. Chirurg-erie 1977;103:1004–1015.

14. Watanabe R, Parke WW. Vascular and neural pathology of lumbosacral spinal nerve rootsin spinal stenosis. J Neurosurg 1986;64:64–70.

15. Evans JG. Neurogenic intermittant claudication. BMJ 1964;2:985–987.16. Olmarker K, Rydevik B, Holm S. Edema formation in spinal nerve roots induced by exper-

imental graded compression. Spine 1989;14:569–573.17. Pedowitz RA, Rydevik BL, Hargrens AR, et al. Neurophysiologic and histologic changes

induced by acute graded compression of the pig cauda equina. Paper presented at the Inter-national Society for the Study of the Lumbar Spine, Miami, FL, 1998.

18. Rydevik B, Hannson TH, Garfin SR. Pathophysiology of cauda equina compression. SeminSpine Surg 1989;1:139–142.

19. Olmarker K, Rydevik B, Holm S, Bagge U. Effects of experimental graded compressionon blood flow in spinal nerve roots: a vital microscopic study on the porcine cauda equina.J Orthop Res 1989;7:817–823.

20. Takata K, Inuoe S, Takahashi K, Ohtsuka Y. Swelling of cauda equina in patients who haveherniation of a lumbar disc. J Bone Joint Surg 1998;70A:361–368.

21. Kobayashi S, Yoshizawa H, Hichiya Y, et al. Vasogenic edema induced by compressioninjury to the spinal nerve root. Spine 1993;18:1410–1424.

22. Delmarter LB, Bohlman HH, Dodge LD, Biro C. Experimental lumbar spinal stenosis. JBone Joint Surg (Am) 1990;72:110–120.

23. Rydevik BL, Myers RR, Powell HC. Pressure increase in the dorsal root ganglion follow-ing mechanical compression: closed compartment syndrome in nerve roots. Spine1989;14:574–576.

24. Olmarker K, Rydevik B, Hansson T, et al. Compression induced changes in the nutritionalsupply of the porcine cauda equina. J Spinal Disord 1990;3:25–29.

25. Myers RR, Murakami H, Powell HC. Reduced nerve blood flow in edematous neu-ropathies: a biomechanical mechanism. Microvasc Res 1986;32:145–151.

26. Corbin JL. Anatomie et Pathologie Arterielles de la Moelle, Masson et Cie, Paris, 1961.27. Crock HV, Yoshizawa H. The Blood Supply of the Vertebral Column and Spinal Cord in

Man, Springer-Verlag, New York, 1977.28. Lazorthes G, Gouaze A, Zadwh JO, et al. Arterial vascularization of the spinal cord. J

Neurosurg 1971;35:253–262.29. Bergmann L, Alexander L. Vascular supply of the spinal ganglion. Arch Neurol Psychiatry

1941;46:761–782.

164 Parke

30. Parke WW. Point of view (a cautionary discussion on the necessity, advantage and pitfallsof animal models in research). Spine 1995;20:765.

31. Parke WW, Whalen JL. The vascular pattern of the dorsal root ganglion and its probablebearing on a compartment syndrome. Spine 2002;27:347–352.

32. Howe JF, Loeser JD, Calvin WH. Mechanosensitivity of dorsal root ganglia and chronicallyinjured axons: a physiological basis for radicular pain of nerve root compression. Pain1977;3:25–41.

33. Yoshizawa H, Kobayashi S, Hachia Y. Blood supply of nerve roots and dorsal root ganglia.Orthop Clin North Am 1991;22:195–211.

34. Lundborg GI, Myers R, Powell H. Nerve compression injury and increased endoneurialpressure: a miniature compartment syndrome. J Neurol Neurosurg Psychiatry 1983;46:1119–1124.

35. Weinstein JN. Part B: Basic Science Perspectives. In: New Perspectives in Low Back Pain(Frymoyer JW, Gordon SL, eds.), American Academy of Orthopedic Surgeons, Park Ridge,IL;1989 pp. 57–110.

36. Hoyland JA, Freemont AJ, Jayson MIV. Intervertebral foramen venous obstruction: a causeof periradicular fibrosis. Spine 1989;14:538–568.

Role of Veins in Chronic Back Pain 165

8Diagnostic and Therapeutic Percutaneous

Transpedicular Approaches to the Spine

Alexander G. Hadjipavlou, MD, George M. Kontakis, MD,Ioannis Gaitanis, MD, and Michael Tzermiadianos, MD

TRANSPEDICULAR BIOPSY

Historical Review and Rationale for the Procedure

Open biopsy advocates prefer transpedicular biopsy (TPB) because it maximizes tissueretrieval, thus providing the highest diagnostic success rate. Open biopsy is especiallyrelied on after failed needle biopsy or in selected presumed primary bone or cartilaginoustumors (1). However, the complications and morbidity associated with an open surgicalprocedure provided incentive for the development of closed needle biopsy techniques.

Historically, preference for closed biopsy of the spine developed because it wasclaimed to be less invasive, less morbid, and more cost-effective than open biopsy.Closed biopsy has also become increasingly accurate as techniques and image modali-ties have evolved. Local anesthesia and an outpatient setting contribute to enhancedcost-effectiveness. Local anesthesia also allows nerve root monitoring during biopsy.Consequently, percutaneous biopsy of spinal lesions has become the biopsy techniqueof choice, but not without potential complications, such as nerve injury, bleeding, pneu-mothorax, and inadequate amount of tissue retrieval for diagnosis (2–5).

The reported diagnostic success rates of closed needle biopsy of the spine are variableand decrease significantly with primary bone tumors (2,3,6,7) and tumors with complexarchitecture and cell pleomorphism (such as giant cell tumors, aneurysmal bone cyst,osteoblastoma, osteosarcoma, or chondrosarcoma) (8,9). Crush artifacts, one of the prob-lems created by small needles (3), predisposes conventional closed biopsy to an inferiorsuccess rate (6,10). Fyfe et al. (10) reported a cadaveric study in which biopsy specimenswith tissue core diameters ≥2 mm enhanced diagnostic accuracy. Because the pedicleaccommodates biopsy instruments that retrieve tissue core diameters >2 mm, the diagnos-tic success rate of a percutaneous TPB should approach the success rate of an open proce-dure. Larger tissue core diameters also avoid the diagnostic problems created by crushartifacts. Therefore, there was room for improvement and a transpedicular approach wasdeveloped as an alternative to the other biopsy methods for vertebral lesions involving thesacrum and thoracic, lumbar, and seventh cervical vertebral (11,12).

Enthusiasm regarding surgery involving the vertebral pedicle is reflected by the ever-increasing information regarding transpedicular fixation (13), morphology (14–21),biomechanics (22), fracture management (13,23), and hemiepiphysiodesis (24).

167

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

Transpedicular fixation techniques have continued to increase in popularity since theirinception (25,26). The pedicular channel also has been used for fracture reduction(27,28), external skeletal fixation (29), decompression (30), thoracic discectomy (31),bone grafting (26), and methylmethacrylate insertion (26).

Despite increasing knowledge of vertebral morphometry and experience withtranspedicular fixation, it was a long time before the pedicle was popularized as a chan-nel for percutaneous vertebral needle biopsy.

The use of the pedicular channel for open biopsy is not a new idea. In 1928, vonLackum (as reported by Duncan and Ferguson (32) in 1936) performed a transpedicularcurettage of a vertebral body giant cell tumor in an 8-yr-old girl. In 1933, Capener (33)described an anterolateral decompression in which the pedicle was removed to accesslesions in the vertebral body. In 1949, Michele and Krueger (34) described a transpedic-ular approach as one of four posterior approaches to the vertebral body. It was not until1979 that Travaglini (35) reintroduced this technique in the English literature.

The belated development of this technique may be attributed to three explanations.First, the proximity of the pedicle to neural elements deterred closed biopsy attemptsbecause of fears of injuring these vital structures. Second, appreciation of the biopsypotential of vertebral body lesions through the pedicle has been limited (36). Third, thelarger tissue samples retrievable with open biopsy made open procedure (with radio-graphic guidance when indicated) the “gold standard” to which all other biopsy proce-dures had to be compared.

In 1983, Roy-Camille et al. (13) first described an open (TPB) technique used in a seriesof 47 patients. In 1990, Rengachary described a transpedicular technique that included ahemilaminectomy, a partial facetectomy, and a partial pediculectomy (19). Also in 1990,Fidler and Niers (37) reported one case of an open TPB. In 1991, Renfrew (7) reported per-cutaneous TPBs in six patients using computed tomography (CT). We have reported thetechnique of TPB as an efficacious, safe, and cost-effective method (12,38–43). In mostcases, it can be performed under local anesthesia, with fluoroscopic guidance.

The Percutaneous Transpedicular Biopsy Technique

The percutaneous procedure requires a high-resolution image intensifier and a radi-olucent operating table that can be precisely tilted. The transverse pedicle width and thepedicle angle in the axial plane are determined from preoperative CT images. The oper-ating table is canted until the pedicular angle in the axial plane is perpendicular to thefloor and the X-ray beam is collinear with the sagittal pedicular angle determined fromlateral views of the vertebral body. A “bull’s-eye” view of the pedicle should beobtained. This procedure is analogous to obtaining perfect circles during distal inter-locking procedures of intramedullary femoral nail. Local anesthesia is obtained byinjecting plain 1% lidocaine hydrochloride along the intended biopsy tract and infiltrat-ing the posterior primary ramus as it emerges from the junction of the transverse processand superior facet of the corresponding joint and adjacent superior and inferior facetjoints. After insertion of the guide pin, the physician makes a small stab wound incisionabout 1 cm long to allow the passage of a modified Kambin dilator (44) (5.35-mm diam-eter; Smith & Nephew) over the guide pin until it reaches bone (Fig. 1). Following this,a cannulated modified Kambin sleeve (6.4-mm diameter; Smith & Nephew) is passedover the dilator and guide pin until it abuts the cortical margins of the pedicle (Fig. 2).The use of a cannulated sleeve prevents clogging of the bone biopsy instrument with

168 Hadjipavlou et al.

Percutaneous Transpedicular Approaches 169

Fig. 1. (A) Modified Kambin-Craig instrumentation (manufactured by Smith & Nephew).Under image intensification, a guide pin is inserted (B) by tapping it gently (C); (D) “bull’s-eye”view into the pedicle.

subcutaneous tissue or muscle fibers and also facilitates the insertion of the instrumentfor discectomy. Next, the physician removes the dilator and advances a toothed, modi-fied Craig biopsy tool (3.2- or 5.15-mm diameter; Smith & Nephew) over the guide pininto the target. This tool has a larger diameter than the conventional Craig needle biopsyand a knob to attach a torque device that will facilitate manual introduction of thebiopsy tool. The larger lumen allows passage of various instruments through the biopsytool. It is important that the surgeon remove simultaneously the Steinmann pin and thebiopsy tool. This method allows the successful removal of a core of bone or pathologi-cal tissue, because the specimen is impacted between the guide pin and the bone biopsyinstrument (12,40) (Fig. 3).

We have demonstrated in the laboratory and in the clinical setting that retrieval ofosteopenic bone and pathological soft tissue is enhanced as tissue is impacted betweenthe biopsy cutting core tool and the guide pin. This expedience holds securely the biopsyspecimen within the cutting core tool. Sufficient space also exists for insertion of instru-ments at various angles and directions to increase tissue sampling and access any vertebral

170 Hadjipavlou et al.

Fig. 2. (A,B) The pin is angled to the lesion intended for biopsy. Next a dilator (C) is passedover the guide pin to dissect the soft tissues, and a cannulated sleeve is inserted over the dilatoruntil it reaches the pedicle. The dilator is then removed, and the toothed cutting biopsy tool (D) isinserted into the sleeve over the guide pin. (Partially reproduced with permission from ref. 40.)

body lesion. The integrity of the inferior and the medial cortical walls of the pediclemust be preserved in order to prevent any spread of hematoma, infection, or tumorinside the spinal canal. Additional tissue can be retrieved using curettes or biopsyforceps through the cannulated sleeve after removal of the guide pin (Fig. 4). Thecannulated sleeve also facilitates insertion of hemostatic agents such as Surgicel (John-son & Johnson Medical) or methylmethacrylate bone cement. The use of bone wax forhemostasis is not recommended because it does not pack well within the pedicle via thecannulated sleeve. Drains for 24 h are used only in cases of infection or benign condi-tions. If a drain is inserted, the patient must return on the first postoperative day forremoval of the drain. We do not advocate drainage in the presence of malignancy.

Percutaneous Transpedicular Approaches 171

Fig. 3. Lateral radiograph demonstrating toothed biopsy cutting tool as it is inserted into ver-tebral body over guide pin (A) using a T-handle torque device (B). As the cutting biopsy tool isbeing inserted, tissue is impacted between the guide pin and the biopsy tool and held firmlyinside the tool. This expedience facilitates retrieval of tissue (C,D). (Partially reproduced withpermission from ref. 40.)

Discussion

As graphed by Misenhimer et al. (17), average cancellous pedicle width (transverseinside diameter) from T1 to L5, measured by sounding, ranges from slightly more than1 mm at T4 to slightly less than 6 mm at L5. Because a biopsy needle that will retrieve atissue core diameter larger than 2 mm has an outside diameter of nearly 3 mm, adequatespace exists in most pedicles for transpedicular retrieval of substantial tissue specimen.

A transverse inner pedicle diameter that measures <3 mm is not a contraindicationfor percutaneous TPB. According to Zindrick et al. (21), the average transverse outside

172 Hadjipavlou et al.

Fig. 4. The biopsy tool can be repositioned in different directions. (A) Further biopsy specimenscan be removed by means of (B) curettage or (C,D) biopsy forceps.

diameters of the pedicular isthmus in the fifth thoracic vertebra is 4.5 mm and in thefifth lumbar vertebra is 18 mm. The narrowest pedicle diameter is 5 mm at T5 thoraciclevel, and the inside pedicle diameter measures <3 mm (45). Band-saw cuts throughthe frontal plane of the vertebral pedicle demonstrated that this is neither circularnor elliptic but egg shaped, with the narrow end superior and the wider end inferior.Furthermore, we have confirmed that the pedicle is mostly cancellous bone with athin shell of cortical bone (12). Finally, the nerve root courses medial to the medialwall of the pedicle and inferior to the inferior wall of the pedicle, whereas the duralsacs lie immediately adjacent to the medial wall of the pedicle. Percutaneous TPB cansafely be performed by cutting through the lateral wall extrapedicularly and avoiding

Percutaneous Transpedicular Approaches 173

Fig. 4. (Continued)

violation of the medial pedicular wall. Inserting bone biopsy instruments through thisarea is minimally problematic. Caution should be taken not to violate the foramen,which provides nutrient vessels to vital nerve tissue structures.

Not only will the pedicle accommodate a variety of biopsy instruments, but the pediclealso will provide access to any vertebral body lesion. In our laboratory study, we haveshown that instruments passed through one vertebral pedicle can access more than 50% ofthe volume of the vertebral body, including tissue directly anterior to the spinal canal(Fig. 5). Furthermore, this volume is accessible without performing a laminectomy,facetectomy, or pediculectomy, as described by others (46). Additional tissue can beobtained by performing multiple passes at various angles. Greater latitude for anglinginstruments exists in the sagittal plane than in the axial plane, because sagittal pediclediameter is greater than transverse diameter. The volume of tissue retrievable throughthe pedicle supports use of the percutaneous transpedicular technique for routine biopsyof vertebral body lesions. In cadaveric specimens, an experimental study showed that a2-mm trephine does not obtain suitable bone core for histological examination, whereasthe amount of samples obtained with a 3.5-mm trephine is adequate for histopathologicalexamination (47).

Fidler and Niers (37) recommended an open transpedicular approach over a percuta-neous procedure. They claim that the open approach facilitates block excision of tissueand prevents dissection of hematoma and damage to the pedicular wall. Violation of thepedicular wall may potentially contaminate the epidural space or the paravertebralstructures. However, using the percutaneous technique as we have described, thesepotential complications can be avoided and the patient can be spared the morbidity andcost associated with an open surgical procedure (12).

174 Hadjipavlou et al.

Fig. 5. In the laboratory, we have shown that through pedicular channels, bone can be retrievedfrom any region of the vertebral body. (Partially reproduced with permission from ref. 12.)

Percutaneous Transpedicular Approaches 175

Renfrew et al. (48) recommended CT-guided percutaneous TPB of the spine. This wasbased on the fact that the proximity of neural elements to the pedicle makes transpedicularbiopsy under fluoroscopy a hazardous procedure. However, high-resolution image inten-sifiers display sufficient details of vertebral elements so as to allow protection of themedial and inferior walls of the pedicle during biopsy, thus avoiding injury to the neuralelements. In our series, there were no advantages of CT guidance over image intensifica-tion (12). Cost-effectiveness is an advantage of image intensification over CT. Moreover,in the presence of spinal deformities, image intensification is easier to maneuver.

Negative results can be encountered as a consequence of technical errors. We believethat pitfalls owing to faulty biopsy instrumentation retrieval techniques can be avoided.

Failures can be encountered when the guide pin technique is not used while retrievingthe biopsy tool (Fig. 6). We encountered no diagnostic problems with crush artifactdespite crowding the biopsy tool with a guide pin. Impaction of tissue between the nee-dle and guide pin facilitates tissue retrieval in osteopenic bone and friable soft-tissuelesions. Another pitfall can be encountered when the pedicle is sclerotic and the lesion inthe vertebral body is lytic. In this situation, dense bone from the pedicle is packed intothe biopsy cannulated instrument and clogs the cutting tool, which makes almost impossi-ble any further retrieval of pathological soft tissue from the vertebral body. This problemprompted us to modify the technique by removing vertebral tissue in sequence. The sur-geon first creates an empty tunnel in the pedicle by removing a core of bone. Then thesurgeon reinserts the empty biopsy tool through the empty pedicle into the pathologicalfriable tissue, and, thus, the tool can retrieve a specimen for biopsy unimpeded (Fig. 7).

The reported complications of this procedure were minor and the incidence rangedfrom 0 to 5.6% (40,42,49–53). In our series (40), we had one technical complication—a retained piece of drainage tube in the pedicle—which was easily retrieved via thepercutaneous transpedicular tract, previously created, using a biopsy forceps underlocal anesthesia. Serious bleeding, which can be encountered in hypervascular tumors,is easily manageable by plugging the pedicle with either methylmethacrylate bonecement or Surgicel (40). To avoid spillage of malignant tumor tissues into the sur-rounding area, we also advocate the use of methylmethacrylate cement to plug thepedicular entrance (Fig. 8). In cases of infection, drainage for suction irrigation can beleft in situ. The reported diagnostic accuracy of PTB ranges from 89 to 99%(40,49,51–55). In our series of 86 procedures, the diagnostic accuracy was 95%. Alldiagnostic failures (four cases) occurred in the first 54 patients of our series (40). Inthe subsequent patients, our success rate was 100% (42). When technical pitfalls areavoided, the diagnostic success rate of TPB is equivalent to that of open biopsy tech-niques and with significantly less morbidity (Figs. 9–18).

In conclusion, we recommend the percutaneous TPB technique over open biopsy orclosed posterolateral biopsy for its safety, minimal morbidity, simplicity, diagnosticaccuracy, and cost-effectiveness. The caliber of the pedicle accommodates biopsyinstruments that are able to access any vertebral body lesion and retrieve sufficienttissue for diagnosis. In addition, the use of local anesthesia provides a reliable monitorof nerve root function. Bleeding is also easily controlled. Furthermore, the techniquecan extend to the upper thoracic levels including the C7 vertebra, provided a high-resolution image intensifier is available.

176 Hadjipavlou et al.

Fig. 6. (A,B) This drawing demonstrates that removal of biopsy specimens through the rightpedicle is greatly facilitated by removing the guide pin of the biopsy cutting tool and the guidepin simultaneously. (B, right) Tissue is packed between the biopsy cutting tool and the guide pin.Using this technique, we have never failed to retrieve vertebral tissue, neither in the laboratorynor in the clinical setting (C). However, if the guide pin technique is not used, the core, cut bythe biopsy tool, might not remain inside the biopsy instrument (especially if the tissue isosteopenic or friable) when the instrumentation is removed (see left pedicle). (Partially repro-duced with permission from ref. 40.)

Percutaneous Transpedicular Approaches 177

Fig. 7. (A) Bone from the pedicle can clog the tip of the biopsy cutting tool and, thus, maynot allow friable tissue from a lytic lesion (b) to enter the biopsy tool. (B) Further insertion ofthe biopsy cutting tool may even crush a soft-tissue lesion against hard bone. (C) First a coreof bone is removed from the pedicle. (D) Then the empty biopsy cutting tool should be rein-serted through the open pedicular channel, to retrieve soft tissue unimpeded (E). (F) Furtherspecimens of friable soft tissue can be removed by mean of biceps forceps. (Modified withpermission from ref. 40.)

178

Fig

. 8.

(A)

A-w

eigh

ted

imag

e sh

ows

a lo

w s

igna

l in

tens

ity, a

nd (

T2-

wei

ghte

d im

age

show

s a

high

sig

nal

inte

nsity

a m

etas

tatic

mix

edle

sion

seen

scan

(la

tera

l ref

orm

at).

,E) A

fter

tran

sped

icul

ar b

iops

y, s

ome

blee

ding

was

enc

ount

ered

, and

this

was

con

trol

led

by in

sert

ing

PMM

Abo

ne c

emen

t int

o th

e ly

tic c

ompo

nent

the

tum

or a

nd th

e pe

dicl

e. P

osto

pera

tivel

y th

e pa

tient

was

pai

n fr

ee.

179

180 Hadjipavlou et al.

MANAGEMENT OF PYOGENIC SPONDYLODISCITIS

Historical Review and Rationale for the Procedure

Because MRI has shown that the pathological lesion involves the disc and the twoadjacent vertebral bodies (56), the term spondylodiscitis is preferred. The natural historyof uncomplicated spondylodiscitis is self-limiting healing. However, a variabledegree of bone destruction frequently takes place during the infectious process (57).Depending on the degree of bone destruction, it is not uncommon for the spine to heal

Fig. 9. An axial CT scan of the T12 vertebra shows that it is affected with (A) solitarymyeloma and (B) its histology. A lateral radiograph demonstrates (C) pathological fracture of L3vertebra and (D) biopsy-revealed lymphoma.

in a kyphotic deformity, which, in turn, may predispose to mechanical low-back pain(58). Reports have indicated that mechanical low-back pain is frequently associatedwith conservative treatment of vertebral osteomyelitis (59). Early diagnosis is crucialfor management of this condition (60–62), because delayed treatment also may resultin serious neurological complications (63).

Percutaneous Transpedicular Approaches 181

Fig. 9. (Continued)

182 Hadjipavlou et al.

Fig. 10. (A) T1 A-weighted magnetic resonance imaging (MRI) image of a lytic lesion isshown. (B) TPB revealed renal cell carcinoma. (C) An axial T1-weighted MRI image of a blas-tic lesion is shown. (D) An axial CT scan shows the removed biopsy core. (E) Histologicalexamination revealed osteoblastoma.

Percutaneous Transpedicular Approaches 183

The treatment of joint infections typically includes surgical debridement, irrigation,and prolonged antibiotic therapy (64–69). Gradually, the percutaneous arthroscopicapproach has superseded open arthrotomy (70,71). A similar concept has been applied suc-cessfully to the treatment of pyogenic spondylodiscitis. Percutaneous discectomy, bymeans of a nucleotome, can evacuate infected disc material as an alternative to opensurgery (72–74). However, reports are scanty and only two or three patients are referred toin each report.

Fraser et al. (75) showed experimentally that during the natural course of discitis,granulation tissue from the subchondral bone would invade the intervertebral disc,resorb the disc space, and heal the infection. Intradiscal invasion of vascular granulationtissue was present in our histopathological studies (76). Successful treatment of discitisentails spontaneous fusion. However, the spine very often may either fail to fuse, devel-oping pseudoarthrosis, or fail to heal in good alignment, resulting in kyphotic deformity.Both conditions may predispose to chronic low-back pain. Spontaneous interbodyfibrous or bony fusion occurs in 6–24 mo (77,78). However, according to Fredericksonet al. (79), spontaneous ankylosis occurs in only 35% of patients. Therefore, it seemsreasonable to assume that any medical manipulation that accelerates the natural healingprocess may prevent these complications (38,41). Although the published data are notfrom prospective randomized studies, there is good evidence in the studies to supportthis concept. Transpedicular drainage of Pot’s abscess, as an adjunct to posterior stabi-lization, was performed successfully to speed up the process of healing (80).

Fig. 10. (Continued)

184 Hadjipavlou et al.

Fig. 11. Sagittal T1-weighted MRI image of a lytic lesion (A) better demonstrated on lateralreformated CT scan. (B) An adequate amount of tissue was retrieved to allow differenthistopathological staining techniques in order to enhance the diagnostic accuracy. The diagnosiswas chordoma. (C) Typical physalipherous cells; (D) cluster epithelioid cells; (E) S1 100 pro-tein stain; (F) Vimentin stain.

Percutaneous Transpedicular Approaches 185

Fig. 11. (Continued)

186

Fig

. 12.

Hyd

atid

dis

ease

seen

(A)

axia

l C

Tsc

an a

nd (

its h

isto

logy

. T1-

and

T2-

wei

ghte

d M

imag

es d

emon

stra

ting

spon

dylo

disc

itis

(C,D

)ca

used

blas

tom

ycos

is a

s se

en o

n hi

stol

ogy

(E).

187

The objective of transpedicular discectomy is to accelerate the natural course of heal-ing by evacuating the bulk of the offending infected disc and, conceivably, by openingchannels through the subchondral bone to speed the process of disc invasion by the repar-ative granulation tissue. For these reasons, and because we had considerable experience inusing the technique of the transpedicular route for vertebral biopsies, we decided todesign a transpedicular approach for discectomy in pyogenic spondylodiscitis (41–43,81).

The Percutaneous Transpedicular Discectomy Technique

Local or general anesthesia is suitable for percutaneous transpedicular discectomy,depending on the severity of pain. The patient is prone, either on a fluoroscopic table inthe radiology suite or on an operating table in the surgical suite, as for a TPB procedure.The target for the pin is the pedicle that is caudal to the affected disc. The tip of theguide pin should be in the center of the pedicle bull’s-eye on fluoroscopic view.

Using an image intensifier, the technician obtains a lateral view to determine cepha-lad angulation of the Steinmann pin in the sagittal plane; this approach is necessary for

188 Hadjipavlou et al.

Fig. 13. (A) T2-weighted MRI image and (B) axial CT scan showing an osteolytic lesion ofa thoracic vertebral body. (C) TPB revealed coccidiomycosis.

reaching the center of the affected disc without violating the confinements of the pedi-cle. The physician then holds the Steinmann pin firmly in this position and gently taps itwith a mallet until its tip reaches the inner annulus along the posterior portion of the disc.Under no circ*mstances should the pin violate the inferior border of the pedicle,because the pin can damage the exiting nerve root. Avoiding an approach through themore cephalad pedicle prevents this danger. Image intensifier views in the oblique and

Percutaneous Transpedicular Approaches 189

Fig. 14. (A) Sagittal T1-weighted MRI image showing a metastatic lesion. Needle biopsyfailed. TPB bull’s-eye (B) through the osteoblastic pedicle of the C7 vertebra (C) revealed anosteoblastic reactive bone with nidus of malignancy (a metastatic lesion from cancer of the breast[D]). (E) A CAT scan demonstrates the biopsy track. Usually needle biopsy fails in osteoblasticlesions. (F) An axial CT of a chondral lesion is shown. (G) TPB revealed chondrosarcoma.

190 Hadjipavlou et al.

Fig. 14. (Continued)

lateral planes may be used to assess the progress of the pin and thus ensure the integrityof the pedicle and the track of the guide pin.

This procedure has three phases. The first phase is similar to the TBP approach. Inthe second phase, discectomy is performed by means of tissue forceps. A modifiedKambin discectomy forceps (Smith & Nephew), which is inserted through the can-nulated sleeve, allows extraction of additional tissue from the disc. These tissuesamples are sent for pathohistological and bacteriological studies. Repositioning ofthe Steinmann pin through the pedicular tract allows direction of the biopsy instru-ment to a different part of the disc. By moving the biopsy forceps into these differentpositions, an adequate discectomy can take place in a piecemeal fashion (Fig. 19). Theset is equipped with one straight and two different angled Kambin flexible discectomyforceps.

The third phase of the procedure involves suction aspiration through the use of aflexible automated nucleotome (Surgical Dynamics, Alameda, CA) (Fig. 20). The flexi-ble automated nucleotome enters through the skin sleeve and the pedicular channel into

Percutaneous Transpedicular Approaches 191

Fig. 14. (Continued)

192 Hadjipavlou et al.

Fig. 15. (A) A lateral radiograph of an L5 vertebral lesion is shown. (B) TPB revealed Pagetdisease of bone. (C) An axial CT scan image of an osteolytic lesion is shown. (D) TPB revealeda giant cell tumor.

Percutaneous Transpedicular Approaches 193

Fig. 16. (A) Axial CT scan image of osteoblastic lesion; (B) sagittal spin echo MRI. (C) TPBrevealed osteosarcoma.

the vertebral body and disc space. The tip of the nucleotome is flexible to a maximumangulation of 90° in order to permit excision of different parts of the disc. The wholeprocedure is performed under fluoroscopic guidance. After completion of the discec-tomy, 10 French metal braided sheaths (Arrow International, Reading, PA) go throughthe pedicular channels into the discs for irrigation and drainage. These sheaths areattached to suction from a vacuum draining bag (Snyder Hemovac, Zimmer Patient

194 Hadjipavlou et al.

Fig. 17. Axial CT scan of (A) an osteolytic vascular lesion as seen on (B) arteriogram. (C)TPB revealed hemangioendotheliosarcoma.

Percutaneous Transpedicular Approaches 195

Fig. 18. (A) Axial CT scan demonstrating a painful osteoid osteoma of pedicle. TPB coredout the whole osteoid osteoma (B) as seen in (C). This biopsy was diagnostic and therapeutic.Three years postoperatively the patient was free of pain.

Care Division, Dover, OH). Irrigation takes place by instilling a solution of 2 g ofcefazolin (Ancef; Smith-Kline Beecham, Philadelphia, PA) and 10 mL of saline. Even-tually, culture results will dictate the choice of antibiotics.

Discussion

Percutaneous transpedicular discectomy for spondylodiscitis is a technically safesurgical procedure and is feasible in the thoracic as well as the lumbar spine. The

196 Hadjipavlou et al.

Fig. 19. (A, B [right]) Diagrammatic demonstration of a guide pin into intervertebral disc(A, lateral lumbosacral view). A 2-mm Steinmann pin is introduced percutaneously rostrallyangled through the pedicle, which is caudal to the affected disc, and advanced to the disc (rightside). (B, left) Axial view of diagrammatic demonstration of pin into disc, with dilator and exter-nal sleeve abutting against pedicle. The toothed biopsy cutting tool removes a core of bone fromthe pedicle and vertebral body to allow easy passage of the dissection forceps (C). The externalsleeve allows easy percutaneous passage of the discectomy instrumentation (D). (Modified withpermission from ref. 41.)

transpedicular tract allows the use of relatively large instruments for aggressive decom-pression without concern about possible spinal cord, nerve root, or vascular injuries.Our technique advocates bilateral access with channels measuring 5.15 mm, whichallow the passage of relatively large discectomy forceps and an automated nucleotome.We strongly urge that access of the intended discectomy level be from the more cau-dally placed adjacent pedicle. Access through a more cephalad pedicle has the potentialof penetrating the inferior borders of the pedicle and damaging the exiting nerve root.

Percutaneous Transpedicular Approaches 197

Fig. 19. (Continued)

We also strongly recommend that the procedure take place under fluoroscopic guid-ance, aiming the guide pin a bull’s-eye into the pedicular center or just superior to thepedicular equator. The procedure also allows the installation of Hemovac tubes (Zim-mer Health Care Division, Dover, OH) for drainage and antibiotic irrigation. Althoughthe procedure can be done safely and effectively under local anesthesia, we advocate

198 Hadjipavlou et al.

Fig. 20. (A) AP and (B) lateral radiograph demonstrating flexible nucleotome within discspace during the procedure, debulking infected disc and evacuating pus and necrotic material.(C) Appearance of nucleotome in action on one side and discectomy by means of Kambin dis-cectomy forceps on right side. (D) Axial CT scan of vertebral body demonstrating drain tubetransversing pedicle. (Partially reproduced with permission from ref. 38.)

general anesthesia because of severe pain in most patients with spondylodiscitis. Localanesthesia is useful in high-risk septic patients or those with other serious medicalconditions. Immediate response after transpedicular discectomy is usually observed in75% of unselected patients (41,43,81). With proper indications, as we have practicedever since the publication of the original article, we have achieved almost a 95% suc-cess rate (Figs. 21 and 22).

Percutaneous transpedicular discectomy is ineffective for the treatment of spondy-lodiscitis with severe neurological deficit caused by large epidural inflammatory tissue

Percutaneous Transpedicular Approaches 199

Fig. 20. (Continued)

200 Hadjipavlou et al.

compressing the neural elements. Therefore, percutaneous transpedicular discectomy iscontraindicated for the treatment of any spinal epidural abscess, or when there is neuro-compression of the cord or the conus medullaris in the thoracic or thoracolumbar spineby inflammatory granulation tissue.

In conclusion, percutaneous transpedicular discectomy is safe and highly effectiveduring the early stages of spondylodiscitis, when bone destruction is not extensive. It isineffective in the presence of infected disc herniation, foraminal stenosis, and excessivebone destruction with spinal deformity. This procedure is contraindicated when there isspinal epidural abscess and neurocompression by deformity, inflammatory tissue, or acombination thereof.

Fig. 21. (A) AP and (B) lateral view of spondylodiscitis of T4–T5 region treated by percuta-neous transpedicular discectomy. (C,D) Five months later there was a complete bony ankylosis.(Reproduced with permission from ref. 41.)

Percutaneous Transpedicular Approaches 201

REFERENCES

1. Mirra JM, Eckardt JJ, Rosen G. Biopsy considerations, in Bone Tumors: Clinical, Radio-logic and Pathologic Correlations (Picci P, Gold RH, eds.), Lea & Febiger, Philadelphia,1989, pp. 31–34.

2. Kattapuram SV, Khurana JS, Rosenthal DI. Percutaneous needle biopsy of the spine. Spine1992;17:561–564.

3. Kattapuramm SV, Rosenthal DI. Percutaneous biopsy of skeletal lesions. Am J Roentgenol1991;157:935–942.

4. Metzger CS, Johnson DW, Donaldson WF. Percutaneous biopsy in the anterior thoracicspine. Spine 1993;18:373–378.

5. Murphy WA, Destouet JM, Gilula LA. Percutaneous skeletal biopsy 1981: a procedure forradiologists—results, review and recommendations. Radiology 1981;139:545–549.

6. Kattapuram SV, Rosenthal DI. Percutaneous needle biopsy of the spine, in Tumors of theSpine: Diagnosis and Clinical Management (Sundaresan N, Schmidek HH, Schiller Al,Rosenthal DI, eds.), WB Saunders, Philadelphia, 1990, pp. 46–51.

7. Robertson RC, Ball RP. Destructive spine lesions: diagnosis by needle biopsy. J Bone JointSurg 1935;17:749–758.

8. Laredo JD, Bard M. Current status of musculoskeletal interventional radiology. RadiolClin North Am 1994;32:377–398.

Fig. 22. (A) A sagittal T2-weighted MRI image of the lumbar spine in a 38-yr-old womandemonstrates changes typical of spondylodiscitis with a small epidural component. (B) A sagit-tal T2-weighted MRI image 2 mo postoperatively, showing resolution of the infection withoutkyphosis. The discectomy accelerated the natural process of healing and prevented kyphoticdeformity. (Reproduced with permission from ref. 38.)

9. Tehranzadch J, Freiberger RH, Glielman, B. Closed skeletal needle biopsy review of 120cases. Am J Roentgenol 1983;140:113–115.

10. Fyfe IS, Henry APJ, Mulholland RC. Closed vertebral biopsy. J Bone Joint Surg (Br)1983;65:140–143.

11. Kornblum MB, Wesolowski DP, Fischgrund JS, Herkowitz HN. Computedtomography–guided biopsy of the spine: a review of 103 patients. Spine 1998;23:81–85.

12. Stringham DR, Hadjipavlou A, Dzioba RB, Lander P. Percutaneous transpedicular biopsyof the spine. Spine 1994;19:1985–1991.

13. Roy-Camille R, Saillant G, Mamoudy P. Biopsie du corps vertebral par voie posterieuretranspediculaire. Rev Chir Orthop 1983;69:147–149.

14. Banta CJ, King AG, Dabezies EJ, Liljeberg RL. Measurement of effective pedicle diameterin the human spine. Orthopedics 1989;12:939–942.

15. Berry JL, Moran JM, Berg WS, Steffee AD. A morphometric study of human lumbar andselected thoracic vertebrae. Spine 1987;12:363–367.

16. Krag MH, Weaver DL, Beynnon BD, Haugh LD. Morphometry of the thoracic and lumbarspine related to transpedicular screw placement for surgical spinal fixation. Spine1988;13:27–32.

17. Misenhimer GR, Peek RD, Wiltse LL, Rothman SLG, Widell EH. Anatomic analysis ofpedicle cortical and cancellous diameter as related to screw size. Spine 1989;14:367–372.

18. Olsewski JM, Simmons EH, Kallen FC, Mendel FC, Severin CM, Berens DL. Morphome-try of the lumbar spine: anatomical perspectives related to transpedicular fixation. J BoneJoint Surg (Am) 1990;72A:541–549.

19. Schmidek HH, Gomes FB, Seligson D, McSherry JW. Management of acute and stablethoracolumbar with and without neurologic deficit. Neurosurgery 1980;7:30–35.

20. Whitecloud TS, Skalley TC, Cook SD, Morgan EL. Roentgenographic measurement ofpedicle screw penetration. Clin Orthop 1989;245:57–68.

21. Zindrick MR, Wiltse LL, Doornik A, et al. Analysis of the morphometric characteristics ofthe thoracic and lumbar pedicles. Spine 1987;12:160–166.

22. Yamagata M, Hiroshi K, Shohei M, et al. Mechanical stability of the pedicle screw for thelumbar spine. Spine 1992;17:551–554.

23. Greenfield RT, Grant RG, Bryant D. Pedicle screw fixation in the management of unstablethoracolumbar spine injuries. Orthop Rev 1992;6:701–706.

24. King AG, MacEwen GD, Bose WJ. Transpedicular convex anterior hemiepiphysiodesisand posterior arthrodesis for progressive congenital scoliosis. Spine 1992;17:S291–S294.

25. Steffee AD, Biscup RS, Sitowski DJ. Segmental spine plates with pedicle screw fixation: anew internal fixation device for disorders of the lumbar and thoracolumbar spine. ClinOrthop 1986;203:45–53.

26. Steffee AD. Segmental spine plates with pedicle screw fixation. Clin Orthop 1986;103:45–53.

27. Olerud S, Karlstrom G, Sjostrom L. Transpedicular fixation of thoracolumbar vertebralfractures. Clin Orthop 1988;227:44–51.

28. Roy-Camille R, Saillant G, Mazel CH. Plating of thoracic, thoracolumbar, and lumbarinjuries with pedicle screw plates. Orthop Clin North Am 1986; 17:147–159.

29. Magerl FP. Stabilization of the lower thoracic and lumbar spine with external skeletal fixa-tion. Clin Orthop 1984;189:125–141.

30. Hardaker WT, Cook WA, Friedman AH, Fitch RD. Bilateral transpedicular decompressionand Harrington rod stabilization in the management of severe thoracolumbar burst frac-tures. Spine 1992;17:162–171.

31. Patterson RH Jr, Arbit E. A surgical approach through the pedicle to protruded thoracicdiscs. J Neurosurg 1978;48:768–772.

32. Duncan GA, Ferguson AB. Benign giant-cell tumor of the fourth lumbar vertebra: a casereport. J Bone Joint Surg (Am) 1936;3:769–772.

202 Hadjipavlou et al.

Percutaneous Transpedicular Approaches 203

33. Capener N. The evolution of lateral rachiotomy. J Bone Joint Surg (Br) 1954;36:173–176.34. Michele JM, Krueger FJ. Surgical approach to the vertebral body. J Bone Joint Surg (Am)

1949;31:873–878.35. Travaglini F. Tumours of the vertebral body: transpedicular biopsy with posterior surgical

stabilization. Bull Hosp Joint Dis 1979;40:1–12.36. Stoker DJ, Kissin CM. Percutaneous vertebral biopsy: a review of 135 cases. Clin Radiol

1985:36:569–577.37. Fidler MW, Niers BBAM. Open transpedicular biopsy of the vertebral body. J Bone Joint

Surg (Br) 1990;72:884, 885.38. Arya S, Crow WN, Hadjipavlou AG, Nauta HJW, Borowski AM, Vierra LA, Walser E. Per-

cutaneous transpedicular management of discitis. JVIR 1996; 7:921–927.39. Borowski AM, Crow WN, Hadjipavlou AG, Chaljub G, Mader J, Cesani F, vanSonnenberg

E. Percutaneous management of pyogenic spondylodiscitis. AJR 1998;170:1587–1592.40. Hadjipavlou AG, Arya S, Crow WN, Maggio WW, Lander P, Nardone EM, Eyal A. Percu-

taneous transpedicular biopsy of the spine. J Intervent Radiol 1996;11:103–108.41. Hadjipavlou AG, Crow WN, Borowski A, Mader JT, Adesokan A, Jensen R. Percutaneous

transpedicular discectomy and drainage in pyogenic spondylodiscitis. Am J Orthop1998;188–197.

42. Hadjipavlou AG, Kontakis GM, Gaitanis I, Katonis PG, Lander P, Crow WN. Effectivenessand pitfalls of percutaneous transpedicular biopsy of the spine. Clin Orthop 2003;411:54–60.

43. Hadjipavlou AG, Mader JT, Necessary JT, Muffoletto AJ. Hematogenous pyogenic spinalinfections and their surgical management: Spine 2000;25:1668–1679.

44. Kambin P (ed). Arthroscopic Microdiscectomy: Minimal Intervention in Spinal Surgery,Williams & Wilkins, Baltimore, 1990.

45. Saillant G. Etude anatomique des pedicules vertebraux: application chirurgicale. Rev ChirOrthop 1976;62:151–160.

46. Ball RP. Needle (aspiration) biopsy. J Tenn State Med Assoc 1934;27:203–207.47. Ward JC, Jeanneret B, Oehlschlegel C, Magerl F. The value of percutaneous transpedicular

vertebral bone biopsies for histologic examination: results of an experimental histopatho-logic study comparing two biopsy needles. Spine 1996;21:2484–2490.

48. Renfrew DL, Whitten CG, Wiese JA, el-Khoury GY, Harris KG. CT-guided percutaneoustranspedicular biopsy of the spine. Radiology 1991;180:574–576.

49. Hsu WC, Lim KE. Computed tomography–guided percutaneous transpedicular biopsy ofthe thoracic spine. Chang Gung Med J 2001;24:368–375.

50. Jho HD. Endoscopic microscopic transpedicular thoracic discectomy: technical note. JNeurosurg 1997;87:125–129.

51. Minart D, Vallee JN, Cormier E, Chiras J. Percutaneous coaxial transpedicular biopsy ofvertebral body lesions during vertebroplasty. Neuroradiology 2001;43:409–412.

52. Moller S, Kothe R, Wiesner L, Werner M, Ruther W, Delling G. Fluoroscopy-guidedtranspedicular trocar biopsy of the spine—results, review and technical notes. Acta OrthopBelg 2001;67(5):488-499.

53. Pierot L, Boulin A. Percutaneous biopsy of the thoracic and lumbar spine: transpedicularapproach under fluoroscopic guidance. Am J Neuroradiol 1999;20:23–25.

54. Ashizawa R, Ohtsuka K, Kamimura M, Ebara S, Takaoka K. Percutaneous transpedicularbiopsy of thoracic and lumbar vertebrae—method and diagnostic validity. Surg Neurol1999;52(6):545–551.

55. Jelinek JS, Kransdorf MJ, Gray R, Aboulafia AJ, Malawer MM. Percutaneous transpedicularbiopsy of vertebral body lesions. Spine 1996;21:2035–2040.

56. Dagirmanjian A, Schils J, McHenry M, Modic MT. MR imaging of vertebral osteomyelitisrevisited. Am J Roentgenol 1996;167:1539–1543.

57. Ambrose GB, Alpert M, Neer CS. Vertebral osteomyelitis: a diagnostic problem. JAMA1966;197:619–622.

204 Hadjipavlou et al.

58. Waldvogel RA, Papageorgiou PS. Osteomyelitis: the past decade. N Engl J Med 1980;303:360–370.

59. Ottolenghi CE. Aspiration biopsy of the spine: technique for the thoracic spine and resultsof twenty-eight biopsies in this region and overall results of 1050 biopsies of other spinalsegments. J Bone Joint Surg 1969;51A:1531–1544.

60. Post MJD, Sze G, Quencer RM, et al. Gadolinium enhanced MR in spinal infection. JComput Assist Tomogr 1990;14:721–729.

61. Sklar EML, Post MJD, Lebwohl NH. Imaging of infection of lumbosacral spine. Neuroimaging1993;3:577–590.

62. Wisneski RJ. Infectious disease of the spine: diagnostic and treatment considerations.Orthop Clin North Am 1991;22:491–501.

63. Eismont FJ, Bohlman HH, Soni PL, et al. Pyogenic and fungal vertebral osteomyelitis withparalysis. J Bone Joint Surg 1983;65A:19–29.

64. Emery SE, Chan DP, Woodward HR. Treatment of hematogenous pyogenic vertebralosteomyelitis with anterior debridement and primary bone grafting. Spine 1989;14:284–291.

65. Estherhai JL Jr, Gelb I. Adult sepric arthritis. Orthop Clin North Am 1991;22:503–514.66. Krodel A, Sturz H, Siebert CH. Indication for and results of operative treatment of

spondylitis and spondylodiscitis. Arch Orthop Trauma Surg 1991;110:78–82.67. Lane JG, Falahaee MH, Wojtys EM, et al. Pyarthrosis of the knee: treatment considera-

tions. Clin Orthop 1990;252:198–204.68. Lee TC, Lu K, Yang LC, Huang HY, Liang CL. Transpedicular instrumentation as an

adjunct in the treatment of thoracolumbar spine tuberculosis with early stage bone destruc-tion. J Neurosurg 1999;91:163–169.

69. Taylor TKF, Dooley BJ. Antibiotics in the management of postoperative disc space infections.Aust NZ J Surg 1978;48:74, 75.

70. Jackson RW. The septic knee arthroscopic treatment. Arthroscopy 1985;1:194–197.71. Ivey M, Clark R. Arthroscopic debridement of the knee for septic arthritis. Clin Orthop

1985;199:201–206.72. Gebhard JS, Brugman JL. Percutaneous discectomy for the treatment of bacterial discitis.

Spine 1994;19:855–857.73. Onik G, Shang Y, Maroon JC. Automated percutaneous biopsy in postoperative diskitis: a

new method. AJNR 1990;1:391–393.74. Vu WY, Siu C, Tong PC, et al. Percutaneous suction aspiration for osteomyelitis: report of

two cases. Spine 1991;16:198–202.75. Fraser RD, Osti OL, Vernon-Roberts B. Iatrogenic discitis: the role of intravenous antibi-

otics in prevention and treatment: an experimental study. Spine 1989;14:1025–1032.76. Lucio E, Adesokan A, Hadjipavlou AG, Crow WN, Adegboyega PA. Pyogenic spondy-

lodiskitis. Arch Pathol Lab Med 2000;124:712–716.77. Garcia A, Grantham SA. Hematogenous pyogenic vertebral osteomyleitis. J Bone Joint

Surg 1960;42A:429–436.78. Sapico FL, Montgomerie JZ. Pyogenic vertebral osteomyelitis: report of nine cases and

review of the literature. Rev Infect Dis 1979;1:754–756.79. Frederickson B, Yuan H, Olans R. Management and outcome of pyogenic vertebral

osteomyelitis. Clin Orthop 1978;131:160–167.80. Guven O, Yalgin S, Karahan M, Esem*nli T. Transpedicular drainage of Pot’s abscess: a

report of two cases. Am J Orthop 1995;24:421–425.81. Crow WN, Borowski AM, Hadjipavlou AG, Walser EM, Arya S, Calme MB, Amps J, Jensen

R, Somisetty S, Alford B, Adesokan A. Percutaneous transpedicular automated nucleotomyfor debridement of infected discs. JVIR 1998;9:161–165.

9Selective Endoscopic Discectomy™

Twelve Years of Experience

Anthony T. Yeung, MD

INTRODUCTION

More than 12 yr have passed (1991) since I learned and adopted arthroscopicmicrodiscectomy (AMD) from pioneer endoscopic spine surgeon Parviz Kambin, who,along with Sadahisa Hijikata, first established the technique for percutaneousnucleotomy in the early 1970s. A cadaver dissection of the traditional posterioranatomy of the lumbar spine compared with the foraminal anatomy clearly illustratesthe feasibility and advantages of the foraminal approach to the lumbar disc (Fig. 1A,B).Kambin’s AMD technique evolved gradually to allow for more dorsal placement of thecannula to effect posterolateral fragmentectomy and resection of the posterior annulusfor stenosis, and it made possible the removal of extruded and sequestered herniateddiscs (1–4) (Fig. 2). Hijikata (5) also recently reviewed his 12 yr of experience withendoscopic discectomy. In 1996, a new design of the operating spine endoscope addingmultichannel irrigation and complementary instrumentation (Fig. 3) (6) allowed furtherdevelopment of endoscopic spine surgery to include the treatment of annular tearscausing discogenic back pain (7,8). Advanced techniques of foraminoplasty for centraland lateral recess stenosis followed (9–12). The third-generation system design changeto the Yeung Endoscopic Spine Surgery (YESS) system gave me more flexibility tomaneuver the endoscope and improved ability to probe spinal anatomy in a consciouspatient (13). Continued evolution of this technique afforded me the ability to betterevaluate the pathological process causing the patient’s discogenic back pain. Condi-tions previously not considered surgical, such as annular tears, were evaluated andmanaged successfully through the endoscope. Synovial facet cysts, inflammatorymembranes containing neoneurogenesis, osteophytosis impinging on sensitive nerves,anomalous furcal nerves, autonomic nerves, and conditions irritating the dorsal rootganglion (DRG) of the exiting nerve were identified as sources of discogenic painand sciatica.

From 1991 to 2004, I treated more than 2300 patients with discogenic pain, degenera-tive conditions of the lumbar spine, and the whole spectrum of disc herniations includingextruded and sequestered fragments (14,15). The success rate in the first 500 patients was432 of 500 (86%) good/excellent results according to the modified MacNab criteria (11).

205

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

A subsequent retrospective study of 219 consecutive patients with radiculopathy sec-ondary to large intracanal noncontained lumbar disc herniations demonstrated a satis-factory outcome in 204 (93.1%) patients based on modified MacNab criteria, but therate was even higher (94.8%) when patients were asked to respond to a study patient-based outcome questionnaire (16). In this chapter, I review my 12 yr of experienceevolving from Kambin’s AMD into the treatment of discogenic back pain and sciaticaby selective endoscopic discectomy.

206 Yeung

Fig. 1. (A) Anatomy of the posterior port provides easier access to the posterior disc andspinal canal at L5-S1 (blue hubbed needle), but with planning, most contained disc herniationscan be removed posterolaterally. (B) Anatomy of posterolateral foraminal port from L2-S1. Onlyin the L5-S1 disc space is access to the spinal canal restricted because of the pelvis and the relativelywide facet (gray hubbed needle in the L5-S1 disc). High lumbar disc herniations from L1 to L3are easier to reach endoscopically through the posterolateral foraminal portal. L4-L5 providesample room for either approach. Note the furcal nerve branches entering the psoas muscle.

ROLE OF EVOCATIVE CHROMODISCOGRAPHY

At Kambin’s 1991 course, Prof. Hans Joerg Leu described the use of indigocarmine dyeto stain and label the nucleus pulposus (NP). To maintain the ability to recognize structuralanatomy, it was necessary to dilute the dye to a 10% solution to effect differential tissue

Selective Endoscopic Discectomy 207

Fig. 2. The dome. Spinal structures in the foramen accessible to visualization and surgical inter-vention and probing via the posterolateral approach include the annulus, disc, pedicle, facet, andepidural space. This approach also visualizes neurocompression in the “hidden zone” of the lateralrecess, a common cause of failed back surgery syndrome (FBSS). (Courtesy of Hal Matthews, MD.)

Fig. 3. Yeung spine scope system. YESS discoscope and partial instrument set. The spinalendoscope is designed with multichannel irrigation and a cannula system that allows access totargeted areas while protecting sensitive nerves. (From ref. 15.)

staining that did not overwhelm the NP with stain. I adopted this adjunctive techniqueinitially to help with a visualized nuclectomy (17). With differential staining, it was easierto recognize NP from annulus and from facet capsule. The epidural space with its epiduralvessels and fat was simple to recognize. When pain was reproduced by discography, theclinical improvement in the patient’s back pain correlated well with concordant painreproduction. The use of discography also helped predict whether the herniation wasextruded or contained, and the nuclear material was clearly stained for easier endoscopicextraction (18) (Fig. 4). I trademarked evocative chromodiscography™ as an integral partof spinal endoscopy. The process of removing the indigocarmine dye-labeled nucleus wastrademarked selective endoscopic discectomyTM and this technique is the focus of thischapter (15) (Fig. 5).

The literature on discography is currently considered controversial only because of thehigh interobserver variability by discographers in reporting the patient’s subjective painas well as the ailing patient’s ability to give a clear response, especially if the painresponse is altered by the use of analgesics or sedation during the procedure. Althoughmuch of the literature that contributes to the controversy of discography points out the pit-falls of depending on discography, the majority of the literature supports its use by clini-cians who know how to use it. The surgeon who is accomplished in endoscopic spinesurgery prefers to do the discography himself or herself in order to decrease the interob-server variability in interpreting the patient’s response. When a discographer compares hisor her own assessment of the patient’s pain response with the report of another discogra-pher, there can be some variability in diagnosis and interpretation. This variability mayresult in unpredictable treatment results. The incidence of “false-positive” discograms,

208 Yeung

Fig. 4. This intraoperative discogram in a patient with a papacentral herniated nucleus pulposus(HNP) by magnetic resonance imaging (MRI) not only confirmed the patient’s concordant backpain and sciatica, but the leakage of contrast to the traversing and exiting nerves alerted the surgeonto look for extruded disc fragments and grade V annular tears, especially in the lateral zone.

however, can be significantly decreased in the hands of an experienced endoscopic sur-geon. False-positive discography should really be false interpretation of positive discog-raphy results. The experienced endoscopic surgeon learns to correlate the patient’sresponse to the discogram pattern of the painful disc that is being treated. There is goodcorrelation of discograms with different types of annular tears and disc herniations. Thesurgical result can then be predicted on the basis of the visualized condition. For example,the discogram can be used to predict the presence of a collagenized disc fragment vs a softherniation; the extrusion of a disc fragment as a noncontained herniation; or the presenceof the type, grade, and location of a painful vs nonpainful annular tear. Discography isused by the surgeon as a means of further identifying concordant discogenic pain in clini-cal situations in which the patient’s clinical presentation is matched with MRI findings.Controversy in the literature has arisen because of the lack of a good spectrum of thera-peutic surgical treatments once the pain is confirmed, and because of the plethora ofarticles pointing out the pitfalls of false interpretation (19,20).

ROLE OF ELECTROTHERMAL THERAPY

Prior to my adoption of AMD, I was using the potassium-trideuterium-phosphate(KTP) laser for laser disc decompression. When I combined the two techniques fornuclectomy, the laser provided hemostasis and better visualization (21–24). I observedthat patients with disc protrusions but with predominant back pain who were not candi-dates for traditional transcanal surgery found relief of their back pain with KTP-assistedAMD. The staining of the disc provided a chromophore that enhanced the efficacy of theKTP laser (25). In 1993, I tested a unipolar electrode by Ellman and introduced to me by

Selective Endoscopic Discectomy 209

Fig. 5. NP stained with indigocarmine dye. The indigocarmine dye stains the NP light blue,helping the endoscopic surgeon target the stained nucleus for extraction. Here, the herniation hasextruded past the unstained annulus. Removal of the extruded herniation will expose the traversingnerve in the epidural space.

Dr. Peter Morrison. Later, an electrode made by Smith and Nephew was used with theEllman unit and the new working channel endoscope. In a retrospective 2- to 4.5-yr fol-low-up study of my first 100 patients by Farouq Al-Hamdan, a spine fellow under AlexHadjipavlou, it was documented that the use of the KTP laser as an adjunct to AMDrelieved back pain as well as leg pain in 65% of the patients. The overall good/excellentresult by MacNab criteria was 89% for sciatica. The KTP laser was initially used to pro-vide hemostasis and better visualization, but its side effect of laser thermal annuloplastyprompted an International Review Board (IRB) study using a temperature-controlledflexible probe by Oratec in lieu of the laser.

In 1998, an IRB-approved study commenced at St. Luke’s Medical Center toevaluate the efficacy of electrothermal treatment in the process of arthroscopicmicrodiscectomy for herniated discs. This study, sponsored by Oratec, using a tem-perature-controlled flexible probe, provided a better tool to contract annular defectscaused by the disc herniation and ablating granulation tissue in annular tears. Thepatient’s response to this application of electrothermal energy provided informationthat electrothermal treatment of the disc was effective in decreasing discogenic backpain, but the fluctuations in temperature control caused me to switch to a bipolarflexible probe (26,27). Oratec investigated the intradiscal electrothermal (IDET)catheter in the same time period. I now use a bipolar flexible probe by Ellmandesigned specifically for thermal annuloplasty that has proven to be as effective asthe Oratec probe, but with a more established and accepted use of electrothermalenergy in spine surgery (27) (Fig. 6).

Rauschning’s cryosections of normal and pathoanatomy have more recently demon-strated inflammation in and around the sensitive DRG and identified granulation tissuein annular tears (28,29) (Figs. 7 and 8). It is also well known that although a spinal struc-ture is capable of pain, spinal pathology on imaging studies does not always correlatewith the debilitating pain that is resistant to conservative management (30). What maybe very painful in one person may be well tolerated or painless in another. Evocativediscography has been shown to be helpful in identifying the disc as a pain generator inaxial back pain and sciatica (18,31–37), and electrothermal treatment of the disc isdemonstrated to be effective in decreasing discogenic back pain.

210 Yeung

Fig. 6. Ellman Bipolar Triggerflex Probe. This bipolar flexible device has provided effectivehemostasis as well as electrothermal shrinkage of disc tissue and annular tears under visualizedcontrol. It offers better control of the energy source by its bipolar design.

Selective Endoscopic Discectomy 211

Fig. 7. (A) Annular tears. This computed tomography (CT) discogram outlines a foraminalHNP and far-lateral annular tear that will irritate the DRG of the exiting nerve. Endoscopic visual-ization of the foramen may reveal the presence of an inflammatory membrane, extruded NP, andgranulation tissue. (B) Granulation tissue and inflammation surrounding small disc fragment inforamen. Grade V annular tears open into the epidural space or psoas muscle, allowing theingrowth of nerves and capillaries and creating an inflammatory response, which, if next to a spinalnerve or the DRG, can cause pain out of proportion to what may be anticipated from traditionalimaging studies. Tears in this area will also result in groin pain by irritating the psoas muscle andgenital-femoral nerve. Patients with annular tears who obtain pain relief from foraminal epiduralblocks may experience more lasting relief of 2 or more years with selective endoscopic discectomyand thermal annuloplasty. (C) Bipolar radiofrequency treatment of annular tears under direct visu-alization. Interpositional disc material should be removed from the annular layers to treat the annu-lar tear effectively. (D) Preoperative endoscopic view of grades V annular tear demonstratinggranulation tissue in tear. (E) Postoperative view of annulus after thermal modulation with EllmanBipolar Triggerflex Probe.

ROLE OF THE LASER

Laser technology also evolved to become more user friendly. The first laser to beapproved by the Food and Drug Administration was the KTP laser, a laser in the blue/greenwavelength spectrum that was effective for soft-tissue ablation, but the bright light limitedits use when direct visualization was desired (22). The Holmium:yttrium-aluminum-garnet(YAG) laser was effective for the ablation of soft tissue as well as bone. Current designsnow include a side-firing irrigated probe and a straight fiber that can be delivered through acurved guide that will angle the laser beam up to 45°. The laser has opened the door for theremoval of osteophytes and lateral stenosis that cause neuropathic pain in patients who

212 Yeung

Fig. 7. (Continued)

Fig. 8. Exiting nerve. The exiting nerve is in the “hidden extraforaminal zone” that is irri-tated by far-lateral annular tears and disc herniations that escape detection by MRI and tran-scanal surgical exploration. It contains the DRG, which, when sensitized by the inflammatorybyproducts of a degenerative disc, is responsible for the “nondermatomal” distribution in patientswith chronic sciatica. The presence of fat around the exiting nerve is a more sensitive indication oflateral recess stenosis than findings on MRI or CT myelogram.

have no other surgical options (15). Figures 9 and 10 demonstrate the use of the laser inforaminoplasty.

ADJUNCTIVE THERAPY WITH CHYMOPAPAIN

Chymopapain is the only minimally invasive technique that has been validated withtwo large double-blind studies and numerous cohort studies that found it effective forthe treatment of contained disc herniations. I have used chymopapain to assist theextraction of large noncontained disc herniations that extrude past the outer annularfibers, with a good/excellent result by MacNab criteria 10% higher than when no chy-mopapain was used (38). If the height of the herniation is greater than the base on MRI,it is likely that the herniated nucleus is collared by the annulus, making it more difficultto remove from within the disc. Chymopapain-treated NP is soft and slippery, makingmechanical removal easier. If chymopapain extravasates along the course of the con-trast agent used for discography, it will theoretically denature and treat the extrudedfragment to make it less inflammetogenic. The extruded fragment is also exposed tophagocytosis and eventual absorption if exposed to the epidural vasculature. In morethan 500 surgical cases of chymopapain-assisted selective endoscopic discectomy, Ihave never had any complications from the use of chymopapain. Because of theabsence of complications when used in conjunction with endoscopic disc removal, I

Selective Endoscopic Discectomy 213

Fig. 9. Role of laser in foraminoplasty. Side-firing laser (Trimedine) or end-firing laser fibers(Lisa) directed by a flexible cannula are used for precise foraminoplasty of the lateral recess.

now do not find it necessary to test routinely for antibodies to chymopapain with theChymofast test unless the patient requests it (Figs. 11–13).

NEUROMONITORING

In 100 consecutive cases, I studied neuromonitoring to determine whether sensory/motorcomplications could be reduced. I specifically studied whether neuromonitoring by continu-ous electromyogram (EMG) or somatosensory evoked potentials (SSEP) would affect theincidence of dysesthesia or help predict decompression of the nerve (39). Although anincrease in conduction velocity was identified when a mechanically compromised nervewas decompressed, I concluded that neuromonitoring was no more effective than monitor-ing the patient for intraoperative pain when a dilute solution of lidocaine (0.5%) was used.There was no difference in the dysesthesia or complication rate of the 100 cases vs a

214 Yeung

Fig. 11. (A) Chymopapain-assisted selective endoscopic discectomy. (B) Preoperative MRIof extruded fragment in horizontal disc at L5-S1 that is anatomically difficult to reach. By usingchymopapain, the results of endoscopic removal have been demonstrated to be improved by10% overall, by reducing the rate of residual HNP or recurrent HNP.

Fig. 10. (A) Postoperative endoscopic view of foraminoplasty of superior articular facet atL5-S1; (B) postoperative view of decompressed exiting nerve after foraminoplasty.

Selective Endoscopic Discectomy 215

Fig. 12. An extruded fragment labeled by indigocarmine and pretreated by chymopapainallowed for easier manual extraction. The chymopapain loosened the fragment, making it easierto remove, aided by suction on the working channel of an endoscope.

Fig. 13. Chymopapain-treated NP. Note the differential staining of the extracted nuclearmaterial. Unstained collagenized disc was extracted from the epidural space where the indigo-carmine dye did not reach. By visualizing the decompressed foramen, successful relief of legpain was immediate postoperatively.

matched number of patients who had no neuromonitoring. Although neuromonitoring maygive the novice surgeon a greater sense of security early in his or her endoscopic practice,analysis of the results of the prospective study of 100 consecutive patients did not shown

216 Yeung

Fig. 14. Foraminal epidurograms. Foraminoepidurography is a new technique for foraminalneedle placement from the far-lateral skin portal mimicking surgical access to the epidural spacethat allows the surgeon to produce an epidurogram that complements the MRI by outlining theposition of the traversing and exiting nerves in the foramen. This information provides the sur-geon with additional information preoperatively and serves as a “practice run” for surgery.

neuromonitoring to be any more useful for avoiding complications than patient feedback onpain during the procedure.

FORAMINOGRAPHY AND THERAPEUTIC INJECTIONS

The efficacy of endoscopic lumbar disc surgery can be enhanced by surgeons per-forming than own discography and foraminal injections. Foraminal epidurography andforaminal therapeutic injections are correlated with identification of pathoanatomy inthe lumbar spine (Figs. 14–17). Surgeons use information gleaned from these injectionprocedures to better select patients for surgical interventional techniques that the theyhave incorporated into their endoscopic surgical practice. Patients with disc protru-sions, annular tears, and foraminal stenosis may get temporary relief with the therapeu-tic injection, but if the response is short-lived, additional information gathered byperforming the epiduralgram will help guide surgeons when they must consider thetechnical feasibility of using the same or similar portal for the insertion of the operatingcannulas. By performing epidurograms, surgeons can obtain additional information ofthe anatomy of the foramen, the outline of the traversing and exiting nerves, and thetherapeutic response afforded by the epidural injection (37). Injection at L5-S1 will alsohelp surgeons with preoperative planning if surgery is eventually required.

PRESENT APPLICATIONS AND FUTURE TRENDS

Indications

Any pathological lesion that is accessible, is visible, is treatable, or requires endoscopicconfirmation through the foramen may ultimately become an indication for diagnosticand therapeutic endoscopy. Patient selection for pain and radiculopathy from disc herni-ation is similar to selection criteria for traditional spine procedures. Endoscopic surgicalindications, however, may be dictated by the limitations of the endoscopic procedureitself with respect to the patient’s anatomy or the surgeon’s skill and experience withendoscopic spine surgery. At L5-S1, anatomical restrictions may cause the surgeon toopt for the posterior transcanal approach (Fig. 15A,B). For herniations from T10 to L4,the foraminal approach provides excellent access to the disc and epidural space. As theexperience of the surgeon increases, previous contraindications become relative,depending partly on the surgeon’s ability to endoscopically visualize, probe, and accessthe pathological lesion. Restrictions are dictated only by anatomical considerations inaccessing the patient’s spinal pathology and the rationale for the endoscopic procedureitself. As the surgeon’s experience increases, former contraindications become relative,depending on the surgeon’s experience, and his or her ability to address the spinal condi-tion to be treated. The three zones within reach of the spine endoscope transforaminallyare illustrated in Fig. 16.

Inclusion Criteria

Discogenic pain as determined by evocative discography implicates the disc as apain generator. Symptomatic disc herniation is the obvious indication, limited only bythe accessibility of endoscopic instruments to the herniated fragment. The ideal lesion

Selective Endoscopic Discectomy 217

Fig. 15. (A) Anatomical limitations at L5-S1. The narrow pelvis in this patient limited accessto the disc, but it was still possible to insert a needle into the disc for discography. Depending onthe type of disc protrusion, the endoscopic surgeon will have a better idea about the prognosis ofthe surgical procedure contemplated. If this patient had an extruded disc herniation, it would bebetter to opt for the transcanal approach. (B) Horizontal L5-S1 disc. The pelvis becomes an evengreater obstacle at L5-S1 if the disc is horizontal, which makes it much more difficult to get intothe epidural space. Removal of the lateral facet could overcome this obstacle. Good preoperativeplanning is enhanced by foraminoepidurography.

for endoscopic discectomy is a far-lateral, extraforaminal disc herniation. Traditionalapproaches to far-lateral disc herniations are more difficult, requiring a paramedianincision through very vascular tissue. The exiting nerve and the DRG are at risk of neu-ropraxia in both approaches (28,29). Although a traditional spine surgeon can accessthe lateral zone of the disc with a paramedian incision, it is easier to access theextraforaminal zone through the foramen. A typical foraminal view of NP extruded pastthe posterior annulus is shown in (Fig. 17). Through this approach to the disc, endo-scopic excisional biopsy and disc space debridement are also ideal for surgicallydebriding infectious discitis (Fig. 18). Currently treated with immobilization and par-enteral antibiotics, discitis is much more effectively treated when augmented by endo-scopic excisional biopsy and debridement. The surgeon will not have to be overlyconcerned about creating dead space for the inflamed or infected disc material to spreadinto the dead space created by a posterior approach. The clinical results are dramatic,and tissue biopsy is more accurate than needle aspiration in identifying the cause ofdiscitis. Even sterile discitis will benefit from intradiscal debridement and irrigation.

Foraminal stenosis in selected patients will respond to foraminoplasty by endoscopictechniques. Lateral recess stenosis is one cause of FBSS that can absolutely be diag-nosed and treated by foraminal decompression (40). The pathoanatomical finding maybe osteophytosis tethering the exiting nerve at the superior vertebral end plate and/or

218 Yeung

Fig. 16. Three zones on foramen accessible by endoscope. Zones II and III are not usuallyvisualized by surgeons using the transcanal approach unless they are experienced in theparamedical approach to the lateral recess.

stenosis and lack of fat around the exiting nerve (Fig. 19). Although trephines, rasps,and burrs can be used, the Ho:YAG side-firing laser is feasible as a visually controlledsoft-tissue and bone ablation device. The cannula chosen for this task has an open sidechannel that will protect the exiting nerve while the laser is used to dissect the tethering

Selective Endoscopic Discectomy 219

Fig. 17. Foraminal view of HNP. The indigocarmine-stained disc tissue has extruded past theposterior longitudinal ligament in this foraminal HNP at L4-L5.

Fig. 18. Discitis. Intradiscal view of discitis after debridement. Usual findings of inflamma-tory disc material and loose end-plate cartilage are readily removed from the disc space. Painrelief is immediate, and abundant tissue is available for laboratory analysis.

osteophyte and scar tissue from the nerve. Endoscopic foraminoplasty has not beenshown to cause increased instability even in spondylolisthesis. When mild degenerativespondylolisthesis is present, the disc bulge can be successfully treated by selectivediscectomy and thermal annuloplasty when there is a sciatic component to the patient’scomplaint. The technique is most useful for lateral recess stenosis, a condition responsiblefor atypical leg pain rather than true intermittent claudication of central spinal steno-sis. In central spinal stenosis, when there is concomitant posterior disc protrusion,decompression of the spinal canal can be effectively accomplished by resecting thebulging annulus in a collapsed disc, thus lowering the floor of the foramen. In isthmicspondylolisthesis, when there is more leg than back pain, this is usually owing toimpingement on the exiting nerve by the pseudoarthrosis at the pars defect. The goal isthen to decompress the compromised exiting nerve by elevating the dome formed by theinferior facet and lamina without further destabilizing the spinal segment.

Exclusion Criteria

Except for pregnancy, there are no absolute exclusion criteria, only relative contraindi-cations depending on the skill and experience of the surgeon. Spinal endoscopy and spinalprobing can be used for diagnostic purposes in extremely difficult or confusing clinicalsituations. Therefore, if endoscopy is helpful for diagnostic purposes, exclusion criteriamay depend mainly on the accessibility of the spinal pathology and the endoscopic skillsof the surgeon. The risks and benefits of the procedure must be weighed against the needto use this fluoroscopically guided procedure under local anesthesia or sedation.

220 Yeung

Fig. 19. Dissecting exiting nerve. In lateral recess stenosis, the scarred exiting nerve isreleased with a bare laser fiber. This picture demonstrates lateral recess stenosis as the mostcommon cause of FBSS.

Future Considerations

The spine scope will eventually be used for any condition for which visual inspectionof the foramen is desired. I have used spinal endoscopy for the following reasons: toinspect a spinal nerve that is suspected of being irritated by orthopedic hardware adjacentto the pedicle, to remove suspected recurrent or residual disc herniations that do not showup on imaging studies, to decompress the lateral recess by foraminoplasty, to removeosteophytes and facet cysts that cause unrelenting sciatica, and to locate painful lateralannular tears or small disc herniations not evident on physical examination or MRI. Insingle- and multilevel discogenic pain, for which the patient has no viable options, endo-scopic discectomy and thermal annuloplasty have been successful for treating chroniclumbar discogenic pain. A minority of patients may continue to have significant backpain, and a few may feel worse, but in the context of a progressive degenerative condi-tion, the results are encouraging and will give most patients relief while awaiting thedevelopment of newer procedures such as nucleus replacement, total disc replacement,and minimally invasive stabilization procedures of the posterior spinal column.

Current Imaging Methods

In my experience, imaging studies are only about 70% accurate and specific for pre-dicting pain. Conditions such as lateral annular tears, rim tears, end-plate separation,small subligamentous disc herniations, intranuclear herniations, anomalous nerves, andmiscellaneous discogenic conditions are cumulatively missed approx 30% of the time.These conditions can be diagnosed by means of spinal endoscopy. Tears that are in the lat-eral and anterior aspect of the disc are routinely missed on MRI studies. Very small discherniations that protrude past the outer fibers of the annulus are also missed because thefragment may be flattened against the posterior longitudinal ligament or nerve, appearingon the MRI as a thickened or bulging annulus, but really containing a subligamentous her-niation. When the nerve root is “swollen” or enlarged, MRI is not always capable of dis-tinguishing it from a conjoined nerve or a nerve with an adherent fragment of disc. Whenthe disc tissue is in direct contact with the nerve, the nerve can be irritated and a painfulinflammatory membrane forms. Even an epidural venous plexus that is inflamed can con-tribute to back pain and sciatica. Anomalous nerve branches known as furcal nerves arenever seen on MRI but can be visualized with spinal endoscopy of the foramen.

When an inflammatory membrane is present, the patient’s pain pattern can be confusing.Diagnostic spinal endoscopy has confirmed “nondermatomal” pain in scores of patientswith proximal thigh, buttock, and groin pain at levels distal to the root origin of theanatomical area.

Technique

Accessing the foramen is simplified and standardized by drawing coordinates on thepatient’s skin to determine the optimal skin window and annular window for positioningthe surgical instruments toward the center of the disc (Fig. 20). Reference points arethe anatomical center of the disc, the superior facet of the inferior vertebra, and theskin window. The needle trajectory must also be in a line of inclination between theend plates of the adjacent vertebrae. Adjustments in the trajectory will be made toaccommodate individual anatomical considerations and the pathology to be accessed.

Selective Endoscopic Discectomy 221

222 Yeung

Fig. 20. YESS technique determination of optimal instrument path using the Yeung instru-mentation trajectory protocol. Interoperative C-arm fluoroscopic imaging allows registration ofinternal structures with surface skin markings. (A) Posteroanterior fluoroscopic exposure enablestopographic location of spinal column midline and transverse planes of target discs. Intersectionsof the drawn lines mark posteroanterior disc centers. (B) Lateral fluoroscopic exposure enablestopographic location of the lateral disc center and allows visualization of the plane of inclinationfor each disc. (C) The inclination plane of each target disc is drawn on the skin from the lateraldisc center to the posterior skin surface. (D) The distance between the lateral disc center and theposterior skin surface plane is measured along each disc inclination line. (E,F) This distance isthen measured from the midline along the respective transverse plane line for each disc. At theend of this measure, a line parallel to midline is drawn to intersect each disc inclination line. Thisintersection marks the skin entry point of “skin window” for each target disc. Needle insertion atthis point toward the target disc at an angle of 25–30° to the surface skin plane will determine thepath of all subsequent instrumentation.

Once the optimal trajectory is established, the cannulas are inserted to allow for endo-scopic surgery under direct visualization.

Endoscopic spine surgery has a very high learning curve but is within the grasp ofevery endoscopic surgeon with proper training. As with any new procedure, the compli-cation rate is higher during the learning curve, and it may vary according to the skilland experience of each surgeon. The endoscopic technique, because of its approach,may pose additional risk for iatrogenic injury, but it is possibly safer than traditionalsurgery because the patient is awake and able to provide immediate input to the surgeonwhen pain is generated. Those surgeons who can master the technique to the extent thatthey prefer endoscopic surgery over traditional surgery for the same condition will havethe ability to perform the surgery without causing the patient undue pain. For most discherniations and discogenic pain, experienced endoscopic spine surgeons will opt for theendoscopic approach as the treatment of choice for their patients.

The future of endoscopic spine surgery is extremely bright. There will soon be anexplosion of new imaging systems, endoscopes, and endoscopic instruments. Refinedtechniques and image-guided systems may help diminish the learning curve. Coupledwith advancements in tissue regeneration and enhancement of tissue healing, and thetrend toward tissue healing instead of removal, regeneration over healing, and arthro-plasty instead of fusion, the spine surgeon may no longer have to consider spine surgeryas paradoxical. As a treatment modality, it will no longer be considered a last resort in adesperate patient. There will be a paradigm shift in the way clinicians view andapproach patients with back pain, especially when endoscopic spine surgery is furthervalidated with outcome studies and becomes routinely available.

REFERENCES

1. Kambin P. Arthroscopic microdiscectomy. Spine J 2003;.3(3 Suppl):60S–64S.2. Kambin P, Casey K, O’Brien E, Zhou L. Transforaminal arthroscopic decompression of

lateral recess stenosis. J Neurosurg 1996;84(3):462–467.

Selective Endoscopic Discectomy 223

Fig. 20. (Continued)

3. Kambin P, O’Brien E, Zhou L, Schaffer JL. Arthroscopic microdiscectomy and selectivefragmentectomy. Clin Orthop 1998;(347):150–67.

4. Kambin P, Savitz MH. Arthroscopic microdiscectomy: an alternative to open disc surgery.Mt Sinai J Med 2000;67(4):283–287.

5. Hijikata S. Percutaneous nucleotomy. A new concept technique and 12 years’ experience.Clin Orthop 1989(238):9–23.

6. Yeung AT. Spinal endoscopy with a multichannel, continuous irrigation, discoscope withintegrated inflow and outflow ports (poster presentation), in North American Spine SocietyAnnual Meeting, New York, 1997.

7. Yeung AT. Arthroscopic electro-thermal surgery for discogenic low back pain: a preliminaryreport, in International Intradiscal Therapy Society Annual Meeting, San Antonio, 1998.

8. Yeung AT. Classification and electro-thermal treatment of annular tears. in American BackSociety Annual Meeting December 12, Las Vegas, 1998.

9. Yeung AT. Minimal access surgery for degenerative conditions of the lumbar spine, in 3rdCongress Meeting of the Chinese Speaking Orthopedic Society and Annual Congress ofthe Chinese Orthopedic Association, Taipei, Taiwan, 1999.

10. Yeung AT. Minimally invasive disc surgery with the yeung endoscopic spine system (YESS).Surg Technol Int 2000;VIII:267–277.

11. Yeung AT. The evolution of percutaneous spinal endoscopy and discectomy: state of theart. Mt Sinai J Med 2000;67(4):327–332.

12. Yeung AT. Selective discectomy with the yeung endoscopic spine system, in The Practiceof Minimally Invasive Spinal Technique, (Savitz MH, Chiu J, Yeung AT, eds.); AAMISMSEducation LLC, New City, NY:2000, pp. 115–122.

13. Yeung AT. Patho-anatomy of discogenic pain, in Minimally Invasive Spine Update, 1999.Disney Magic Cruise Nov 5–8.

14. Yeung AT, Tsou PM. Posterolateral endoscopic excision for lumbar disc herniation: surgicaltechnique, outcome, and complications in 307 consecutive cases. Spine, 2002;27(7):722–731.

15. Yeung AT, and Yeung CA. Advances in endoscopic disc and spine surgery: foraminalapproach. Surg Technol Int 2003;11:253–261.

16. Tsou PM, and Yeung AT. Transforaminal endoscopic decompression for radiculopathy sec-ondary to intracanal noncontained lumbar disc herniations: outcome and technique. SpineJ 2002. 2(1):41–48.

17. Yeung AT. The value of an intra-operative discogram: its role in arthroscopic microdiscec-tomy, in International Intradiscal Therapy Society Annual Meeting and San FranciscoSpine Institute Course, 1995.

18. Yeung AT. The role of provocative discography in endoscopic disc surgery, in The Practiceof Minimally Invasive Spinal Technique, (Savitz MH, Chiu J, Yeung AT, eds.) AAMISMSEducation LLC, New City, NY:2000, pp. 231–236.

19. Carragee EJ, Tanner CM, Khurana S, et al. The rates of false-positive lumbar discography inselect patients without low back symptoms. Spine 2000;25(11):1373–1380; discussion, 1381.

20. Carragee, EJ,Tanner CM, Yang B, Brito JL, Truong T. False-positive findings on lumbardiscography: reliability of subjective concordance assessment during provocative disc injec-tion. Spine 1999;24(23):2542–2547.

21. Yeung AT. Enhancement of KTP/532 LDD and AMD procedures with a vital dye, in AAOSAnnual Meeting Scientific Exhibit, 1992.

22. Yeung AT. Considerations for use of the KTP laser for disc decompression and ablation, inSpine: State of the Art Reviews, 1993.

23. Yeung AT. Enhancement of KTP/532 LDD and AMD procedures with a vital dye, in Inter-national Intradiscal Therapy Society March 10–14, Phoenix, 1993.

24. Yeung AT. Enhancement of KTP/532 laser disc decompression and arthroscopic microdisce-tomy with a vital dye, in AAOS Audio Video Program, AAOS 61st Annual Meeting, 1994.

224 Yeung

25. Yeung AT. Laser as an adjunct to arthroscopic microdiscectomy, in North American SpineSociety Annual Meeting, New York, 1997.

26. Yeung AT, Gore SA. Evolving methodology in treating discogenic back pain by selectiveendoscopic discectomy (SED). J Minimally Invasive Spinal Tech 2001;1:8–16.

27. Yeung AT, Morrison PC, Felts MS, Carter JL. Intradiscal thermal therapy for discogeniclow back pain, in The Practice of Minimally Invasive Spinal Technique, (Savitz MH, ChiuJ, and Yeung AT eds.), AAMISMS Education LLC, New City, NY:2000.

28. Yeung AT. Macro-and micro-anatomy of degenerative conditions of the lumbar spine(best paper presentation award), in International Intradiscal Therapy Society 16th AnnualMeeting, 2003.

29. Yeung AT. Rauschning’s anatomy for minimally invasive spine surgery, in Spine Acrossthe Sea 2003, July 27–31, 2003.

30. Modic MT, Masaryk TJ, Ross JS, Carter JR. Imaging of degenerative disk disease. Radiol-ogy 1988;168(1):177–186.

31. Aprill C, Bogduk N. High-intensity zone: a diagnostic sign of painful lumbar disc on mag-netic resonance imaging. Br J Radiol 1992;65(773):361–369.

32. Aprill CN. Diagnostic Disc Injection, in The Adult Spine: Principles and Practice Fry-moyer JW, ed. Lippincott-Raven, Philadelphia, 1996, pp. 539–562.

33. Bini W, Yeung AT, Calatayud V, Chaabant A, Seferlis T. The role of provocative discogra-phy in minimally invasive selective endoscopic discectomy. Neurocirugia (Austr), 2002.13(1): p. 27–31; discussion, 32.

34. Guyer RD, Ohnmeiss DD. Lumbar discography. Spine J 2003;3(3 Suppl): p. 11S–27S.35. Guyer RD, Ohnmeiss DD. Lumbar discography: position statement from the North Ameri-

can Spine Society Diagnostic and Therapeutic Committee. Spine 1995;20(18): p. 2048–2059.36. Yeung AT. Evocative lumbar discography: treatment of multi-level cases, in 1st World

Congress American Academy of Minimally Invasive Spinal Medicine and Surgery, 2000.37. Yeung AT. Discography, foraminal epidurography and therapeutic foraminal injections: its

role in endoscopic spine surgery, in International 22nd Course for Percutaneous Endo-scopic Spinal Surgery and Complementary Techniques, 2004.

38. Yeung AT. Intra-operative chemonucleolysis as an adjunct to arthroscopic microdiscec-tomy, in International Intradiscal Therapy Society Annual Meeting, 1997.

39. Yeung AT, Porter J, Merican C. SEP as a sensory integrity check in selective endoscopicdiscectomy using the Yeung endoscopic spine system, in 2nd World Congress AmericanAcademy of Minimally Invasive Spinal Medicine and Surgery, 2001.

40. Yeung AT. Failed back surgery syndrome, in The Practice of Minimally Invasive SpinalTechnique, (Savitz MH, Chiu J, Yeung AT, eds.), AAMISMS Education LLC, 2000 pp.293–296.

Selective Endoscopic Discectomy 225

10Minimally Invasive Posterior Fusion

and Internal Fixation With the Atavi® System

Richard D. Guyer, MD and Terry P. Corbin, BS

INTRODUCTION

Although interest in minimally invasive posterior internal fixation and fusion hasrecently exploded, it is not because the technical developments are also recent. Thefoundations of these procedures were laid more than 60 yr ago, when the first endoscopeswere used to examine patients’ spinal nerves within the cauda equina (1). From thatbeginning, the evolution was slow until several technological advances facilitated asafe and effective procedure with a reasonable learning curve.

The major milestones in this evolution are provided in Table 1 and are discussed inthis chapter. The trend toward minimally invasive spinal surgery must be creditedfirst to Lyman Smith, the developer of chymopapain for chemonucleolysis (2).Although chemonucleolysis has been proven to have limitations, the initial interestgenerated has led to many other minimally invasive discectomy approaches, includ-ing mechanical techniques by Kambin (3) and Hijikata (4), automated percutaneousdiscectomy (5), and laser techniques (6,7). Dr. Parviz Kambin made numerous contri-butions to the advancement of this field. In particular, he defined the “triangularworking zone” for posterolateral approaches to the disc (8). He also first used sequen-tially larger dilators to minimize the trauma of introduction of larger cannulae formore sophisticated intradiscal and foraminal procedures (9,10). He developed a can-nula with a side port through which an endoscope could be inserted for visualizingtools and anatomical structures (Fig. 1), and he later used an endoscope with a fieldof view at a 70° angle to the long axis of the cannula to view the surgical field.

As the pioneers refined the mechanical discectomy techniques, especially withthe addition of endoscopic visualization (11), the possibility of interbody fusionbecame a reality. Kambin (9) developed a technique using a powered reamer todecorticate the end plate. He subsequently reported that the fusion rate was only57%, leading him to investigate the use of internal fixation to supplement interbodyfusion (12).

Several clinicians developed pedicle screw internal fixation approaches, followingthe early work of Magerl (13), who used external fixation via screws into the pedicles to

227

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

228 Guyer and Corbin

Table 1Milestones in Evolution of Minimally Invasive Posterior Fusion

Year Innovator Milestone

1938 Pool First report of endoscope use to visualize dorsal nerve roots1955 Ottolenghi, Craig Developed posterolateral approach for percutaneous

transpedicular biopsy of vertebral body1963 Smith Beginning of minimally invasive spine surgery trend as a

result of publication on chemonucleolysis1968 Wiltse Muscle-splitting approach1983 Kambin Muscle dilators used to prepare path for cannula1987 Goldthwaite, White Described pioneering efforts to decorticate transverse

processes percutaneously1987 Kambin Triangular working zone defined for posterolateral

approaches to the disc defined1991 Kambin Performs laminotomy through a cannula with endoscopic

visualization1991 Kambin Performs endoscopic-assisted lumbar interbody fusion with

percutaneous pedicle screw internal fixation1994 Goldthwaite et al. Patent application filed describing foraminotomy procedure

performed through cannula with endoscopic visualization1999 Knirk et al. Validates concept of pedicle screw placement through

Endius FlexPosure retractor in cadavers2001 Jahng et al. Confirms accurate pedicle screw placement possible

through FlexPosure in sheep2004 Winer et al. Describes patient benefit of Atavi technique for

posterolateral fusion with pedicle screw internal fixation

Fig. 1. Kambin cannula with endoscope in side port.

temporarily stabilize the spine while interbody graft consolidated. Leu et al. (14) alsoused external fixation, inserted percutaneously, to make the procedure less invasive.Kambin used subcutaneous plates to connect the pedicle screws; the internal fixationhardware was removed in a second procedure approx 8 mo after implantation. Hereported a 90% fusion rate using this construct (12). The Kambin instrument set is illus-trated in Fig. 2; a fluoroscopic image of the pedicle screws and further description ofthe Kambin procedure is included in Chapter 5.

Kambin achieved a good fusion rate with his technique, but there are limitations tothis approach. Regan and Guyer (15) described the significant learning curve associatedwith procedures based on the Kambin anthroscopic microdiscectomy (AMD) techniques.The Kambin fusion approach requires a second surgery, as well. To avoid the need for asecond surgery, Goldthwaite and White (16) used a different approach based on AMD:a percutaneous posterolateral fusion. They did not use internal fixation in their initialwork, speculating that fusion could be enhanced with osteoinductive factors to obviateinternal fixation.

Although it has been more than 16 yr since this speculation, growth factorssuch as rhBMP-2 are not yet available for use in posterolateral fusions. In1996, Boden et al. (17) confirmed the feasibility of posterolateral decortica-tion and graft placement through an endoscope working channel in an animalmodel for growth factor research. The initial clinical studies of rhBMP-2 forposterolateral fusion have been slowed by changes to the carrier needed tokeep the growth factor from leaking out of the fusion bed. Because no

Minimally Invasive Fusion and Fixation 229

Fig. 2. Kambin instrument set for percutaneous interbody fusion and pedicle screw internalfixation.

osteoinductive material is available to speed up and increase the likelihood ofposterior lumbar fusions, internal fixation is still an important part of thespine surgeon’s armamentarium.

THE ENDIUS ATAVI® MINIMALLY INVASIVE INTERNALFIXATION CONCEPT

Endius developed an adjunct to the Kambin minimally invasive fusion technique thatoffers significant benefits. Its paradigm was to improve the available working space andvisualization in the hope that this would reduce the learning curve associated witharthroscopic fusion. This was accomplished by designing the FlexPosure® retractor,which has an expanding skirt (Fig. 3). The original application was discectomy anddecompression, but the early users quickly steered the development effort towardfusion and internal fixation. Fusion using the FlexPosure retractor is termed an Ataviprocedure by its developers.

With the FlexPosure skirt expanded, there is sufficient space for a one- or two-levelpedicle screw placement. Recently, oval retractors have been introduced, with fixeddimensions of 24 × 30 mm at the top, expanding to 40 × 80 mm at the bottom of theskirt. These retractors are used with the components given in Table 2 to perform thecomplete array of one- and two-level spine procedures.

Knirk and Osuna (18) conducted the first test of the FlexPosure retractor for deliveryof internal fixation hardware. They demonstrated feasibility of accurate pedicle screwplacement in cadavers. This was followed by an in vivo demonstration of feasibility in

230 Guyer and Corbin

Fig. 3. FlexPosure retractor in position and expanded.

a sheep model by Jahng et al. (19). Only 6% of the screws in this study were misplaceddespite the small size of sheep pedicles.

Based on these results, in late 2000 several centers started a human study to eval-uate the procedure and equipment. Shortly thereafter, hemostasis was improved withthe addition of a bipolar sheath for the MDS™ microdebrider, which is used for soft-tissue removal. At the same time, a three-chip camera and a scope retractor mountwere introduced, improving the arthroscopic image. These improvements reducedthe incidence of procedures converted to open surgery owing to difficulties withvisualization.

The preliminary study report (20) demonstrated that the theoretical advantages ofminimally invasive fusion were achieved with this procedure. The average blood losswas only 280 mL, and the average hospital stay was 3.5 d. These are considerablylower than for comparable open surgery cases compiled by one of the investigational sites.In this early series, the longer-term outcomes were also good: the fusion rate was 88% andthe average improvement in the Oswestry Disability Index was from 52 to 23 at 3 mofollow-up. Similarly, the visual analog 11-point pain scale score dropped from an averageof 7.4 preop to 2.6 at 3 mo.

Most of these improvements were obtained in the first 6 wk, which again suggests thatthe theoretical benefits of minimally invasive fusion are being achieved. This rapidimprovement is attributed to less damage to paraspinal muscles during the procedure. Kimand Fox (21) noted that multifidus intramuscular pressure is 30–40 torr (mmHg) lower

Minimally Invasive Fusion and Fixation 231

Table 2Major Components of the Endius Atavi System

Description/use

Instrument/equipment

FlexPosure retractor Retractor with fixed dimension at top and expanding skirt at bottom

MDS microdebrider with Shaver with Bipolar outer sheath; used for soft-tissue removal Bipolar sheath (including nucleotomy) with simultaneous hemostasis

Flex Arm® Holds FlexPosure retractor and endoscope; easilyrepositioned with vacuum release button

Endoscope/camera Endoscope features 30° view; high-resolution three-chip camera

Decompression instruments Family of instruments optimized for use in minimally invasive decompression; includes angled Kerrisons, osteotomes, curettes, probes

Internal fixation implants

Wave frame plate Simple titanium plate-pedicle screw system for single-level fixation system posterolateral fusions

TiTLE rod fixation system State-of-the-art titanium top-loading rod and pedicle screw system for one- and two-level fusions; unique features incorporated for MIS approach

MiCOR precision Used in MiLIF™ lumbar interbody fusion procedurebone allograft

in Atavi fusions than in open cases. This results in muscle that looks healthier at closureand has less atrophy and edema (22). The net is less postoperative pain and better functionfor the patient. Preserving the multifidus is particularly important because it contributesmore than two-thirds of the stiffness of the lumbar spine—resistance to flexion/extension,lateral bending, and rotation (23).

The original internal fixation system used in this feasibility study was called the DiamondPlate. This construct required assembly of two screws, four washers, one plate, and two nutswithin the 2l × 35 mm operative field in the retractor. The Wave® Frame was introduced tosimplify the technique; this plate has integral washers for easier assembly (Fig. 4). The Flex-Posure retractor was recently expanded into a family with several new members:

• A 2l-mm-diameter FlexPosure that expands to 25 × 40 mm.• A 24-mm-diameter FlexPosure that expands to 30 × 63 mm, sufficient for a two-level

posterolateral fusion with pedicle screw internal fixation.• Two oval FlexPosure retractors with top dimensions of 24 × 30 mm, opening at the bottom to

40 × 50 or 40 × 80 mm, sufficient for a three-level fusion.

During the study any necessary spinal stenosis decompression was performed througha small midline incision. Subsequently, Hartjen et al. (24) have developed instrumentsand techniques for decompression through the FlexPosure retractor, eliminating the needfor an additional incision. Endius has introduced a new visualization method by providinga mount for the FlexPosure that has an integrated surgical light. In this mode, procedurescan be conducted under direct vision through loupes or an operating microscope (Fig. 5). Thisfeature is particularly useful for decompression around the nerve roots. The depth percep-tion afforded by the binocular vision adds to procedure safety. For the new user, the tran-sition from an open technique to Atavi with direct vision is very easy.

232 Guyer and Corbin

Fig. 4. Wave frame internal fixation system.

Two other improvements have been made to the internal fixation capabilities of theAtavi procedure. In September 2002, Endius introduced the TiTLE titanium rod andscrew system (Fig. 5). This is a low-profile top-loading system optimized for assemblyin minimally invasive fusion techniques. A notable feature is the friction in the multiaxialtulip; the tulip stays in position so that rod placement is easier. This system includes acomplete set of instruments for distraction and compression. Whereas the Wave Frameplate was only available for single-level fusions, the TiTLE system can be used to makeconstructs up to three levels with pedicle screws at each level.

The most recent addition to the Atavi line is the MiCOR™ allograft block. Thesecrescent-shaped implants are useful for minimally invasive interbody fusion procedures.A number of traditional posterior or lateral approaches to the interbody space can beaccomplished through the FlexPosure retractor with the MiCOR graft and TiTLE internalfixation hardware (22).

The Atavi procedure incorporates the muscle-splitting approach to the lumbar spinefirst espoused by Wiltse et al. (25). With the refinements introduced over the last 4 yr,all of the procedures described in Wiltse and Spencer’s subsequent review (26) can beperformed: decompression of lateral herniations and spinal stenosis, fusion, and pediclescrew internal fixation. Because the Atavi procedure is useful in most indications forlumbar spine surgery, the new user can overcome the learning curve and stay proficientrelatively easily. With the addition of the oval FlexPosure retractor, the learning curveis markedly diminished (Fig. 6).

ATAVI PROCEDURE COMPARED TO EMERGING MINIMALLYINVASIVE FUSION/INTERNAL FIXATION TECHNIQUES

In addition to the Atavi procedure, several other approaches to minimally invasiveposterior fusion with internal fixation have been described in the literature. In 2000,Muller et al. (27) described a “keyhole” approach for endoscopically assisted internal

Minimally Invasive Fusion and Fixation 233

Fig. 5. TiTLE rod and pedicle screw system, endoscope view.

fixation. They used ports developed for thoracoscopic surgery for introducing the pedi-cle screws. The screws were connected by rods bluntly tunneled between the screws.The initial results in cadavers and a small number of patients demonstrated feasibilityand screw placement accuracy of 93%. A limitation of the equipment used was the rela-tive lack of room in the port when the endoscope was in position; the endoscope wasremoved when the screws were inserted. The ports must be removed to pass the rodbetween the screws, a blind procedure.

Foley et al. (28) described an internal fixation technique called the Sextant procedureby its manufacturer, Medtronic Sofamor Danek. This equipment is used to percutaneouslyinsert pedicle screws and accurately connect them with rods. The primary application isinternal fixation to supplement an anterior lumbar interbody fusion or posterior lumbarinterbody fusion; there is no provision for posterolateral fusion. Like the Muller tech-nique, the Sextant procedure involves blind tunneling for the rod passage with no provi-sion for hemostasis in the muscle. A total of six small incisions are needed for a one-levelfusion. Internal fixation for a two-level fusion can be provided by placing screws in theend vertebrae, skipping the intervening vertebra.

Spinal Concepts recently introduced the Pathfinder system for minimally invasiveposterior internal fixation (Fig. 7). The system is similar to the Sextant approach in thatit allows accurate blind placement of the rod through the pedicle screws. Rather thanthe multiple small incisions for Sextant patients, Pathfinder patients have one larger inci-sion over the pedicles on each side. The surgical field is limited with the Pathfinder tech-nique, similar to working through the early Kambin straight cannula.

Both the Sextant and the Pathfinder techniques depend heavily on intraoperative flu-oroscopy in contrast to the Atavi approach, which allows placement of pedicle screwsthrough either direct or endoscopic visualization.

CONCLUSION

The Endius Atavi system has incorporated the advantages of the Wiltse muscle-sparingparaspinal approach to the lumbar spine and the Kambin arthroscopic technique with a

234 Guyer and Corbin

Fig. 6. Oval FlexPosure retractor with direct vision.

Minimally Invasive Fusion and Fixation 235

Fig. 7. Spinal Concepts Pathfinder equipment for internal fixation.

unique expanding retractor as the foundation for a family of products for minimallyinvasive lumbar spinal surgery. With the addition of the oval FlexPosure retractor, thelearning curve is markedly reduced, and the larger surgical field allows applicationof similar techniques to open surgery. Although the equipment is new and the clinicalfollow-up is relatively short, there appears to be significant benefit for the patient interms of reduced rehabilitation time and postoperative pain. The system can be used formost lumbar spine surgical procedures, which justifies the investment of time in learningthe nuances.

REFERENCES

1. Pool JL. Myeloscopy: diagnostic inspection of the cauda equina by means of the endoscope.Bull Neurol Inst NY 1938;7:178–189.

2. Smith L, Garvin PJ, Gesler RM, Jennings RB. Enzyme dissolution of the nucleus pulposus,Nature 1963;198:1311–1312.

3 Kambin P, Gellman H. Percutaneous lateral discectomy of the lumbar spine: a preliminaryreport. Clin Orthop 1983;174:127–l32.

4. Hijikata S. Percutaneous discectomy: a new treatment method for lumbar disk herniation.J Toden Hosp l975;5:5–13.

5. Onik G, Helms C, Ginsburg L, et al. Percutaneous lumbar discectomy using a new aspirationprobe. AJR 1985;144:1137–1140.

6. Ascher PW, Holzer P. Laser denaturation of the nucleus pulposus of herniated intervertebraldiscs, Arthroscopic Microdiscectomy: Minimal Intervention in Spinal Surgery (Kambin P,ed.), Urban & Schwarzenberg, Baltimore, 1991, pp. 137–140.

7. Sherk HH, Black J, Rhodes A, et al. Laser discectomy. Clin Sports Med 1993;12:569–577.8. Kambin P, Brager M. Percutaneous posterolateral discectomy: anatomy and mechanism.

Clin Orthop 1987;223:145–154.9. Kambin P. Posterolateral percutaneous lumbar interbody fusion, Arthroscopic Microdiscec-

tomy: Minimal Intervention in Spinal Surgery (Kambin P, ed.), Urban & Schwarzenberg,Baltimore, 1991, pp. l17–121.

10. Kambin P. Endoscopic spinal surgery past and future. Video presentation of endoscopiclaminotomy and foraminotomy. Presented at the annual meeting of the International Societyfor Minimally Invasive Spinal Surgery, January 1991, Zurich, Switzerland.

11. Schreiber A, Suezawa Y. Transdiscoscopic percutaneous nucleotomy in disc herniation.Orthop Rev 1986;15:35–38.

12. Kambin P. Arthroscopic lumbar intervertebral fusion, in The Adult Spine: Principles andPractice, 2nd ed. (Frymoyer JW, Ducker T, Hadler N, et al., eds.), Lippincott-Raven, NewYork, 1997, pp. 2037–2047.

13. Magerl F. Stabilization of the lower thoracic and lumbar spine with external skeletal fixation.Clin Orthop 1984;189:125–141.

14. Leu H, Hauser R, Schreiber A. Percutaneous lumbar spine fusion. Acta Orthop Scand Suppl1993;251:116–119.

15. Regan JJ, Guyer RD. Endoscopic techniques in spinal surgery. Clin Orthop 1997;335:122–139.

16. Goldthwaite N, White AH. Toward percutaneous spine fusion, in Lumbar Spine Surgery:Technique & Complications, Mosby, St. Louis, 1987, pp. 512–522.

17. Boden SD, Moskovitz PA, Morone MA, Toribitake Y. Video-assisted lateral intertransverseprocess arthrodesis: validation of a new minimally invasive lumbar spinal fusion techniquein the rabbit and nonhuman primate (rhesus) models. Spine 1996;21:2689–2697.

18. Knirk J, Osuna R. Endoscopically assisted posterolateral fusion with pedicle screwinstrumentation: feasibility study. Paper presented at the North American Spine Society,Chicago, 1999.

19. Jahng TA, Fu TS, Cunningham BW, Dmitriev AE, Kim DH. Endoscopic instrumentedposterolateral lumbar fusion with Healos and recombinant human growth/differentiationfactor-5. Neurosurg 2004;54:17l–l80.

20. Winer M, Goodwin C, Hartjen C, Davison, T. Minimally invasive spine surgery with pediclescrew internal fixation; the Endius Atavi™ system, in Emerging Spine Surgery Technologies(Corbin TP, Connolly PJ, Boden SD, Yuan HA, Bao Q-B, eds.), Quality Medical Publishing,St. Louis, in press.

21. Kim KD, Fox A. The Atavi™ approach to minimally invasive spine surgery, in SurgicalTechniques in Spinal Instrumentation (Kim DH, Vaccaro AR, Fessler RG, eds.), Thieme,New York, in press.

236 Guyer and Corbin

22. Winer MH, Alamin TF. Minimally invasive lumbar interbody fusion with pedicle screwinternal fixation: the Endius Atavi™ system in The Practice of Minimally Invasive SpinalTechnique, 2nd (Savitz MH, ed.), AAMISMS Education LLC, Richmond, VA, in press.

23. Wilke HJ, Wolf S, Claes LE, Arand M, Wiesend A. Stability increase of the lumbar spinewith different muscle groups: a biomechanical in vitro study. Spine 1995;20:192–198.

24. Hartjen C, Martin R, Sweeney T. Technique for Decompression During EndoscopicPosterolateral Fusion Surgery, Endius, 2002.

25. Wiltse LL, Bateman JG, Hutchinson RH, Nelson WE. The paraspinal sacrospinalis-splittingapproach to the lumbar spine. J Bone Joint Surg Am 1968;50:919–926.

26. Wiltse LL, Spencer CW. New uses and refinements of the paraspinal approach to the lumbarspine. Spine 1988;13:696–706.

27. Muller A, Gall C, Marz U, Reulen HJ. A keyhole approach for endoscopically assistedpedicle screw fixation in lumbar spine instability. Neurosurg 2000;47:85–95.

28. Foley KT, Holly LT, Schwender JD. Minimally invasive lumbar fusion. Spine 2003;28:S26–S35.

Minimally Invasive Fusion and Fixation 237

11Vertebral Augmentation

for Osteoporotic Compression Fractures

Daisuke Togawa, MD, PhD

and Isador H. Lieberman, MD, MBA, FRCS(C)

INTRODUCTION

Osteoporosis is a systemic disease currently afflicting approx 44 million Americans;this figure will increase as the population ages. It results in progressive bone mineralloss and concurrent changes in bony architecture that leave bone vulnerable to fracture,often after minimal or no trauma. The spine is the most common site of osteoporoticfracture, with vertebral compression fracture (VCF) occurring in 20% of people over theage of 70 yr, and up to 50% of women 80 yr and older (1,2). Overall, 700,000 people peryear in the United States suffer a VCF, exceeding even the frequency of hip fractures (3).Osteoporotic VCFs have been shown to be associated with up to a 30% age-adjustedincrease in mortality (4). The cost to society of managing osteoporotic VCF patients inthe United States in 1995 was $746 million (5). Possible acute complications of vertebralfracture include cord compression, urinary retention, and ileus (6). Long-term consequencesinclude considerable pain (reported in 35% of detectable VCFs) (7) as well as pulmonarycompromise (a 9% loss in predicted forced vital capacity with each vertebral fracture)(8). Other chronic sequelae include deconditioning, deformity, insomnia, and depression,resulting in substantial physical, functional, and psychosocial impairment (8,9).

Nonoperative Management of VCFs

Two-thirds of patients with acute, painful VCFs experience pain improvement regard-less of the treatment applied. Traditional, nonoperative management includes bed rest,analgesics, and bracing. This type of medical management, however, fails to restore spinalalignment, and the lack of mobility itself can result in secondary complications, includ-ing worsening osteoporosis, atelectasis, pneumonia, deep vein thrombosis, decubitus ulcer,and pulmonary embolism. An alternative approach is supervised ambulatory mobilityby a physiotherapist plus hydrotherapy (10). In one-third of patients, severe pain, limitedmobility, and poor quality of life persist despite appropriate nonoperative management.Whether the pain has resolved or not, no patient after a VCF spontaneously achieves arealigned spine, corrected sagittal contour, or restoration of vertebral height.

239

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

Operative Management of VCFs

Historically, the only alternative to nonoperative management of symptomatic vertebralfractures was open surgical decompression (anterior or posterior decompression and stabi-lization via internal fixation hardware and bone grafting), and this was usually reserved forthose patients with gross spinal deformity or neurological impairment (<0.5%) (6). Thissurgical caution came about because of the adverse risk/benefit ratio in this elderly popu-lation with poor bone quality and multiple comorbid conditions.

Percutaneous vertebroplasty (PVP) is a minimally invasive method that involves thepercutaneous injection of polymethyl methacrylate (PMMA) into a collapsed vertebralbody to stabilize the vertebra. Originally developed for osteolytic metastasis, myeloma, andhemangioma, the procedure resulted in quick, effective pain relief and a low complicationrate (11–13). PVP is now also increasingly used for the treatment of osteoporotic vertebralfractures (9). However, PVP does not expand the collapsed vertebra, potentially lockingthe spine in a kyphotic posture. In addition, the PMMA bone filler has associated problems(epidural leakage, thermal necrosis, inability to integrate with bone, handling difficulties,toxicity to patient and operator) (2,14).

Kyphoplasty is an advanced minimally invasive technique with a number of poten-tial advantages over PVP, including lower risk of cement extravasation and betterrestoration of vertebral body height (15). A cannula is introduced into the vertebralbody, followed by insertion of an inflatable bone tamp, which, when deployed, reducesthe compression fracture and restores the vertebral body toward its original height, whilecreating a cavity to be filled with bone cement. The cement augmentation is thereforedone with more control into the low-pressure environment of the preformed cavity withviscous, partially cured cement.

PERCUTANEOUS VERTEBROPLASTY

Background

Percutaneous vertebral augmentation (vertebroplasty, or PVP) was first reported byGalibert and colleagues in 1984 and initially involved augmentation of the vertebralbody with PMMA to treat a hemangioma. PVP was reportedly not performed in theUnited States until 1994. Originally targeted for osteolytic metastasis, myeloma, andhemangioma, PVP resulted in early appreciable pain relief and a low complication rate(13,16). Its indications subsequently expanded to osteoporotic vertebral collapse withchronic pain, and then further to include treatment of asymptomatic vertebral collapseand even prophylactic intervention for at-risk vertebral bodies (17). Nevertheless, thetreatment of acute fractures in ambulatory patients and prophylactic treatment remaincontroversial (18). In fact, vertebral augmentation itself is somewhat controversial, withquestions concerning a lack of defined indications, expected complications, outcome mea-sures, and the need for long-term follow-up data (2).

An open question in PVP is the mechanism of pain relief. The most intuitive expla-nation involves simple mechanical stabilization of the fracture. However, another pos-sibility is the analgesic result from local chemical, vascular, or thermal effects ofPMMA on nerve endings in surrounding tissue (9,19). Supporting this concept is thelack of correlation between cement volume and pain relief (20,21). Further evidenceagainst an effect resulting solely from mechanical stabilization is the fact that PVP

240 Togawa and Lieberman

typically does not restore lost vertebral body height and therefore does not correctaltered biomechanics (1,18).

Technique

Injection of opacified PMMA is performed via a transpedicular or paravertebralapproach under continuous fluoroscopic guidance to obtain adequate filling and to avoidPMMA leakage. For complex or high-risk cases, computed tomography (CT) and fluoro-scopic guidance are sometimes combined (11,18). In routine cases, PVP can be performedunder local anesthesia with slight sedation in less than 1 h (1), although general anesthesiais sometimes required because pain may intensify during cement injection (9). PrecedingPMMA injection, intraosseous venography is often used to determine the filling patternand identify sites of potential PMMA leakage (outline the venous drainage pattern, confirmneedle placement within the bony trabeculae, and delineate fractures in the bony cortex).However, some clinicians have dispensed with routine venography (1).

Contraindications to vertebroplasty include coagulopathy, absence of facilities to performemergency decompressive surgery in the event of a complication, and extreme vertebralcollapse (>65–70% reduction in vertebral height) (9).

Results

From 1985 to March 2004, 329 articles on PVP were published in peer-reviewedjournals. Vertebroplasty data from more than 1000 patients have been reported in sev-eral case series (1,11,13,18,22–34). The longest reported follow-up is 3 yr, althoughthe first article on vertebroplasty was published in 1987. Reportedly, pain has beenreduced in 70–90% of patients. There have been no reported cement failures, andonly two reported cases of fracture progression in the treated vertebral bodies(caused by inadequate cement fill). Barr et al. (18) reported the results of 47 patients treatedwith vertebroplasty with an average follow-up of 18 mo. Their article outlines marked tocomplete pain relief in only 63% of patients with osteoporotic VCFs. Vertebroplasty, how-ever, does not address the spinal deformity. In addition, this technique requires ahigh-pressure cement injection using low-viscosity cement, thus increasing the riskof cement leaks through the fracture clefts or the venous sinuses. Evans et al. (26)reported their retrospective results of 245 cases with an average follow-up of 7 mo. Inthier study, pain score was significantly decreased and the ability to participate in anactive daily lifestyle was significantly improved following vertebroplasty (26).

Complications

The principal risk of PVP, which involves the forced injection of low-viscosity PMMAcement into the closed space of the collapsed vertebral body, is cement extravasation.Extravasation rates are as high as 65% when used to treat osteoporotic fractures (13,28).The likelihood is greater when using cement with a liquid rather than paste consistency,or with higher PMMA volume (25). However, in most settings, the majority of extrava-sations have no clinical relevance, at least in the short term (1).

The consequence of an extravasation depends on its location. In epidural or foraminalextravasation, nerve root compression and radiculopathy is the major risk. Thisoccurred in 11 of 274 patients (4%) treated by Deramond et al. (11). Three of thosepatients required surgical nerve root decompression. Other clinicians have described a

Vertebral Augmentation 241

5% rate of radiculopathy as well (6,20,35). Extravasation into perivertebral veins cancause cement embolism to the lungs; deaths attributed to cement embolism have beendocumented. However, two reported deaths attributed to pulmonary embolism were feltto be unrelated to the procedure; no cement material was detected by chest X-ray ofthe first patient (9,36), and the second pulmonary embolism arose from deep venouslower-extremity thrombosis (11). On the other hand, extravasation into adjacent disksor paravertebral tissue, although common, generally produces no symptoms and car-ries little clinical significance; many such extravasations can be avoided by carefulneedle positioning (11).

Other operative and long-term complications of PVP are specific to PMMA as a filler(1,17,37). The physician may work with PMMA in large batches in order to keep it liq-uid and to extend the working time for vertebroplasty. However, its high polymerizationtemperature (86–107°C within cement core) (38) can damage adjacent tissue, includingthe spinal cord and nerve roots (14), leading to an inflammatory reaction and transitoryexacerbation of pain (9). When injecting PMMA monomer, physician vigilance and cau-tion is required. Absorption of PMMA monomer during the injection can inducehypotension by virtue of its cardiotoxic and arrythmogenic properties (39). Placing amaterial in the spine affords proximity and access to the chest and the heart. Therefore,vertebral augmentation with PMMA demands meticulous attention to technique.

Overall, the risk of complications that carry clinical significance following PVP forosteoporotic vertebral fracture is felt to be 1–3%, and most potential complications canbe avoided with good technique (11).

KYPHOPLASTY

Background

Kyphoplasty is an advanced surgical technique that has evolved from a marriage of ver-tebroplasty with balloon angioplasty. It has a number of potential advantages, includinglower risk of cement extravasation and better restoration of vertebral body height. A can-nula is introduced into the vertebral body, via a transpedicular or extrapedicular route,followed by insertion of an inflatable bone tamp, which, when deployed, reduces the com-pression fracture and restores the vertebral body toward its original height. This then cre-ates a cavity to be filled with bone cement. The cement augmentation can now becompleted with more control into the low-pressure environment of the preformed cavitywith viscous, partially cured cement. Using a cannula for bone filler with a steel stylet as aplunger enables the operator to apply cement at considerably higher viscosity than is possi-ble with injection through a 5-cc syringe and an 11-gage needle. Both the higher cementviscosity and controlled fill reduce the risk of cement extravasation. Filling is performedunder continuous lateral fluoroscopic guidance similar to vertebroplasty. The procedurecan be performed under general anesthesia or local with intravenous sedation; mostpatients are able to return home the same day as the procedure.

Technique

With the patient under general or local anesthesia in prone position on a radiolu-cent spinal frame (Fig. 1A), two C-arms are positioned for anteroposterior and lateralfluoroscopic images (Fig. 1B). Once positioned, the C-arms and patient are not moved,to ensure repeatable images throughout the case. Two 3-mm incisions are made at the

242 Togawa and Lieberman

Vertebral Augmentation 243

Fig. 1. (A) Patient and (B) operation room setup.

vertebral level parallel to the pedicles in both planes. Then a guide wire or biopsy nee-dle is advanced into the vertebral body via a transpedicular or extrapedicularapproach, depending on the fracture configuration and the patient’s anatomy. Theguide wire is exchanged for the working cannula using a series of obturators. Oncethe working cannula is positioned, the surgeon reams out a corridor to accommodatethe inflatable bone tamp (IBT) and positions the IBT under the collapsed end plate. Todeploy the IBT, inflation proceeds slowly under fluoroscopy until maximum fracturereduction is achieved or the balloon reaches a cortical wall (Fig. 2). At this point thesurgeon deflates and removes the IBT, mixes the cement, prefills the cement cannulae,

and allows the cement to partially cure in the cement cannulae. Once partially cured,PMMA is slowly extruded into the vertebral body through each pedicle under continu-ous lateral fluoroscopic guidance (Fig. 3). This technique permits a controlled fill. Inmost instances, the volume of cement can slightly exceed that of the bone cavity tointerdigitate filler from the central bolus with the surrounding bone. Once filling iscomplete and the cement has hardened, the surgeon removes the cannula and closesthe 3-mm incisions.

Results

In our ongoing Institutional Review Board-approved study (15,40,41), more than900 consecutive kyphoplasty procedures were performed in more than 300 patientsbetween April 1999 and February 2004. The mean age was 69 yr (range: 35–89 yr). Themean duration of symptoms was 7 mo. Outcome data were obtained by administeringthe Short Form-36 health survey (SF-36), and visual analog scale (VAS) for pain rating.

244 Togawa and Lieberman

Fig. 3. Cement deposition.

Fig. 2. Fracture reduction using IBT.

Additionally, the patients underwent detailed neurological and radiographical examina-tions pre- and postoperatively. Perioperative and clinical follow-up revealed that theprocedure was well tolerated, with improvement in pain and early mobilization. Thelevels treated ranged from T3 to L5 with 47% of the vertebrae at the thoracolumbarjunction. Length of stay ranged from 0.5 to 9 d (mean: 1.1 d). In our experience, noclinically significant cement leaks and no perioperative complications were attributableto the IBT or tools. Pre- and postoperative SF-36 data are available on more than 230(72%) patients with follow-up ranging from 1 wk to 59 mo (mean: 14 mo). SF-36scores improved in every category, statistically significant in all but the general healthmodality (Fig. 4). Physical function improved from 22.0 to 36.0 (p ≤ 0.0001). Rolephysical improved from 9.3 to 27.3 (p ≤ 0.0001). Bodily pain improved from 22.4 to41.9 (p ≤ 0.0001). Vitality improved from 31.4 to 40.7 (p ≤ 0.0001). Social functionimproved from 37.7 to 61.2 (p ≤ 0.0001). Role emotional improved from 54.8 to 65.5(p = 0.030). Mental health improved from 63.1 to 68.0 (p < 0.001). General health didnot change significantly, with a score from 51.3 to 49.2 (p = 0.067). The VAS scoresimproved from a preoperative level of 7.0 to an initial postoperative level of 3.2 (p <0.0001). At last follow-up examination, the value had not changed significantly, with ascore of 3.4 (p < 0.0001).

Ledlie and Renfro (42) reported functional and radiographic outcomes in the first 96kyphoplasty patients with 133 fractures. Their follow-up period was a minimum of 12 mo,and the mean patient age at the time of surgery was 76 yr (51–93 yr). Regarding pain asrated by the patient using a 10-point VAS, the mean score was decreased to 1.4 at the 1-yrfollow-up, whereas the mean preoperative VAS score was 8.6. Ambulatory status was

Vertebral Augmentation 245

Fig. 4. SF-36 scores from ongoing Institutional Review Board-approved study. VAS, visual ana-log pain score; Osw, Oswestry disability index; PF, physical function; RP, role physical; BP, bodilypain; GH, general health; V, vitality; SF, social function; RE, role emotional; MH, mental health.

also improved postoperatively. More than 90% (27/29, with 1-yr follow-up) of thepatients were ambulatory at 1 yr, whereas only 35% (28/79) of the patients were ambu-latory preoperatively.

Phillips et al. (43) also recently reported their early radiographic and clinical resultsof kyphoplasty. In their study, 29 patients with 61 fractures between T6 and L5 wereevaluated. The mean age of the patients was 70 yr. Their clinical information includingpain relief, improvement in activity, and satisfaction with the surgical procedur, as wellas their sagittal spinal alignment on the standing radiographs, was assessed and followedup to 1 yr. Average pain scores were significantly decreased to 2.6 and 0.6 at 1 wk and 1 yr,respectively, whereas average pain score was 8.6 preoperatively.

In addition to good clinical results, height restoration by kyphoplasty has beenreported in several studies. Our initial results showed height restoration in 70% of 70fractured vertebrae treated with kyphoplasty. In patients in whom the vertebral fractureswere reduced by kyphoplasty, vertebral height was increased by a mean of 46.8%.

Garfin et al. (44) reported in a prospective multicenter series that the average anteriorand midline height were 83 ± 14 and 76 ± 14% before treatment, respectively, but wereincreased to 99 ± 13 and 92 ± 11% after treatment, respectively. In vertebral bodieswith 15% or more of the estimated height lost, the average anterior and midline heightwere 68 ± 12 and 64 ± 13% before treatment, respectively, but improved to 84 ± 14 and90 ± 12% after treatment, respectively (44).

Ledlie and Renfro (42) reported from radiographic measures anterior and midline pointsof the fractured vertebrae using the two nearest normal vertebrae as reference points. At1 yr, the anterior vertebral height was 85% of the predicted height and midline height was89%, whereas their preoperative heights were 66 and 65%, respectively.

Phillips et al. (43) also reported that local kyphosis improved by a mean of 14° inpatients with reducible fractures.

Complications

In our series of patients (15,40), cement extravasation was seen in <10% of cases.No problems were identified clinically as a result of these extravasations immediatelyafter surgery or at final follow-up. In one patient, a myocardial infarction occurred as aresult of fluid overload during the procedure.

In a separate prospective multicenter series reported by Garfin and Reilley (45), therewere six major complications out of 600 cases associated with the kyphoplasty procedure.Four of these complications (0.75%) were neurological (45). These were directly attributableto surgeon error and breach of technique.

To date, no reports of primary or secondary infection of the cement mantle have beenpublished. In our series of more than 300 patients, we had no primary infections. We did,however, encounter one hematogenous infection 2 yr after the kyphoplasty in a patientreceiving multiple blood and platelet transfusions for Waldenström’s macroglobulinemia.

Ledlie and Renfro (42) reported that asymptomatic cement leaks were noted in 9% ofvertebral bodies treated, but no device- or procedure-related complications were reported.

Phillips et al. (43) reported that asymptomatic cement leaks were observed in 6 of 61(9.8%) vertebral fractures. In this series as well, there were no clinical consequencesattributable to the bone tamp or cement deposition.

246 Togawa and Lieberman

VERTEBROPLASTY vs KYPHOPLASTY

Although both vertebroplasty and kyphoplasty provide excellent pain relief, kypho-plasty has the potential to improve spine biomechanics and decrease the risk of cementextravasation. PVP usually will not expand the vertebral body or regain normal spinealignment. Hiwatashi et al. (46) reported an increase in vertebral body height after ver-tebroplasty to measure the vertebral heights in preoperative magnetic resonance imag-ing (MRI) and postoperative CTs. The heights of 85 vertebral bodies in 37 patientswere measured before and after vertebroplasty in this study. The results showed anaverage increase in vertebral body height of 2.5 mm anteriorly, 2.7 mm centrally, and1.4 mm posteriorly. However, the investigators did not distinguish height correctionsfrom the positioning, and it is still unclear how much was corrected by the procedureitself. In addition, the significance of this methodology to measure the height betweenMRI and CT is uncertain. Preliminary data indicate that kyphoplasty may restore near-normal height, preventing kyphosis that leads to respiratory and digestive problems.Restoration of height and sagittal alignment may also work to protect vulnerable vertebrallevels above or below the site(s) treated by minimizing force transfer.

The vertebroplasty technique is much more prone to cement leaks than kyphoplasty,because the PMMA is injected in a liquid state and will take the path of least resistancethrough any cracks in surrounding bone. In administering vertebroplasty, the operatorinjects the liquid cement, typically pausing or stopping once a leak becomes evident. Onthe other hand, in kyphoplasty, the expanded balloon creates a cavity and pushes bone tothe edges of the cavity, thus sealing off potential fissures and cracks. Greater placementcontrol is possible in a kyphoplasty, in which the operator can fill the cavity with a moreviscous cement to the point at which the cement bolus reaches and interdigitates with thebony margins. The initial kyphoplasty findings show lower rates of cement extravasationcompared with published results of vertebroplasty series, supporting the hypothesis thatfilling with high-viscosity cement into a previously formed cavity may be an improvementover the injection of low-viscosity liquid cement into the unreduced vertebral body.

CONCLUSION

Osteoporotic VCFs pose a significant clinical problem including spinal deformity,pain, reduced pulmonary function and mobility, as well as an overall increase in mortalityin the elderly. Traditional medical and surgical options in many cases prove inadequate.

PVP is a relatively noninvasive technique that has gained increased acceptance overthe last decade in the treatment of symptomatic osteoporotic vertebral fractures. Theavailable clinical studies describe pain relief achieved in >90% of symptomatic osteo-porotic fractures, with only infrequent, mostly minor, complications. Some of the draw-backs of PVP stem from the use of PMMA, because of its toxicity and poor handlingcharacteristics, rather than from the procedure itself.

Kyphoplasty is a modification of PVP that may add a margin of safety by virtue of alower observed incidence of cement leakage. Kyphoplasty has been shown to be worthwhilein acute vertebral fractures to predictably restore vertebral height and to facilitate a con-trolled fill of the vertebral body. Favorable outcomes in early trials appear to implythat kyphoplasty permits early mobilization, which has the potential to decrease mortality.

Vertebral Augmentation 247

Considering the greater mortality that is associated with osteoporotic compression fractures,early mobilization in patients with vertebral fractures is of prime importance.

The next logical step beyond treatment of evident vertebral fractures is prophylacticaugmentation. Prevention of osteoporotic vertebral fractures with a combination of phar-macologics and timely reinforcement of at-risk osteoporotic vertebrae is the ultimate goalaside from prevention of osteoporosis itself. It is here that new osteoconductive syntheticcomposites will figure more prominently as an emerging alternative to cement. Advancesin minimally invasive surgical techniques, imaging, and synthetic engineering are rapidlychanging the treatment protocols available for osteoporotic compression fracture.

REFERENCES

1. Heini PF, Walchli B, Berlemann U. Percutaneous transpedicular vertebroplasty with PMMA:operative technique and early results. A prospective study for the treatment of osteoporoticcompression fractures. Eur Spine J 2000;9:445–450.

2. Einhorn TA. Vertebroplasty: an opportunity to do something really good for patients. Spine2000;25:1051, 1052.

3. Riggs BL, Melton LJ 3rd. The worldwide problem of osteoporosis: insights afforded byepidemiology. Bone 1995;17:505S–511S.

4. Kado DM, Browner WS, Palermo L, Nevitt MC, Genant HK, Cummings SR. Vertebralfractures and mortality in older women–a prospective study: Study of Osteoporotic Frac-tures Research Group. Arch Intern Med 1999;159:1215–1220.

5. Ray NF, Chan JK, Thamer M, Melton LJ 3rd. Medical expenditures for the treatment ofosteoporotic fractures in the United States in 1995: report from the national osteoporosisfoundation. J Bone Miner Res 1997;12:24–35.

6. Bostrom MP, Lane JM. Future directions: augmentation of osteoporotic vertebral bodies. Spine1997;22:38S–42S.

7. Cooper C, Atkinson EJ, O’Fallon WM, Melton LJ 3rd. Incidence of clinically diagnosedvertebral fractures: a population-based study in Rochester, Minnesota, 1985–1989. J BoneMiner Res 1992;7:221–227.

8. Leech JA, Dulberg C, Kellie S, Pattee L, Gay J. Relationship of lung function to severity ofosteoporosis in women. Am Rev Respir Dis 1990;141:68–71.

9. Cotten A, Boutry N, Cortet B, Assaker R, Demondion X, Leblond D, Chastanet P,Duquesnoy B, Deramond H. Percutaneous vertebroplasty: state of the art. Radiographics1998;18:311–320; discussion 320–313.

10. Patel U, Skingle S, Campbell GA, Crisp AJ, Boyle IT. Clinical profile of acute vertebralcompression fractures in osteoporosis. Br J Rheumatol 1991;30:418–421.

11. Deramond H, Depriester C, Galibert P, Le Gars D. Percutaneous vertebroplasty with polymethyl-methacrylate: technique, indications, and results. Radiol Clin North Am 1998;36:533–546.

12. Chiras J, Depriester C, Weill A, Sola-Martinez MT, Deramond H. [Percutaneous vertebralsurgery: Technics and Indications]. J Neuroradiol 1997;24:45–59.

13. Jensen ME, Evans AJ, Mathis JM, Kallmes DF, Cloft HJ, Dion JE. Percutaneous poly-methylmethacrylate vertebroplasty in the treatment of osteoporotic vertebral body compressionfractures: technical aspects. Am J Neuroradiol 1997;18:1897–1904.

14. Deramond H, Wright NT, Belkoff SM. Temperature elevation caused by bone cement poly-merization during vertebroplasty. Bone 1999;25:17S–21S.

15. Lieberman IH, Dudeney S, Reinhardt MK, Bell G. Initial outcome and efficacy of“kyphoplasty” in the treatment of painful osteoporotic vertebral compression fractures.Spine 2001;26:1631–1638.

16. Galibert P, Deramond H, Rosat P, Le Gars D. [Preliminary note on the treatment of verte-bral angioma by percutaneous acrylic vertebroplasty]. Neurochirurgie 1987;33:166–168.

248 Togawa and Lieberman

17. Bai B, Jazrawi LM, Kummer FJ, Spivak JM. The use of an injectable, biodegradable calciumphosphate bone substitute for the prophylactic augmentation of osteoporotic vertebrae andthe management of vertebral compression fractures. Spine 1999;24:1521–1526.

18. Barr JD, Barr MS, Lemley TJ, McCann RM. Percutaneous vertebroplasty for pain reliefand spinal stabilization. Spine 2000;25:923–928.

19. Mathis JM, Petri M, Naff N. Percutaneous vertebroplasty treatment of steroid-inducedosteoporotic compression fractures. Arthritis Rheum 1998;41:171–175.

20. Cotten A, Dewatre F, Cortet B, Assaker R, Leblond D, Duquesnoy B, Chastanet P, Clarisse J.Percutaneous vertebroplasty for osteolytic metastases and myeloma: effects of the percentageof lesion filling and the leakage of methyl methacrylate at clinical follow-up. Radiology1996;200:525–530.

21. Dean JR, Ison KT, Gishen P. The strengthening effect of percutaneous vertebroplasty. ClinRadiol 2000;55:471–476.

22. Cotten A, Duquesnoy B. Vertebroplasty: current data and future potential. Rev Rhum EnglEd 1997;64:645–649.

23. Cortet B, Cotten A, Boutry N, Dewatre F, Flipo RM, Duquesnoy B, Chastanet P, DelcambreB. Percutaneous vertebroplasty in patients with osteolytic metastases or multiple myeloma.Rev Rhum Engl Ed 1997;64:177–183.

24. Weill A, Chiras J, Simon JM, Rose M, Sola-Martinez T, Enkaoua E. Spinal metastases:indications for and results of percutaneous injection of acrylic surgical cement. Radiology1996;199:241–247.

25. Martin JB, Jean B, Sugiu K, San Millan Ruiz D, Piotin M, Murphy K, Rufenacht B, MusterM, Rufenacht DA. Vertebroplasty: clinical experience and follow-up results. Bone1999;25:11S–15S.

26. Evans AJ, Jensen ME, Kip KE, DeNardo AJ, Lawler GJ, Negin GA, Remley KB, BoutinSM, Dunnagan SA. Vertebral compression fractures: pain reduction and improvement in func-tional mobility after percutaneous polymethylmethacrylate vertebroplasty retrospective reportof 245 cases. Radiology 2003;226:366–372.

27. Cohen JE, Lylyk P, Ceratto R, Kaplan L, Umanskyt F, Gomori JM. Percutaneous vertebro-plasty: technique and results in 192 procedures. Neurol Res 2004;26:41–49.

28. Cortet B, Cotten A, Boutry N, Flipo RM, Duquesnoy B, Chastanet P, Delcambre B. Percu-taneous vertebroplasty in the treatment of osteoporotic vertebral compression fractures: an openprospective study. J Rheumatol 1999;26:2222–2228.

29. Cyteval C, Sarrabere MP, Roux JO, Thomas E, Jorgensen C, Blotman F, Sany J, Taourel P.Acute osteoporotic vertebral collapse: open study on percutaneous injection of acrylic surgicalcement in 20 patients. Am J Roentgenol 1999;173:1685–1690.

30. Maynard AS, Jensen ME, Schweickert PA, Marx WF, Short JG, Kallmes DF. Value of bonescan imaging in predicting pain relief from percutaneous vertebroplasty in osteoporoticvertebral fractures. Am J Neuroradiol 2000;21:1807–1812.

31. Grados F, Depriester C, Cayrolle G, Hardy N, Deramond H, Fardellone P. Long-termobservations of vertebral osteoporotic fractures treated by percutaneous vertebroplasty.Rheumatology (Oxf) 2000;39:1410–1414.

32. Kim AK, Jensen ME, Dion JE, Schweickert PA, Kaufmann TJ, Kallmes DF. Unilateraltranspedicular percutaneous vertebroplasty: initial experience. Radiology 2002;222:737–741.

33. Kaufmann TJ, Jensen ME, Schweickert PA, Marx WF, Kallmes DF. Age of fracture andclinical outcomes of percutaneous vertebroplasty. Am J Neuroradiol 2001;22:1860–1863.

34. Ryu KS, Park CK, Kim MC, Kang JK. Dose-dependent epidural leakage of polymethyl-methacrylate after percutaneous vertebroplasty in patients with osteoporotic vertebral com-pression fractures. J Neurosurg 2002;96:56–61.

35. Kuivaniemi H, Tromp G, Prockop DJ. Mutations in collagen genes: causes of rare and somecommon diseases in humans. FASEB J 1991;5:2052–2060.

Vertebral Augmentation 249

36. Padovani B, Kasriel O, Brunner P, Peretti-Viton P. Pulmonary embolism caused byacrylic cement: a rare complication of percutaneous vertebroplasty. Am J Neuroradiol1999;20:375–377.

37. Cunin G, Boissonnet H, Petite H, Blanchat C, Guillemin G. Experimental vertebroplastyusing osteoconductive granular material. Spine 2000;25:1070–1076.

38. Leeson MC, Lippitt SB. Thermal aspects of the use of polymethylmethacrylate in large meta-physeal defects in bone: a clinical review and laboratory study. Clin Orthop 1993:239–245.

39. Phillips H, Cole PV, Lettin AW. Cardiovascular effects of implanted acrylic bone cement.Br Med J 1971;3:460, 461.

40. Coumans JV, Reinhardt MK, Lieberman IH. Kyphoplasty for vertebral compression fractures:1-year clinical outcomes from a prospective study. J Neurosurg 2003;99:44–50.

41. Dudeney S, Lieberman IH, Reinhardt MK, Hussein M. Kyphoplasty in the treatment ofosteolytic vertebral compression fractures as a result of multiple myeloma. J Clin Oncol2002;20:2382–2387.

42. Ledlie JT, Renfro M. Balloon kyphoplasty: one-year outcomes in vertebral body heightrestoration, chronic pain, and activity levels. J Neurosurg 2003;98:36–42.

43. Phillips FM, Ho E, Campbell-Hupp M, McNally T, Todd Wetzel F, Gupta P. Early radiographicand clinical results of balloon kyphoplasty for the treatment of osteoporotic vertebral com-pression fractures. Spine 2003;28:2260–2265.

44. Garfin SR, Yuan HA, Reiley MA. New technologies in spine: kyphoplasty and vertebroplastyfor the treatment of painful osteoporotic compression fractures. Spine 2001;26:1511–1515.

45. Garfin SR, Reilley MA. Minimally invasive treatment of osteoporotic vertebral body com-pression fractures. Spine J 2002;2:76–80.

46. Hiwatashi A, Moritani T, Numaguchi Y, Westesson PL. Increase in vertebral body heightafter vertebroplasty. Am J Neuroradiol 2003;24:185–189.

250 Togawa and Lieberman

12Principles of Transthoracic, Transperitoneal, and Retroperitoneal Endoscopic Techniques

in the Thoracic and Lumbar Spine

Geoffrey M. McCullen, MD and Hansen A. Yuan, MD

INTRODUCTION

Endoscopes are rigid, straight, or angled systems that provide visualization, light, andmagnification to anatomical areas, thereby avoiding larger open incisions. The endoscopeconsists of optical fibers and a light source. Each fiber delivers a separate piece ofvisual information to a camera and a video-integrated system. The camera processes themultiple image components into picture elements known as “pixels.” To increase pic-ture quality and clarity, the number of optical fibers and pixels would have to beincreased. Given the size constraints of an endoscope, an increase in the number ofoptical fibers would require a decrease in fiber size. However, if the fiber becomes toosmall, the capacity to transmit light is significantly impeded. Presently, the maximumnumber of pixels in a camera system given the size constraints of the straight 10-mm-diameter thoracic or lumbar endoscope is 30,000. Zero and 30° angled scopes are mostcommonly used.

Imaging advances have assisted the safe implementation of minimally invasivestrategies. In laparoscopic interbody fusions, the exact midline of the disc must be iden-tified using a true anteroposterior fluoroscopic image with symmetrical pedicles andflat end plates. Radiation safety precautions should be followed to minimize the riskswhile working under fluoroscopy. Frameless stereotaxy, developed in 1992, was initiallydesigned for intracranial use. The technique links the anatomy to a preoperatively acquiredimage. In endoscopic approaches, navigational systems have been difficult to applybecause of problems with registration (precisely correlating anatomical landmarks withimage reference points). External landmarks are not reliable as implanted fiducials forregistration. A frame attached to a pedicle screw (placed percutaneously) can serve as astable, fixed reference. A computed tomography (CT) scan is subsequently performed. Regis-tration, using the geometry of the frame as fiducials, has been successful when usedwith endoscopic spine surgery (1,2).

Intraoperative nerve monitoring can assess for nerve compression or irritability.Mechanically elicited electromyograph activity recorded in the muscles innervated bythe lumbar nerve roots can alert the surgeon to nerve proximity.

251

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

ADVANTAGES OF ENDOSCOPIC TECHNIQUES

The principal purpose of endoscopic techniques is to approach the spine through portalsrather than larger skin incisions. At the target site, the same operative procedure is per-formed using an endoscopic approach as is performed using an open approach. Benefitsinclude decreased soft-tissue disturbance, leading to lesser postoperative scarring andpain as well as reduced ultimate healing time. In addition, the entire operating team isable to watch the monitor during the procedure.

DISADVANTAGES OF ENDOSCOPIC TECHNIQUES

Spinal endoscopic procedures are technically demanding and require a dedicated effortto safely overcome the “learning curve.” The vascular or thoracic surgeon and the spinesurgeon should train together in the laboratory before performing live surgery on humans.The surgeon should always be prepared to convert the case to an open procedure withopen laparotomy and thoracotomy instruments and vascular instruments close at hand.

THORACOSCOPIC SPINAL PROCEDURES (SEE TABLE 1)

In 1807, in Frankfurt, Germany, Bozzini was the first recorded individual to use anendoscope. Known as the “lichtleiter,” this device used candle illumination to examinebody orifices (3). Lens and light amplification improvements followed. Throughout the1920s, Jacobaeus, in Sweden, was the first to perform both laparoscopic and thoracoscopicprocedures in humans (4,5). He used a cystoscope and a heated platinum lighting loop.Thoracoscopic intrapleural pneumolysis was performed on patients with tuberculosis.

With improvements in medical management of tuberculosis, interest in thoracic endo-scopic approaches waned from 1960 to 1990. In the early 1990s, renewed interest wasexperienced in thoracoscopy for the inspection and treatment of pleural diseases and forendoscopic pulmonary resection. During these procedures, excellent visualization of thethoracic spine was recognized. Thoracoscopic spine procedures began with the drainageof an intervertebral disc abscess (6,7).

Similar to open thoracotomy, the majority of thoracoscopic procedures are performedin the lateral decubitus position requiring a dual lumen endotracheal tube for selectivelung ventilation. Thoracoscopic spine surgery indications include anterior release of large(>80°) and fixed (corrects to <60° with push-prone views) scoliotic curves, Scheuermann’skyphosis >90° that fails to correct to <50° with hyperextension, anterior fusion in skele-tal immaturity to decrease the incidence of postoperative crank-shaft, decompressivediscectomy, and corpectomy. Endoscopic anterior instrumentation has been developedand is still evolving. Problems encountered include obtaining safe bicortical screw pur-chase and difficulties performing compression or distraction and rod rotation.

The recent introduction of the prone position for thoracoscopic spinal proceduresoffers benefits including a more familiar orientation, gravity-assisted retraction, gravity-assisted correction of kyphosis, elimination of the need for repositioning, and use of astandard single-lumen endotracheal tube (8,9). Simultaneous posterior exposure andprone thoracoscopic release has been reported (10).

Principles of Technique

With thoracoscopy, it is possible to visualize from T4 to L1 and thoracoscopic right orleft approaches are possible. Compared to open procedures, thoracoscopic techniques

252 McCullen and Yuan

cause less acute and chronic postoperative pain/intercostal neuralgia and improve pul-monary function (11,12). In addition, improved shoulder girdle strength and range ofmotion (12), decreased cost, and reduced hospital stay have been reported with the tho-racoscopic technique (11,13).

Thoracoscopic procedures are usually contraindicated in those patients who haveundergone multiple anterior thoracic procedures with expected scar and adhesions. Patientswith neuromuscular deformity and a history of pneumonia or empyema often have thickpleural adhesions. Those patients with significant restrictive lung disease will be unable totolerate single-lung ventilation.

Lateral thoracoscopy requires single-lung ventilation, a double-lumen endotrachealtube, and high airway pressures. Correct placement of the double-lumen endotrachealtube followed by confirmation with fiberoptic bronchoscopy is necessary. The tubecan dislodge, or tracheal tears may occur when turning the patient from supine to thelateral position.

When approached from the convex side of a scoliotic curve, single-lung ventilationmust occur in the smaller lung on the concave side of the curve. Large scoliotic curves(>90°) result in a smaller chest cavity, limiting the space available to perform the endo-scopic procedure successfully (12). Patients with right idiopathic scoliosis have a moreposterior aorta (14). Whereas an open thoracotomy approach allows circumferentialexposure of the spine so that a finger can be placed around the opposite side to protectthe far-side vasculature during placement of screws, with thoracoscopic fusion suchprotection is not possible.

Prone thoracoscopy uses double-lung ventilation, with decreased tidal volumes andincreased respiratory rate. There is less anesthetic preparation time with prone vs lateralthoracoscopy (8,9). Postoperative oxygen requirements are decreased with double-lungvs single-lung ventilation (9). The lateral position must be used for discectomy aboveT4 using an axillary portal anterior to the pectoralis major.

Intercostal neuralgia after thoracoscopy is not uncommon, occurring in approx 7%of cases (15). Softer, flexible trocars have helped reduce the development of intercostalneuralgia. The sixth or seventh intercostal space is the safest region for entry for the firstthoracic port. All subsequent port placements should be performed under direct visual-ization. Perforation of the diaphragm and parenchymal lung injury are best avoided bydirectly visualizing all instruments when introduced into the chest cavity.

With endoscopic instrumentation, screw pullout at the cephalad screw is usually theresult of poor screw placement and unicortical purchase. Long screw tips adjacent tothe aorta on the contralateral side are a cause for concern. Tension pneumothorax canoccur secondary to overadvancement of a guide wire during cannulated screw instru-mentation (16).

Results

Thoracoscopic procedures have been compared with open procedures in clinical andlaboratory studies. In the laboratory, thoracoscopic discectomy for release/fusion isequal to the open technique in the percentage of disc removal (76% for open, 68% forthoracoscopic) (17) and in the adequacy of the biomechanical release (18). In scoliosisanterior release/fusion, the percent curve correction, blood loss, and complication rateare similar when comparing open and endoscopic methods (9,12,19). The endoscopic

Principles of Endoscopic Techniques in Spine 253

technique is 28% more expensive, reflecting the cost of the expensive disposable tools(20). Thoracoscopic release requires a 50% longer operating time compared to openthoracotomy (14). The “learning curve” demonstrates improvement in operating times,with early thoracoscopic release taking 29 min/disc level, improving to 22 min/levelwith experience (20). Thoracic disc excision for radicular and myelopathic patients hasdemonstrated a 70% clinical success rate, with a mean operative time of 173 min, bloodloss of 259 cc, and average hospital stay of 4 d (21).

LUMBAR TRANSPERITONEAL AND RETROPERITONEALENDOSCOPIC SPINAL PROCEDURES

In 1901, Ott (22) used a cystoscope to visualize structures within the pelvis. In 1902,Kelling (23) was the first to induce pneumoperitoneum in dogs. Oxygen followed byCO2, was used for insufflation. In 1938, Veress developed the insufflation needle thatbears his name and is still in use today.

In 1991, Obenchain and Cloyd described transperitoneal laparoscopic lumbar discec-tomy (24) (see also Table 2). Performed in a supine position, the technique is used toaccess the L4-L5 and L5-S1 levels. Laparoscopic instrumentation with BAK cylindricalinterbody devices (Sulzer Spine Tech, Minneapolis, MN) was first reported in 1995 (25).Typically three or four 1-cm portal sites are prepared: one or two for retraction, one forworking instruments, and one for the endoscope. The posterior peritoneum is opened.Blunt dissection in the retroperitoneal space is performed in order to avoid injury to theparasympathetic plexus. Laparoscopic approaches to the L5-S1 level have been com-monly performed because this space is below the bifurcation. The L4-L5 level can bedifficult to expose when the level is above the bifurcation and the iliolumbar vein must beidentified, ligated, and divided in order to retract the great vessels from left to right.

A balloon-assisted endoscopic retroperitoneal gasless (BERG) technique allowingthe use of conventional instruments and avoiding the complications of CO2 insufflationhas been described (26). The patient is placed in the supine position with a radiolucentsupport placed under the patient’s left flank. An incision is made midway between theiliac crest and the costal margin along the midaxillary line. The external and internaloblique muscles are dissected bluntly. The transversus abdominus muscle is entered,exposing the retroperitoneal fat. Blunt finger dissection is used followed by introductionof a dissecting balloon. An endoscope is placed through the cannula into the balloonwhile the balloon is being inflated, allowing visualization of the expanding retroperitonealspace and close observation of the “receding line” of the peritoneum. A 2-cm incision ismade 2 cm off the anterior abdominal midline. Dissection continues until the balloon isidentified. A fan-shaped retractor is placed and is used to elevate the anterior abdominalwall. A retractor is used to displace the peritoneum and intra-abdominal contents medially.The balloon is deflated. The endoscope is used to visualize in a gasless working cavity.

The endoscopic retroperitoneal approach was first developed for urological surgicalprocedures and adapted for lumbar spine interbody fusion (27). The procedure can beperformed using gas insufflation, balloon insufflation (gasless), or a combination of thetwo techniques. A lateral decubitus positioning is utilized. A 2- to 3-cm incision is createdat the intended level (L2-L5) that is centered on a line between the eleventh rib and theanterior superior iliac spine. Blunt dissection is performed through the muscles using anendoscopic trocar until the fat of the retroperitoneal space is identified. A dissection

254 McCullen and Yuan

balloon is placed and inflated until a retroperitoneal cavity has been created. The balloonis removed and either CO2 insufflation (to a pressure of 5 mmHg) or a self-retainingretractor system is placed. Typically, three ports are used: retraction, endoscope, andworking instruments.

Principles of Technique

In transperitoneal laparoscopy, insufflation is required for visualization. To maintainpneumoperitoneum, the use of suction is limited. CO2 insufflation can cause hypercapnia,elevation of the peak pulmonary pressures and mean arterial pressure, and CO2 embolism(secondary to decreased diaphragm movements and increased CO2 absorption). Despite aTrendelenburg positioning and the use of multiple ports, the small bowel mobilization andretraction remains a problem. Vascular mobilization can be even more difficult. Routinepreoperative magnetic resonance imaging or CT scanning can be used to classify vascularanatomy (28). If the bifurcation of the great vessels is above the L4-L5 disc space, alaparoscopic approach to L4-L5 is technically more feasible. Otherwise, the iliolumbarvein should be identified, mobilized, and ligated for exposure to the L4-L5 level.

Previous intra-abdominal or retroperitoneal operative interventions may create scarthat adversely affects tissue mobilization and visualization. Laparoscopic procedures areunable to visualize the neural elements and to directly address spinal canal stenosis.Fusion for internal disc derangement with tall discs is a relative contraindication becauseit is more difficult to obtain adequate disc distraction. The larger interbody devices thatwould be required for fusion of tall disc spaces cannot be delivered through existingdevices that would maintain pneumoperitoneum.

Retroperitoneal, lateral disc exposure can be used for access to L4-L5 and above (2).Lateral endoscopic retroperitonoscopy has a reduced risk of small-bowel adhesions andautonomic plexus dysfunction (27). Performed in the lateral decubitus position, thisprocedure allows the intra-abdominal contents to “fall away” from the spine. In theretroperitoneal endoscopy, the peritoneum is left intact, decreasing the postoperativecomplications related to manipulation of the bowel and disruption of the peritoneum. Inaddition, the intact peritoneum serves as a retractor aiding in the control of the bowel.

Principles of Endoscopic Techniques in Spine 255

Table 1Thoracoscopic Spinal Endoscopy

Advantages Disadvantages Options

Improved visualization Large (>90°) scoliosis Lateral positioning: double-curves creating smaller lumen endotracheal tube,available “working space” single-lung ventilation

Less tissue dissection to Inability to perform Prone positioning: double-accomplish approach circumferential exposure for lung ventilation,

far-side tissue protection simultaneous anterior/posterior procedures

Similar discectomy extent and Difficult but improving release capability when anterior instrumentationcompared to open thoracotomy

With lateral retroperitoneal laparoscopy, the psoas is often very large and difficult tomobilize. A muscle-splitting approach through the psoas may lead to injury of the gen-itofemoral nerve or elements of the lumbosacral plexus. A transpsoas dissection at L5-S1would risk the L4 and L5 nerve roots, the femoral nerve, and the obturator nerve and,therefore, is not recommended. The psoas muscle can be split more anteriorly than thedorsal fourth of the lumbar vertebral body from the cranial third of the L3 vertebralbody and above. When the psoas is split at the caudal two-thirds of L3 or at L4 there isa risk of injury to the genitofemoral nerve (29).

Results

In a study comparing transperitoneal laparoscopic vs mini-open approach, thecomplication rate was 20% in laparoscopic vs 4% in mini-open (30). Sixteen percent ofthe transperitoneal laparoscopic approaches have been considered “inadequate,” allowingone rather than two cages to be placed (30). Approximately 10% of laparoscopic proce-dures have required conversion to open for repair of vessel lacerations or to close tears inthe peritoneum (31,32). There is no significant difference between minilaparotomy andtransperitoneal laparoscopic approach when comparing analgesia requirements, time toresuming oral intake, or length of hospitalization (30,33). Laparoscopy costs more($1374/case on average) (28). Laparoscopic operative time averages 167 min for singlelevel and 215 min for multiple levels (28).

Laparoscopic complications include vascular and peritoneal/visceral injuries. Duringlaparoscopy, the insufflation pressure should be decreased to 10 mmHg or less duringstages within the case to check for areas of venous bleeding that could otherwise go unrec-ognized at case completion. Retrograde ejacul*tion rate among males is high after lum-bosacral laparoscopy—16–25% (28,31)—compared with a 6% rate with mini-open (34).Avoiding monopolar electrocautery and limiting the degree of dissection along the left sideof the aorta and the left iliac artery may help to minimize the risk of ejacul*tory dys-function (28). Ureteral injury has been reported (35). During transperitoneallaparoscopy, the sigmoid colon mesentery is approached from the right. The right ureter,

256 McCullen and Yuan

Table 2Laparoscopic Spinal Procedure

Advantages Disadvantages Options

Anterior interbody fusion Typically requires CO2 Prone: transperitoneal,insufflation; elevated retroperitoneal gasless mean arterial pressure, (BERG), L4-L5 and L5-S1;hypercapnia, embolism, complications: vesselunrecognized venous mobilization, retrogradebleeding ejacul*tion, ureteral injury

No direct canal Lateral: retroperitoneal,decompressive capability L4-L5 to L2-L3;

complications: psoas,genitofemoral, and lumbosacral plexus injuries

Interbody graft size/shapelimitations

traveling over the right iliac artery, must be identified before making the posterior peri-toneum incision. In lateral endoscopic transpsoas approaches, a 30% rate of transientparesthesia in the groin/thigh region has been reported (2).

CONCLUSION

Endoscopic spinal procedures are a relatively recent addition to the spine surgeon’sarmamentarium. The techniques offer the surgeon an enhanced visualization of the opera-tive target site with less skin, soft tissue, and muscle disruption. Although there are definitebenefits, these procedures are technically challenging and there are associated risks.

As new modalities are developed, care should be directed to prevent inventing newindications to justify the technique. The core indications for surgical intervention shouldnot change. A “learning curve” should be expected and preparations made for it. Endo-scopic technology will continue to evolve by merging it with biomedical advancementsin robotics and image guidance systems.

REFERENCES

1. Bergey D, Regan J. Lateral endoscopic transpsoas spinal fusion: review of technique andclinical outcomes in a consecutive series. Spine J 2003;3:166S.

2. Brody F, Rosen M, Tarnoff M, Leberman I. Laparoscopic lateral L4-5 disc exposure. SurgEndosc 2002;16(4):650–653.

3. Bush RB, Leonhardt H, Bush IV, Landes RR. Dr. Bozzini’s Lichtleiter: a translation of hisoriginal article (1806). Urology 1974;3:119–123.

4. Jacobeaus JC. Possibility of the use of the cystoscope for investigation of serious cavities.Munch Med Wochenschr 1920;57:2090–2091.

5. Jacobeaus JC. The cauterization of adhesions in pneumothorax treatment of tuberculosis.Surg Gynecol Obstet 1921;32:493–500.

6. Mack MJ, Regan JJ, Bobechko WP, et al. Application of thorascopy for diseases of thespine. Ann Thorac Surg 1993; 56:736–738.

7. Regan JJ, Mack MJ, Picetti G. A technical report of video-assisted thorascopy (VATS) inthorascopic spinal surgery: preliminary description. Spine 1995;20:831–837.

8. King AG, Mills TE, Loe WA Jr, Chutkan NB, Revels TS. Video-assisted thorascopicsurgery in the prone position. Spine 2000;25 (18):2403–2406.

9. Succato DJ, Elerson E. Positioning for anterior thorascopic release and fusion: prone ver-sus lateral. Spine 2003;E 28(18):2176.

10. Lieberman IH, Salo PT, Orr RD, Kraetschmer B. Prone position endoscopic transthoracicrelease with simultaneous posterior instrumentation for spinal deformity. Spine2000;25(17):2251–2257.

11. Dickman CA, Detweiler PW, Porter RW. Endoscopic spine surgery. Clin Neurosurg.2000;46:526–553.

12. Newton PO, Marks M, Faro F, et al. Use of video-assisted thoracoscopic surgery to reduceperioperative morbidity in scoliosis surgery. Spine 2003;28:S249–S254.

13. Han PP, Kenny K, Dickman CA. Horacoscopic approaches to the thoracic spine: experi-ence with 241 surgical procedures. Neurosurgery 2002;51:88–95.

14. Succato DJ, duch*ene C. MRI analysis of the position of the aorta relative to the spine: acomparison between normal patients and those with idiopathic scoliosis. J Bone Joint Surg2003;85A:1461–1469.

15. McAfee PC, Regan JR, Zdeblick T, et al. The incidence of complications in endoscopic ante-rior thoracic and lumbar spinal reconstructive surgery. Spine 1995;20:1624–1632.

Principles of Endoscopic Techniques in Spine 257

16. Roush TF, Crawford AH, Berlin RE, Wolf RK. Tension pneumothorax as a complication ofvideo-assisted thorascopic surgery for anterior correction of idiopathic scoliosis. Spine2001;26(4):448–450.

17. Huntington CF, Murrell W, Betz RR, et al. Comparison of thoracoscopic and open thoracicdiscectomy in a live ovine model of anterior fusion. Spine 1998;23(15):1699.

18. Wall EJ, Bylski-Austrow DI, Shelton FS, Crawford AH, Kolata RJ, Baum DS. Endoscopicdiscectomy increases thoracic spine flexibility as effectively as open discectomy: amechanical study in a porcine model. Spine 1998;23(1):9–16.

19. Huang EY, Acosta JM, Gardocki, RJ, et al. Thoracoscopic anterior spinal release andfusion: evolution of a faster, improved approach. J Pediatr Surg 2002;37(12):1732–1735.

20. Newton PO, Shea KG, Granlund KF. Defining the pediatric spinal thoracoscopy learningcurve: sixty-five consecutive cases. Spine 2000;25(8):1028–1035.

21. Anand N, Regan JJ. Video-assisted thorascopic surgery for thoracic disc disease: classificationand outcome study of 100 consecutive cases with a two year minimum follow-up. Spine2002;27(8):871–879.

22. Ott DV. Illumination of the abdomen. J Akusk Ahensk Boliez 1901;15:1045–1049.23. Kelling G. Uberoesophagoskopie, gastoskopie and kalioskope. Munch Med Wochenschr

1902;52:21.24. Obenchain TG. Laparoscopic lumbar discectomy. J Laparoendosc Surg 1991;1:145–149.25. Zucherman JF, Zdeblick TA, Bailey SA, Mahvi D. Hsuk Y, Kohrs D. Instrumented laparo-

scopic spinal fusion: preliminary results. Spine 1995;20:2029–2034.26. Thalgott JS, et al. Balloon-assisted endoscopic retroperitoneal gassless (BERG) lumbar

discectomy and fusion, 15th Annual Meeting of the Joint Section on Disorders of the Spineand Peripheral Nerves, Orlando, 1999.

27. McAfee PC, Regan JJ, Geis WP, Fedder IL. Minimally invasive anterior retroperitonealapproach to the lumbar spine: emphasis on the lateral BAK. Spine 1998;23:1476–1484.

28. Kleeman TJ, Michael Ahn U, Clutterbuck WB, et al. Laparoscopic anterior lumbar inter-body fusion at L4-5. Spine 2002; 27:1390–1395.

29. Moro T, Kikuchi S, Konno S, Yaginuma H. An anatomic study of the lumbar plexus withrespect to retroperitoneal endoscopic surgery. Spine 2003;28(5):423–428.

30. Zdeblick TA, David SM. A prospective comparison of surgical approach for anteriorL4-5 fusion: laparoscopic versus mini anterior lumbar interbody fusion. Spine 2002;25(20):2682–2687.

31. Escobar E, Transfeldt E, Garvey T, Ogilvie J, Graber J, Schuttz L. Video-assisted versusopen anterior lumbar spine fusion surgery: a comparison of four techniques and complica-tions in 135 patients. Spine 2003;28(7):729–732.

32. Regan JJ, Yuan H, McAfee PC. Laparoscopic fusion of the lumbar spine: minimally inva-sive spine surgery. A prospective multicenter study evaluating open and laparoscopic lumbarfusion. Spine 1999;24:402–411.

33. Rodriguez HE, Connolly MM, Dracopoulos H, Geisler FH, Pobielski FJ. Anterior access tothe lumbar spine: laparoscopic versus open. Am Surg 2002;68(11):982–983.

34. Kaiser MG, Haid RW Jr, Subach BR, Miller JS, Smith CD, Rodts GE Jr. Comparison ofmini-open versus laparoscopic approach for anterior lumbar interbody fusion: a retrospec-tive review. Neurosurgery 2002;51:97–105.

35. Guingrich JA, McDermott JC. Ureteral injury during laparoscopy-assisted anterior lumbarfusion. Spine 2000;25(12):1586–1588.

258 McCullen and Yuan

13Use of Laser in Minimally Invasive Spinal Surgery

and Pain Management

John C. Chiu, MD, DSC and Martin H. Savitz, MD, PhD

INTRODUCTION

Arthur L. Schawlow and Charles H. Townes are credited with the invention of thelaser—light amplification by stimulated emission of radiation—dating back to 1958 withthe publication of “Infrared and Optical Lasers.” The work of Schawlow and Townestraced back to the 1940s, when there was an attempt to create a device for studyingmolecular structure (1). Extending the research from microwaves to the infrared regionof the spectrum required a series of mirrors to focus the shorter wavelengths. In 1960, apatent was granted for the laser. Townes was awarded the Nobel Prize in Physics in 1964,and Schawlow in 1981 (2).

Numerous applications in medicine have been adapted to the specialties of ophthal-mology, plastic surgery, urology, vascular surgery, general surgery, gynecology, neuro-surgery, and orthopedics. Ascher (1) in Germany was one of the first investigators toemploy the CO2 and neodymium lasers in neurosurgery. His experience with hemostasisand vaporization in the resection of tumors of the brain and spine was extensive.

LASER DISCECTOMY

In 1985, Ascher (3) applied laser technique to disc surgery. Measurements of theintradiscal pressure were made before and after laser discectomy with a saline manometerin order to document objectively the pathophysiological effect. In 1990, Yonezawa et al.(4), in Japan, used the neodymium:yttrium-aluminum-garnet (Nd:YAG) laser through adouble-lumen needle with a bare quartz fiber. The tip-type pressure transducer was sim-ilarly able to record the preoperative and postoperative intradiscal pressure. In 1992,Davis (5) employed the potassium-trideuterium-phosphate (KTP) laser for lumbar discablation and considered the success rate (32 of 40 patients, or 80%) equivalent to theresults obtained by Onik (6) with his nucleotome and automated percutaneous lumbardiscectomy. Again in 1992, Choy et al. (7) reported percutaneous laser discectomy(PLD) with the Nd:YAG laser; 333 cases of herniated, nonsequestered lumbar discswere diagnosed by computed tomography (CT) and/or magnetic resonance imaging(MRI). The success rate at an outpatient surgical facility was 78.4% over a 5-yr period,

259

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

and the follow-up was 12–62 mo (average: 26 mo). One-third of repeat MRIs showed amoderate decrease in disc herniation. Kambin (8) reported the effects of the CO2,Nd:YAG, and holmium (Ho).YAG lasers on the nucleus pulposus in cadaveric discs in1991. Sherk et al. (9), beginning in 1993 and continuing to the present, have performedlaser discectomy on a series of patients.

Between 1994 and 1999, nine series of PLDs were reported (Table 1), totaling morethan 4300 patients. The overall success rate was about 80%. Hellinger (10) operated on38 thoracic disc herniations. Isolated postoperative complications including transientfoot drop, permanent traumatic neuropathies, pleuritis, pneumothorax, and complexregional pain syndrome (11) were <1% (12).

A variety of wave configurations and powers created a 1.5-cm3 defect in the centerof the disc. Adjoining tissue was studied histologically for possible thermal effect (8).Thermocouples were placed in adjacent nerve roots, end plates, and posterior longitudinalligaments to record temperature changes. Intradiscal pressure measurements werealso made before and after laser ablation. All of the lasers caused minimal temperatureelevations to 102°F in the nerve roots. Histological slides confirmed that the free-beamNd:YAG caused an excessive thermal change in the end plates and nerve roots. Furtherfindings included the fact that 1200–1500 J of laser energy diminished the intradiscalpressure by 25–50%. The free-beam CO2 was not practical because the laser couldnot easily be delivered into the disc space. The Nd:YAG contact fiber and theHo:YAG fiber were readily inserted through a needle, and both delivered effectiveamounts of laser energy.

The published advantages of PLD (21–25) include simplicity of minimally invasivetechnique, small caliber of instruments, documented reduction in intradiscal pressure,low rate of complication, and no spinal instability. Blind laser nucleolysis through aposterolateral approach (26,27) has a number of limitations and disadvantages: minimalflexibility, inability to reach subligamentous fragments, no documentation of area ofvaporization of collagenized nucleus, and lack of control of thermal spread to nerveroot and end plates. Blind PLD, like automated percutaneous lumbar discectomy,occupies an important place in the history of minimally invasive surgery and is still

260 Chiu and Savitz

Table 1Studies of PLDs From 1994 to 1999

Authors/ref. Year Laser Cases

Ohnmeiss et al. (13) 1994 Nd:YAG 204Simons et al. (14) 1994 Nd:YAG 150Schatz and Talalla (15) 1995 Nd:YAG 500Liebler (16) 1995 Nd:YAG 333

KTP 117Siebert et al. (17) 1996 Nd:YAG 180Casper et al. (18) 1996 Ho:YAG 50Nerubay et al. (19) 1997 CO2 50Pedachenko et al. (20) 1998 Nd:YAG 273Hellinger (8) 1999 Nd:YAG 2535

performed by some spinal surgeons. Most minimally invasive spinal surgeons prefermechanical removal of disc material following proper positioning of instruments in thetriangular working zone (Fig. 1) (6) (see Chapter 4). Other surgeons use both instru-ments and laser modulation under endoscopic control (28–32) (Fig. 1). Monitoring thethermal effects of the laser on neural tissue mitigates almost all complications. The useof the laser can also be extended and expanded under direct visualization.

Interventional procedures on the spine in cases of discogenic pain syndromes shouldnot merely be measured by their success rate but also by the resulting amount of postop-erative complications (21–25). Particular attention should be devoted to sequelae resultingfrom spinal interventions including injections in close proximity to the spine. Nerve blocksor paravertebral infiltrations can result in disasters such as paraplegia (24,25). DuringOctober and November of 1989, Siebert and colleagues (15–17) (abdominal position) andHellinger (21) (lateral position) introduced nonendoscopic percutaneous laser disc decom-pression and nucleotomy with the Nd:YAG laser in Germany, after Ascher (1) and Choy

Laser in Spinal Surgery and Pain Management 261

Fig. 1. The triangular working zone represents ample access for a percutaneous approach tolumbar disc space.

Fig. 2. Drawing of anterior approach to cervical disc and insertion of spinal needle followedby nucleotome(28).

et al. (29,30) had undertaken the first global attempts in 1986. Topological elaboration ofthe cervical spine (Fig. 2) and the thoracic spine soon followed (Fig. 2 and 3). Complica-tions of PLD have included paraspinal hematoma, perforation in the peripheral ileum,vasovagal reactions, intradiscal abscess, worsening of preexisting foot drop, postoperativecholangitis, and postoperative cerebrovascular accident (21–25).

LASER FACET RHIZOTOMY

Lumbar facets or zygapophyseal joints are synovial arthroses richly innervated withnerve endings from the medial branch of the posterior primary ramus (33). Osteoarthri-tis of the lumbar facet joint is a common cause of disabling low-back pain. Facet jointarthritis on MRI and CT scan does not necessarily correlate with symptoms. Somepatients can have extensive facet arthritis on imaging study and be clinically asymp-tomatic, whereas other patients can have subtle evidence of arthritis radiologically withclassic symptoms of arthritis clinically (34). Many patients report referred sciatica aswell. Currently, the treatment options include apophyseal joint nerve blocks for short-term relief and facet joint denervation for long-term relief by cryotherapy or radiofre-quency. The results of facet rhizotomy have been variable, with a significant number ofpatients requiring repeat procedures or experiencing inadequate relief of pain.

Few reports of laser-assisted lumbar facet rhizotomy have appeared in the literature(35). By means of a 17-gage needle (Fig. 4), the Ho:YAG straight-firing laser probe(Trymedine, Irvine, CA) was directed at the medial branch of the dorsal ramus, the nervethat gives rise to the articular branches at each level. Each facet joint receives its nervesupply from multiple segments above and below (34). The laser probe was also directed

262 Chiu and Savitz

Fig. 3. The arrow represents the posterolateral approach to thoracic disc space.

onto the facet joint to thermocoagulate the finer capsular branches with a total of 500 J ofenergy. There was a significant difference in pain relief between patients with unoperatedbacks vs patients who had undergone prior spinal surgery. Laser has the advantage of ther-mocoagulating a relatively larger area in the vicinity of the probe tip than a radiofrequencyprobe. More recently, Katzman (36) reported arthroscopic laser-assisted facet joint dener-vation. A large number of patients who failed to be relieved by radiofrequency facet rhizo-tomy experienced marked relief of pain following the laser procedure.

ENDOSCOPIC LASER FORAMINOPLASTY

With accumulated experience with endoscopically assisted mechanical and laserlumbar discectomy (37), the need for a method to decompress more effectively the lat-eral recess and intervertebral neural foramen from very large or extruded disc protru-sions, recurrent discs, scar tissue, and spondylitic spurs became evident (38). The mostfrequently seen lumbar spinal disc disease in the elderly is spinal and lateral foraminalstenosis (39). Lateral stenosis may be congenital or degenerative when secondary to acutedisc disease and spinal trauma. Transforaminal Microdecompressive Endoscopic AssistedDiscectomy and Foraminoplasty (TF-MEAD) (Fig. 5) is a new system of more aggressivemechanical instrumentation and laser application.

Positioning of a Steerable Spinoscope (Karl Storz Endoscopy, Culver City, CA) witha flexible tip that can bend up to 90° and rotate to reach 360° is checked throughout theprocedure by fluoroscopy in two planes (Fig. 6). At the involved nerve root distribution,sterile needle electrodes are placed for continuous intraoperative neurophysiologicalelectromyograph monitoring (40). If a pain provocation test and discogram were not

Laser in Spinal Surgery and Pain Management 263

Fig. 4. Fluoroscopic view of needle placed on facet at L5-S1.

done preoperatively, they are done at the outset. Under fluoroscopy the extendedside of the appropriate cannula is turned to face the nerve root for retraction and pro-tection. The cannula retractors have various duck bill extensions. Larger, moreaggressively toothed trephines are then inserted and rotated to cut through annulus,disc protrusion, spur, or spondylitic bar. A slim rongeur, spinal disc forceps, or pitu-itary forceps and curettes can aid in decompressing the foramen and the lateral recess(Fig. 5). Biting forceps, a discectome, and an Ho:YAG laser with continuous irrigationare used consecutively to perform intradiscal discectomy. A lower-energy nonablativelaser is applied for shrinking and tightening of the disc (laser thermodiskoplasty)(41,42). Laser thermodiskoplasty can also cause sinovertebral neurolysis or denervation.The discectome is employed to remove charred debris from use of the laser.

The disc space and neural foramen can be directly visualized and examined byendoscopy to confirm adequate disc decompression and to perform further decompres-sion when necessary. If the foramen is compromised, the depth of insertion of theendoscope is adjusted; the nerve root is again protected by the duck bill extension; andspurs are removed with curettes, bone punches, and Kerrison rongeurs. TF-MEADeffectively treats spinal pathology at multiple levels and bilaterally. Many elderlypatients suffering symptoms caused by lateral spinal stenosis and disc problems can besuccessfully treated.

LASER SYMPATHECTOMY

Two common syndromes in which sympathetic pain appears are causalgia and reflexsympathetic dystrophy. True causalgia follows partial injury to a major nerve trunk such as

264 Chiu and Savitz

Fig. 5. Endoscopic view of lumbar mechanical decompressive foraminoplasty and discec-tomy: (A) disc removal with cutter forceps; (B) removal of disc fragment below nerve; (C)curette for osteophytic decompression; (D) rasp for osteophytic decompression; (E) bone punch,rongeur for foramen decompression; (F) postforaminoplasty disk defect (arrow).

the sciatic nerve or its large branches. Reflex sympathetic dystrophy may occur followingminor trauma to the neural structures that accompanies fractures, soft-tissue injuries, andsurgical incisions. Clinical characteristics include burning, poorly localized dermatomaldistribution of stabbing pain, hyperesthesia, vasomotor alterations leading to trophicchanges, changes in skin temperature, alteration of sweating patterns, piloerection, andswelling. Other conditions complicated by sympathetic dysfunction are amputationstump pain, circulatory insufficiency in the legs, arteriosclerotic disease of the lower limbs,intermittent claudication, and arterial embolism.

Lumbar disc surgery, including percutaneous endoscopic discectomy, may causemechanical trauma to the somatic nerves, and a certain percentage of patients docomplain of burning pain, lower-extremity swelling, and color changes and hypersensi-tivity of the skin. Diagnostic lumbar sympathetic blocks can identify sympathetic nervoussystem involvement as the causative factor. The sympathetic nervous system becomes

Laser in Spinal Surgery and Pain Management 265

Fig. 6. Drawing and X-ray appearance of Steerable Spinoscope with bendable tip in situ.

involved owing to damage to A delta and C fibers that develop hypersensitivity to circu-lating norepinephrine, pressure, and movement. Spontaneous firing causes the typicalpain of sympathetic origin. Shortly afterward, small neuromas form that sprout smallmyelinated and unmyelinated fibers. Normally silent fibers then generate, even in theabsence of stimulation, an ongoing barrage of impulses that traverse the afferent fibers tothe spinal cord.

In the past, permanent interruption of the lumbar sympathetic chain has been accom-plished by open surgery or phenol or alcohol injection (43). Subsequently, percutaneouslumbar sympathectomy by radiofrequency lesions involved less morbidity. The Ho:YAGlaser has now been found to be even more effective and longer lasting than radiofrequency.

Anatomically, the lumbar sympathetic chain lies at the anterolateral border of thevertebral bodies. The aorta is positioned anteriorly and slightly medial to the chain onthe left side. The inferior vena cava is more closely approximated to the chain on theright in an anterior plane. Many other small lumbar arteries and veins are positionednear the sympathetic chain. The psoas muscle is situated posteriorly. Blockade of thesympathetic nerves can also be performed with spinal, epidural, or peripheral nerveblocks, but relief of pain after lumbar sympathetic block will most clearly confirmthe painful etiology as sympathetically mediated. Most fibers headed for the lowerextremity pass through the second and third lumbar ganglia, so that a sympathetic blockplaced at this level provides almost complete sympathetic denervation to the lowerextremity. The pain relief obtained is usually immediate and can be long-lasting, out-lasting the duration of action of the local anesthetic.

Sites for needle placement for the sympathetic chain at the L2, L3, and L4 vertebrallevels are identified on projection fluoroscopy (Fig. 7). Small skin incisions are madewith a scalpel blade, and a 20-gage, 15-cm radiofrequency needle with a 10-mm activetip is advanced to the anterolateral aspect of the vertebral column. Aspiration for bloodand cerebrospinal fluid is performed to make certain that the needle is not in a bloodvessel or the intrathecal space. Further confirmation is provided by injecting radiocon-trast dye to outline the sympathetic chain. Stimulation is then performed with 50 Hz at0.8–1.0 V of radiofequency stimulation to ensure that no somatic nerve is involved.The stylet of the needle is removed, and a guide wire inserted through the needle. Theneedle is removed, and a 12-gage,10-cm cannula is advanced until it makes contactwith the anterolateral aspect of the vertebra at the contrast site outlining the sympa-thetic chain. A side-firing Ho:YAG laser probe is passed through the cannula, and laserheat is applied at 5 Hz, 10 W for a total of 90–l00 J. The laser probe is rotated superi-orly, medially, inferiorly, and laterally to thermocoagulate the sympathetic chain at allthree levels.

Application of Ho:YAG laser heat or radiofrequency heat is more precise and hasfewer complications than the use of a neurolytic solution for chemical sympathectomy.There is no spread to the psoas muscle, somatic nerves, or subarachnoid space; noureteral strictures occur; hypotension is less frequent; postsympathectomy sympathalgiais virtually absent; and impotence is rare. Potential complications include puncture ofmajor blood vessels or the renal pelvis, genitofemoral neuralgia, perforation of the disc,and puncture of the ureter. Fewer postoperative thromboembolic phenomena occur in theelderly, because an operation and bed rest are avoided. The procedure can be repeatedwith minimal morbidity, and anatomical landmarks are not altered if a repeat procedure

266 Chiu and Savitz

is needed. By comparison, surgical or chemical sympathectomy induces an extensivefibrous reaction and obliterates the potential space in which the sympathetic chain lies,making the space impossible to identify during a subsequent procedure.

REFERENCES

1. Townes CH and Schawlow AI. Microwave Spectroscopy. Dover, NY:1975.2. Hecht J. Laser Pioneers. Academic, NY:1992.3. Ascher PW. Status quo and new horizons of laser therapy in neurosurgery. Lasers

Surg Med 1985;5:499–506.4. Yonezawa T, Onomura T, Kosaka R. The system and procedures of percutaneous intradis-

cal laser nucleotomy. Spine 1990;15:1175–1185.5. Davis JK. Early experience with laser disc decompression: a percutaneous method. J Fla

Med Assoc 1992;79:37–39.6. Onik G, Mooney V, Maroon JC, et al. Automated percutaneous discectomy: a pro-spective

multi-institutional study. Neurosurgery 1990;26:228–232.7. Choy DS, Case RB, Fielding W, et al. Percutaneous laser nucleolysis of lumbar disks. N Engl

J Med 1987;317:771, 772.8. Kambin P. Arthroscopic Microdiscectomy, Urban & Schwarzenberg, Baltimore, 1991.9. Sherk HH, Black J, Rhodes A. Laser discectomy. Clin Sports Med 1993;12:569–577.

10. Hellinger J. Technical aspects of the percutaneous cervical and lumbar laser disc decom-pression and nucleotomy. Neurol Res 1999;21:99–102.

Laser in Spinal Surgery and Pain Management 267

Fig. 7. Fluoroscopic view of multiple needle placements at L2, L3, and L4 levels for lasersympathectomy.

11. Hendler N, Raja SN. Reflex sympathetic dystrophy and causalgia, in Handbook of Pain Man-agement (Tollison CD, Satterthwaite JR, Tollison JW, eds.), Williams & Wilkins, Baltimore:1994, pp. 484–496.

12. Choy DSJ. Percutaneous laser disc decompression (PLDD): 12 years’ experience with 752procedures in 518 patients. J Clin Laser Med Surg 1998;16:325–331.

13. Ohnmeiss DD, Guyer RD, Hochschuler SH. Laser disc decompression: the importance ofproper patient selection. Spine 1994;19:2054–2058.

14. Simons P, Lensker E, von Wild K. Percutaneous nucleus pulposus denaturation in treat-ment of lumbar disc protrusions. Eur Spine J 1994;3:219–221.

15. Schatz SW, Talalla A. Preliminary experience with percutaneous laser disc decompressionin the treatment of sciatica. Can J Surg 1995;38:432–436.

16. Liebler WA. Percutaneous laser disc nucleotomy. Clin Orthop 1995;310:58–66.17. Siebert WE, Barendsen BT, Tollgard J. Percutaneous laser disc decompression. Experience

since 1989. Orthopade 1996;25:42–48.18. Casper GD, Hartman VL, Mullins LL. Results of a clinical trial of holmium:YAG laser in

disc decompression utilizing a side-firing fiber: a two-year follow-up. Lasers Surg Med1996;19:90–96.

19. Nerubay J, Caspi I, Levinkopf M. Pecutaneous carbon dioxide laser nucleolysis with 2 to5 year follow-up. Clin Orthop 1997;312:45-48.

20. Pedachenko EG, Khizhniak MV, Tanaseichuk AF, et al. Laser puncture diskectomy indiscogenic lumbosacral radiculitis. Lik Sprava 1998;1:143–145.

21. Hellinger J. Erfahrungen mit der perkutanen laserkoagulation des discus intervertebralis.Orthop Mitteilungen 1991;3:157.

22. Hellinger J. Ein neuer Weg der Bandscheibenchirurgie. Arztliche Praxis 1992;44:21 22.23. Hellinger J. Percutaneous laser disc decompression, in: Lasers at the Dawn of the 3rd

Millennium, Antypas G, ed. Monduzzi, Bologna, 1996, pp. 201–205.24. Hellinger J. Die Laserosteotomie als Zugangsmöglichkeit zur lumbalen und zervikalen

perkutanen Nukleotomie. Lasermedizin 1992;8:105.25. Hellinger J, Kornelli H, Einenkel RN. Der computerisierte spine-motion-Test mit integriertem

Rücken-Muskel-EMG zur Quantifizierung lokaler vertebragener Befunde, in Wirbelsäulendi-agnostik, Schmitt E, Lorenz R, eds. Enke, Stuttgart, 1998, pp. 32–36.

26. Craig F. Vertebral body biopsy. J Bone Joint Surg 1956;38:93–102.27. Castro WHM, Halm H, Schinkel V. Neodymium-YAG 1064 nm Laservaporisation von

lumbalen Bandscheibenvorfällen: Klinische Frühergebnisse, Laser 92, Shaker, Aachen1992, pp. 187–189.

28. Chiu J. Endoscopic lumbar foraminoplasty. In: Kim D, Fessler R, Regan J, eds, EndoscopicSpine Surgery and Instrumentation. Thieme Medical, NY:2004; pp.212–229.

29. Choy DS, Ascher PW, Ranu HS, et al. Percutaneous laser disc decompression: a new thera-peutic modality. Spine 1992;17:949–956.

30. Choy DS, Case RB, Fielding W, et al. Percutaneous laser nucleolysis of lumbar disks. N EnglJ Med 1987;317:771, 772.

31. Savitz MH. Same-day microsurgical arthroscopic lateral-approach laser-assisted (SMALL)fluoroscopic discectomy. J Neurosurg 1994;80:1039–1045.

32. Stöhr M. In Iatrogene Nervenläsionen, Thieme, Stuttgart, 1996, pp. 47–52.33. Suseki K, Takahashi Y, Takahashi K. Innervation of the lumbar facet joints: origins and

functions. Spine 1975;22:477–485.34. Bough B, Thakore J, Davies A. Degeneration of the lumbar facet joints: arthropathy and

pathology. J Bone Joint Surg 1999;72A:275, 276.35. Kantha S. Laser lumbar facet rhizotomy. J Minim Invasive Technol 2001;1:31, 32.36. Katzman S. Arthroscopic laser assisted facet joint denervation. J Minim Invasive Technol

2003;3:48–50.

268 Chiu and Savitz

37. Chiu JC, Clifford TI, Negrer F, et al. Microdecompressive percutaneous discectomy: spinaldiscectomy with new laser thermodiskoplasty for nonextruded herniated nucleus pulposus.Surg Technol Int 1999;VIll:343–351.

38. Knight M, Goswami A, Patko J, Buxton N. Endoscopic foraminoplasty: an independentprospective evaluation, in Laser in the Musculoskeletal System (Gerber BE, Knight M,Seibert WE, eds.), Springer-Verlag, Berlin, 2001, pp. 320–329.

39. Haag M. Transforaminal endoscopic microdiscectomy: indications and short-term tointermediate-term results. Orthopade 1999;7:615–621.

40. Clifford T, Chiu JC, Rogers G. Neurophysiological monitoring of peripheral nerve functionduring endoscopic laser discectomy. J Minim Invasive Spinal Technol 2001;1:54–57.

41. Chiu J, Clifford T. Microdecompressive percutaneous discectomy: spinal discectomy withnew laser thermodiskoplasty for non extruded herniated nucleus pulposus. Surg TechnolInt 1999;VIII:343–351.

42. Chiu J. Anterior endoscopic cervical microdiscectomy. In:Kim D, Fessler R, Regan J, eds,Endoscopic Spine Surgery and Instrumentation. Thieme Medical, New York: 2004; pp48–58.

43. Ewing M. The history of lumbar sympathectomy. Surgery 1971;70:790–796.

Laser in Spinal Surgery and Pain Management 269

14Minimally Invasive Techniques

in Pain Management

James Reynolds, MD and Garrett Kine, MD

INTRODUCTION

Epidurals were the first minimally invasive technique for the relief of lumbosacralpain. This spurred the development of the subspecialty of pain management in anesthesia,which has evolved and added many techniques to alleviate or reduce lumbosacral pain.Specialists including physiatrists, orthopedic surgeons, and neurosurgeons also performthese procedures. This chapter discusses the proven techniques used to relieve lumbosacralpain, and the clinical presentation of herniated discs, spinal stenosis, facet syndrome,and pain arising from the sacroiliac joint and coccyx. Pain may arise from one or all thestructures in the lumbosacral spine, complicating treatment. The technique for injectionsor other proven pain-relieving methods are described. Newer techniques for the treatmentof the painful disc are in the early stages of development and have not completelyevolved. The therapeutic benefits have not been proven by randomized blinded studies.These techniques are discussed but not described in detail.

THE FACET JOINT

The facet joint is a synovial joint between the articular processes of the laminae. Thejoint has an intracapsular superior recess and an extra-articular inferior recess. Thefacet joint will accept 1 to 2 mL and excess will extravasate into the epidural space, notinto the paraspinal muscle (1). Problems in the facet joint account for 15% of patientswith low-back pain (2). The patient with low-back pain arising from the facet joint doesnot have clinical features that discriminate the facet joint from other causes of low-backpain (3). There are no findings on clinical examination that consistently identify thefacet as the cause of pain. Classic teaching saying that low-back pain with extension orwith extension and rotation is caused by the facet was not confirmed with zygoapophysealjoint block (4). Patients with pain arising from the facet do not have pain in the midline.The pain arising from the facet joint is always lateral to the midline and may radiate to thebuttocks and distally to the toes. Anesthetizing the medial branch of the zygoapophysialjoint will relieve pain arising from this joint (5). The medial branch can be anesthetized byinjecting anesthetic solution at the base of the transverse process just above the midpoint.

271

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

Contrast must be used to avoid an intravascular injection, a complication that occurs8% of the time (5,6). Degenerative changes identified on computed tomography (CT) donot correlate with the presence or absence of pain arising from the zygoapophyseal joint(2). Discogenic pain and pain arising from the zygoapophyseal joint do not usually occurin the same patient (7). Disc degeneration always occurs before facet degeneration (8).Facet joint injections do not show good long-term relief of low-back pain (9). Random-ized, placebo-controlled trials have not shown a significant difference between corticos-teroid injections and placebo (10). However, there are a small number of patients whor*spond very well to facet injections and benefit from injections every 3–6 mo.

RADIOFREQUENCY MEDIAL BRANCH ABLATION

Patients with short-term relief of low-back pain from facet injections may be givenlonger relief by radiofrequency medial branch ablation. Pain arising from the facet jointshould be proven by two injections using anesthetic agents with different durations ofaction. Lidocaine has a rapid onset and brief duration of action. A confirmatory injectionwith bupivacaine should have a longer duration of action of at least 3 h. The patientwith at least 50% relief of low-back pain from injections of both the short- and longer-lasting anesthetic agents are considered appropriate responders. The injection may be inthe joint of the facet or at the medial branch (3). Provocation of low-back pain by injec-tion of contrast into the facet joint is not a good indicator of the facet joint as the sourceof the low-back pain. The relief of pain by anesthetic agents injected into the joint or atthe medial branch to the facet is the only appropriate indicator that the facet is thesource of pain in the lower back (11). There is conflicting evidence that radiofrequencyablation of the medial branch to the facet in the lumbar spine relieves low-back pain(12,13). When inadequate diagnostic criteria were used to select patients to undergoradiofrequency facet joint denervation, the results of facet joint denervation were notshown to be effective in the relief of low-back pain (14). When a single diagnostic injec-tion was used, radiofrequency lumbar facet denervation was shown to be more effectivethan a sham procedure (15). When very strict diagnostic criteria were used, includingtwo diagnostic medial branch blocks with a very high percentage of pain relief, excellentrelief of low-back pain was obtained, with 87% of patients experiencing at least a 60%reduction in their visual analog scale (VAS) (16). Relief of back pain was maintained forat least 6 mo, but the majority of patients sustained relief for 12 mo. No complicationsoccurred in the study (16). Other studies have shown good relief of pain from repetitionof the radiofrequency denervation of the lumbar facet joint (17). The patient undergoingradiofrequency denervation of the lumbar facet joint experiences an increase in low-backpain for 2–4 wk after the procedure. Therefore, it is best to wait until the low-back pain hasreturned to significant levels before recommending a repeat radiofrequency neurotomy ofthe lumbar facet. Patients experiencing satisfactory relief from the injection of corticos-teroids in the facet joint for 3 mo can be managed with quarterly injections until therelief from the facet injection no longer lasts 3 mo.

EPIDURAL STEROID INJECTION INTO LUMBAR SPINE

There are three portals to deliver steroids to the epidural space in the lumbar spine: inter-laminar, caudal, and transforaminal. Blind placement of epidural steroids into the epiduralspace in the lumbar spine through the caudal or interlaminar route results in misplacement

272 Reynolds and Kine

of the corticosteroid in 14–38% of lumbar epidural corticosteroid injections. The use ofcontrast during the epidural procedure assists in visualizing the proper placement of thecorticosteroid solution and prevents intravascular injection of the solution (18–20).Transforaminal lumbar epidural injection is the most precise technique for delivery ofthe corticosteroid into the lumbar epidural space (21). Selective nerve root block andtransforaminal epidural lumbar injection differ in the amount and placement of thesolution into the lumbar epidural space. Selective nerve root block is only 1 to 2 mL ofvolume and is meant to be diagnostic as well as therapeutic. The bevel of the needlefaces distally and is placed within the epiradicular membrane. Diagnostic selectivenerve root injection results in a minimal increase in the risk of neural injury and reducesthe chance of a false negative (22).

The efficacy of therapeutic epidural injections for significant long-term improvementhas been questioned, but in such studies fluoroscopic imaging and contrast were not usedto ensure that the corticosteroid reached the pathology (23,24). Studies that showedepidural steroids to be effective all were reformed with fluoroscopic guidance through atransforaminal approach. The use of a single injection through a periradicular corticos-teroid injection of a lumbar nerve root showed short-term improvement compared with aninjection without corticosteroids (25). Studies that allowed one to four injections had afavorable long-term outcome. Seventy-five percent of patients had a 50% or greaterreduction in pain and a return to normal or near-normal activity (26). Other studiesshowed similar relief of sciatica, with 77% of patients avoiding surgery (27). A singleperiradicular membrane injection of the L4, L5, or S1 nerve roots containing corticos-teroid was superior to the same injection without corticosteroid for patients with a con-tained herniated nucleus pulposus. Fewer than half of the patients with the containedherniated disc who received injections with corticosteroids chose to undergo surgerycompared with the control subjects. The result was a cost savings of $12,666 perpatient. Patients with an extruded herniation did not experience statistically significantimprovement compared with the control subjects (28). A randomized double-blindstudy showed that selective nerve root injections reduced the number of patients electingto undergo decompression in the lumbar spine for spinal stenosis. The patients receivedselective nerve root injections with anesthetic agent with or without corticosteroid. Theycould elect to receive as many as three injections, but always with the same solution—eitherwith or without corticosteroids. The group receiving injections with corticosteroids electedto undergo surgical decompression less than half as frequently as the control group (29).

Infiltration of a lumbosacral nerve root was shown to be predictive of a successfulsurgical procedure when the patient had reproduction by needle placement and relief ofsymptoms with infiltration of an anesthetic agent (30). Injection of anesthetic agentsand corticosteroid in the perineural membrane of a lumbar nerve root was shown to beuseful in predicting the response to surgical procedures in patients with radicular symptomsin the lumbar spine present for more than 1 yr. For those patients who experiencedrelief of radicular pain during the anesthetic phase and relief for 1 wk from the corticos-teroids, 85% had a successful surgical outcome. Those patients without relief from thecorticosteroid phase had an unsuccessful outcome 95% of the time (31).

The International Spinal Injection Society has specific recommendations for the fre-quency and number of spinal injections. Injections should be 7–14 d apart, and no morethan four in a 6 mo period (31,32). The potential side effects and possible complications

Techniques in Pain Management 273

include vasovagal reactions, facial flushing, insomnia, and bleeding (32). These effectsare reactions to medication rather than true complications. Other procedure-relatedcomplications are dural puncture and associated headaches, infection within the epiduralspace, nerve injury, bleeding, transient weakness or numbness, and paraplegia (32).Adverse reactions can include reaction to the contrast and fluid retention with associatedperipheral swelling. Contraindications are known hypersensitivity to medications usedin the injection, malignancy, infection, bleeding diathesis or anticoagulation, congestiveheart failure, or uncontrolled diabetes (32). Celestone™is a corticosteroid containingbetamethasone, and the flare resolves faster than prednisolone or triamcinolone acetate,thus making Celestone the preferred corticosteroid for epidural injections.

SACROILIAC INJECTIONS

The sacroiliac joint is the great imitator. It can produce groin pain that mimics thepain of a degenerative hip joint, back pain similar to discogenic pain, and pain radiatingon the posterior portion of the leg even below the knee that is similar to sciatica (33). In14% of patients, pain from the sacroiliac joint was referred to the foot (34). Anatomicalstudies of innervation of the sacroiliac joint using special staining techniques have beenconducted in animals. The sensory nerve fibers to the dorsal side of the sacroiliac jointarise from L4 to S2 and from the ventral side from L1 to S2 (34,35). The L4 and L5nerve roots converge to begin the formation of the lumbosacral trunk. As this portion ofthe lumbosacral trunk reaches the pelvic brim, the combined roots of L4 and L5 arewithin 5 mm of the sacroiliac joint (36). No study has determined whether the cause ofthe pain is referred pain, direct irritation of the lumbosacral plexus, or another source.

Physical examination does not correlate or always lead to the diagnosis of sacroiliacjoint pain. Classic physical examination of the sacroiliac joint includes the Patrick test,also known as the FABER test or figure 4 test; Gaenslen’s test; and the posterior sheartest (37–39). CT showed abnormal joints in 57.5% of patients with pain arising fromthe sacroiliac joint and 31% of matched control subjects (40). In the intact pelvis, thedegree of motion of the sacroiliac joint was <2° (41). Even in patients with postpartumsacroiliac pain, no more than 2º of motion on any axis could be demonstrated (42). Theonly symptom that correlated with sacroiliac pain proven by diagnostic blocks wasgroin pain (33). Others have found the sacroiliac as the source of persistent pain after alumbosacral fusion. In a group of patients undergoing diagnostic injection to determinethe source of persistent lumbosacral pain after lumbosacral fusion, 32% of patients werediagnosed with sacroiliac pain as the cause of persistent pain after lumbosacral fusion.Seventy-five percent relief of pain was the criterion used to determine that the sacroiliacjoint was the remaining cause of pain that persisted after a lumbosacral fusion (43).

Injection of the sacroiliac joint is difficult without image guidance. Injection into thesacroiliac joint was successful in only 22% of attempts done without image guidance.CT performed immediately after blind sacroiliac injection showed that in 24% ofpatients the dye entered the epidural space (44). Injection of the sacroiliac joint ofpatients with inflammatory spondyloarthropathy has been shown to be effective forshort-term relief of 1–3 mo for symptoms arising from the sacroiliac joint, but it did notprovide significant relief at 6 mo (45). In patients without spondyloarthropathy, injectionof a corticosteroid was shown to be effective for relief of pain arising from the sacroiliacjoint when compared with an injection of saline (46). Injection of the painful sacroiliac

274 Reynolds and Kine

joint resulted in an improvement in function for more than 1 yr. The number of injec-tions per patient was 2.1 (47). The use of radiofrequency for denervation of the sacroil-iac joint is being explored. Of patients undergoing such a procedure, 36.4% reported atleast a 50% reduction in pain for at least 6 mo, with an average duration of relief of 12mo (48). Many details must be considered when performing such a procedure, and theefficacy is still to be determined.

THE COCCYX

Coccydynia historically has been viewed as having a psychological origin, but a verycomprehensive study showed “no evidence of neurosis” (49). One of the factors thatcontribute to this myth is the common association of an additional source of pain. Studieshave shown that as high as 77% of patients with coccydynia have associated conditionscausing back pain. These conditions include painful degenerative discs, herniated discs,and sacroiliac dysfunction (50). The symptoms of coccydynia are classic and localizedto the “tail bone.” Palpation of the coccyx is always painful. Injections of the sacrococ-cygeal joint should give good relief of pain arising from the coccyx. The patient will beable to distinguish pain arising from another source as being a different pain from coc-cydynia. Between 60 and 75% of patients with coccydynia will get significant relief ofpain from sacrococcygeal joint injections (49,50).

TECHNIQUES OF LUMBAR SPINAL BLOCKADE

IntroductionIntent

Injection in and around the spine may be performed with both diagnostic and therapeuticintent. As in any diagnostic test, there exists the possibility of both true positive and falsepositive responses as well as true negative and false negative responses. Injections may beused as a “rule-out” procedure. In this instance, if a properly performed procedure in anawake, cooperative patient does not produce relief of the regionalized pain, then, more thanlikely, it is a true negative response. That particular structure is not a significant pain genera-tor in this individual. A false negative response may occur in an individual with multiplepain generators. In addition, there exists the possibility of multiple behavior and psycholog-ical issues that predispose a patient to perceive that he or she is actually not feeling better.

Interpretation of positive responses carries with it inherent difficulties as well. Falsepositive responses may occur owing to placebo response, anesthetizing of structuresother than the target, and other considerations. The use of concomitant intravenous med-ications is controversial during the performance of diagnostic blocks and adds yetanother complex factor to interpretation of the results. All this having been said, truepositive responses can be very rewarding to both the patient and the medical practitioner.These true positive responses lead to a diagnosis; lend credence to the patient’s paincomplaint; and may lead the patient down the path to cointerventions with the ultimategoal of recovery, or at least a higher functioning and more satisfying standard of living.

Scientific inquiry is based on the notion of quantifying data, introducing a singlevariable, then carefully observing the resultant change. In the case of diagnostic injectionprocedures, careful preblock condition of the patient should be assessed with provocativemaneuvers, the block should be performed accurately with proper fluoroscopic visual-ization, and then the results should be evaluated and quantified through the same

Techniques in Pain Management 275

provocative maneuvers. Independent assessment of the individual by a qualified personshould yield more accurate data.

The therapeutic intent of injection procedures has been the mainstay of their existencefor many decades. The duration of the perceived benefit is variable. Often, injectionsgive a window of relief where cointerventions may be administered. Injections are alsooften used in conjunction with physical therapy. In addition, injections may help peopleover very difficult times of flare, and they may be employed as a maintenance therapywhen all other available options have been exhausted.

Disclaimer

The techniques described as follows are those of the authors. Certainly, other techniquesexist that are safe and sound. These techniques are generally consistent with thoseoffered by the guidelines of the International Spinal Injection Society.

This chapter is not a substitute for proper training. Any individual performing theseinjections should be fully cognizant of the medications being used, be aware of theirpotential side effects, and be prepared for any potential adverse reactions. Resuscitativemediations and equipment should at all times be immediately available.

Lumbar Transforaminal EpiduralsBackground

Transforaminal epidurals have also been called selective epidurals or selective nerve rootblocks. The transforaminal route has emerged in the last couple of decades as a generallysuperior technique compared to the interlaminar and caudal routes for administration ofmedications to the spinal canal. It has the advantage of enabling the physician to deliverthe injectate to the ventral epidural space, in closer proximity to disc pathology. Theavailability and increasing quality of fluoroscopy equipment have certainly also led tothis emergence. Although interlaminar and caudal epidurals may certainly be attemptedwithout radiological guidance, fluoroscopy has enabled verification of placement ofmedication and minimized several safety concerns.

Positioning

The patient is generally placed prone on the fluoroscopy table. A pillow under theabdomen may help open the foramen and contribute to patient comfort. Although the patientcould be placed in an oblique position, a C-arm fluoroscopy machine negates the need forthis. Prone positioning also allows for bilateral procedures without the need to reposition.

Imaging

In the standard subpedicular technique, the target point is the “safe” triangle. A needleplaced here is just inferior to the pedicle, lateral to the dural sleeve, and medial to animaginary line dropped from the most lateral aspect of the pedicle. The pedicle is oftendescribed as a clock and the optimal needle position would then be 6:00. It is better toerr on the safer side, and this is sometimes described as being at the 5:30 position on theright of the spine and the 6:30 position on the left.

The C-arm should be adjusted in a cephalocaudal angulation so that the inferior edgeof the pedicle is a sharp line. This should correspond to also having the X-ray beamsparallel to the inferior and superior end plates of the vertebral body. The C-arm can thenbe moved through slight obliquity to see if this target area can be reached without any

276 Reynolds and Kine

Techniques in Pain Management 277

Fig. 1. (A) Anteropostenor (AP) view of left L4 transforaminal epidural utilizing single-needletechnique. (B) Lateral view of L4 transforaminal epidural. Contrast is seen to travel in ventralepidural space more in a cephalad than a caudad direction. A filling defect at L4-L5 suggests aprotrusion. (C) AP view of left L5 transforaminal epidural with double-needle technique.

bony interference. Sometimes the inferior edge of the transverse process, the lamina, orthe superior articular process of the facet may interfere with a direct shot at the “safe”triangle. Slight C-arm adjustments will evaluate this. A single straight-needle technique

can be used if there is no osseous encroachment (Figs. 1A,B). Otherwise, a double-nee-dle technique can be used to direct a curved needle through an introducer to go aroundany bony obstacles (Fig. 1C).

Once the needle is placed at the target point, the initial injection should be underreal-time fluoroscopy with a nonionic hypoallergenic contrast agent. An extension tubecan be used to keep the operator’s hands out of the direct X-ray beam. Care should betaken to ensure that there is no sign of intrathecal spread as well as no indication ofintra-articular injection in a radicular artery.

Needle Techniques

Generally, one should strive to use the thinnest needle possible for any blockingprocedure. With transforaminal epidurals, usually a 25- to 22-gage needle should suffice.It becomes difficult to inject through a 27-gage needle with particulate steroids.

As mentioned under Imaging, if an unobstructed view of the target area can be seen witha fluoroscopy machine, then a single straight needle should be sufficient. Placing the needleaccurately and using the bevel for slight adjustments is all that is typically necessary.However, if the aforementioned lateral border of the lamina, inferior edge of the trans-verse process, or superior articular process of the zygapophyseal joint is in the way of adirect shot to the “safe” triangle, then a double-needle technique can be employed. Thiswould also be useful if there has been surgical modification of the spine such as anintertransverse fusion mass. A 6-in., 25-gage needle fits through a 31/2-in., 20-gageintroducer. In a larger individual, or in a situation in which one really has to maintain astrong curve of the longer needle, a 6-in., 22-gage needle fits through a 31/2-in., 18-gageintroducer. This latter combination is useful, e.g., in an intertransverse fusion mass,where it helps keep the needle from going too ventral into the body.

278 Reynolds and Kine

Fig. 1. (Continued)

Injectate

After needle placement, initial injection should be of contrast only under real-timefluoroscopy. This initial injection of 0.5–1 mL will give an indication of the spread of themedication. The majority of medicine should be traveling centrally into the epiduralspace. If the majority of medicine remains external to the neural foramina, then attemptsshould be made to reposition the needle to achieve better spread of medication.

The choice of anesthetic should be based on multiple considerations. In the case oftransforaminal injections, there is the possibility of a significant motor blockade. This isespecially the case when multiple levels are being injected simultaneously. Therefore,medications such as lidocaine would often be preferred over a longer-acting agent suchas bupivacaine. It is generally recognized that optimal medical care would dictate thatan individual be kept under medical observation while any significant motor, sensory,or hemodynamic events are possible. Therefore, with the use of lidocaine, an individualcan be recovered in <1 h. Obviously, all anesthetics should be preservative free,because there is always the potential risk of an unintentional dural puncture.

A more concentrated anesthetic may be diluted with a contrast agent to achieve asolution of lesser concentration that can be visualized. An example of this would bediluting 2% preservative-free lidocaine in a 50/50 mixture with the contrast agent. Aglucocorticoid can then be added to the solution and the total mixture injected underfluoroscopic observation to help judge the extent of spread of the medication. The volumesused will vary from individual to individual. If there is diagnostic intent, such as tryingto predict the clinical effect of a single-level foraminotomy, then one would choose touse a more concentrated anesthetic with less contrast and far smaller volumes.

Other Considerations

An S1 transforaminal injection is basically the same as the subpedicular approach inthe lumbar region. The big difference is that access is achieved through the dorsal foramen.Typically, the dorsal foramen is located approx 15º oblique to the easier seen ventralforamen. Consequently, lateralizing the C-arm 15º oblique often provides better visual-ization. A straight single needle is usually all that is necessary. A slight caudocephaladangulation regarding the sacrum allows for more cephalad spread of medicine reachingtoward and above the L5-S1 disc space. The volumes necessary to reach this disc spacemay be somewhat greater than are typically found in the lumbar region.

If at any time a nerve is touched, the needle should be immediately repositioned.Often the paresthesia is very transient and there should be no continued pain. If the paindoes continue, one should consider discontinuing the procedure. Furthermore, if anunintentional dural puncture occurs or injectate in a radicular artery is noticed, oneshould consider abandoning the procedure at that point and possibly returning after theartery or dural sleeve has had adequate time to heal.

Lumbar Interlaminar EpiduralBackground

Lumbar interlaminar epidurals have also been called translaminar epidurals. This isthe traditional approach that has been used for many decades for surgical levels ofanesthesia or for labor and delivery. Most of the initial studies regarding epiduralsteroid injections were performed without fluoroscopy and were mostly uncontrolled.

Techniques in Pain Management 279

More recently, controlled studies have questioned the efficacy of the interlaminar route.Logic dictates that there may be a subset population for whom the interlaminar routeis more efficacious. An example of this is the aging population in whom neurogenicclaudication becomes symptomatic owing to central canal stenosis. A interlaminarroute should theoretically deliver 100% of the medication into the neuraxis. Controlledstudies of this patient population are still forthcoming.

Positioning

The patient is placed prone on the fluoroscopy table. Generally, a pillow or bolster isplaced under the abdomen to induce a gentle flexion. If a patient is unable to lie prone,the procedure can be performed in the lateral position with the more dominant painfulside down. It is believed that gravity may help lateralize the medication toward the siteof pathology.

Imaging

The interlaminar space is visualized in the AP projection so that the spinal processesappear end on. Some cephalocaudal angulation may be used to better visualize theinterlaminar spaces. If necessary, lateral fluoroscopic imaging can show the depth of theneedle; however, it is generally advised to visualize needle movement in the AP direc-tion to ensure that the needle tip is close to the midline.

After the needle is properly placed, the first injection should be nonionic contrastunder real-time fluoroscopy (Fig. 2). Visual inspection should include possible venousuptake as well as possible intrathecal spread of the contrast. If an inadvertent intrathecalspread is suspected, a lateral fluoroscopic image can confirm or deny this suspicion.

280 Reynolds and Kine

Fig. 2. AP view of L5-S1 interlaminar epidural with right paravertebral approach usingCrawford needle.

The classic myelogram type of pattern should be easily recognized if it is present inthe lateral imaging. The contrast materials are relatively hyperbaric compared to thecerebrospinal fluid and, therefore, slight table tilting can confirm or deny the presenceof a dural puncture.

Needle Techniques

The procedure may be performed with a midline approach or a paravertebralapproach. The paravertebral approach has the advantage of not going through thesupraspinous and interspinous ligaments. Going through these ligaments can sometimeslead to a false feeling of loss of resistance when, in actuality, the needle is too shallow.Additionally, these ligaments are supporting structures in the posterior compartment.

An entry point is made perpendicular to the lamina just inferior to the interlaminarspace. After anesthetizing the skin with a small bore needle, often a longer needle, suchas a 22-gage, 31/2-in. spinal needle, is used to anesthetize the deeper structures and isbrought in direct contact with the lamina. This gives the physician direct knowledge ofthe expected depth of the interlaminar space.

Several specially designed needles are used for the actual epidural injection. If asingle injection is desired without the use of a catheter, a Crawford-type needleshould be considered. It is specifically designed for this purpose. The shallow bevelof the needle is placed so that it is parallel to the fibers of the ligamentum flavum.The needle is directed from slightly paravertebral toward the midline of the interlam-inar space. The stylette is kept within the needle so as not to allow any organic mate-rial to potentially clog the hollow bore of the needle. With a paravertebral approach,the first real resistance that should be met at the appropriate depth is the ligamentumflavum. At this point, the stylette is removed. Most interventional specialists prefer aloss-of-resistance-to-fluid technique. This technique involves placing 1 to 2 cc ofeither normal saline or 1% preservative-free lidocaine in a standard 5-cc syringe or aspecialty syringe designed for this purpose. With the needle tip engaged in the liga-mentum flavum, there is a great deal of difficulty injecting any fluid. As the needle isslowly advanced, there will be an immediate loss of resistance to injection of thefluid. At this point, all additional needle movements should be curtailed. Care shouldbe taken so that one hand holding the needle additionally contacts the patient. Thatway, there is less risk of sudden advancement of the needle should the patient move.At this point, the loss-of-resistance syringe should with replaced with an injection ofcontrast material under real-time fluoroscopy. Extension tubing that has been purgedof air with the contrast is helpful in keeping the operator’s hands out of the directfluoroscopic beam.

At times, it is desirable to thread a catheter. Radiopaque catheters are available forthis purpose. For this procedure, a blunt curved-tip needle, such as a Tuohy needle, is used.The loss-of-resistance sequence is the same as with the Crawford needle. If unilateraldepositing of medicine is desired, one should consider performing the paravertebralapproach from the contralateral side. The point of entry is then at the midline, and typicallythe catheter will then thread to the desired side without difficulty. Catheters are generallythreaded in a cephalad direction; however, experienced hands can thread the catheter ina caudad direction as well. Catheters can often be directed toward the neural foramina ifthis is the site of pathology.

Techniques in Pain Management 281

Injectate

Initial injections should once again be solely with contrast under real-time fluo-roscopy. After confirmation of correct placement, a glucocorticoid and anesthetic mix-ture may be instilled. The volume used will vary with the patient’s size, the amount ofepidural fat and venous vasculature, and the degree of stenosis and other pathology. Atotal volume may range from 1 or 2 mL up to 10 mL or more. If multiple segmentalpain generators are suspected, the higher volumes might be employed. On the otherhand, in an elderly patient with spinal stenosis, it would be prudent to use muchsmaller volumes.

Other Considerations

The interlaminar technique generally deposits a greater amount of medicine in thedorsal epidural space rather than the ventral. If the patient’s pathology is a disc abnor-mality, certainly a transforaminal approach may prove more beneficial than an inter-laminar one. A transforaminal epidural can be purposely delivered toward the ventralepidural space.

If there is any suspicion of a dural puncture, it may prove prudent to abandon theprocedure. An injection can occur that is subdural, but not truly subarachnoid. Whenthis occurs, patients have a stronger than normal sensory block that is often delayed bysome 5–10 min. Vasodilation can occur with subsequent hypotension. Additionally,patients with blockade are at risk of falling when they first try to ambulate. They shouldbe kept in the recovery area until all signs of blockade have dissipated.

Caudal EpiduralsBackground

The caudal approach for performing epidurals has a long history. It is still being usedfor administering several types of surgical levels of anesthesia as well as in the managementof pain under certain circ*mstances.

Transforaminal epidurals have generally supplanted the use of caudal epidurals.However, there are certain clinical presentations for which a caudal may be desirable.In advanced degeneration of the spine, the interlaminar spaces may no longer be accessi-ble. In individuals with large intertransverse fusions, the transforaminal route becomesmore technically challenging. In extremely obese individuals, the caudal canal canoften be reached when standard needle equipment may not be long enough to reachother avenues into the epidural space.

Positioning

The patient is generally prone on the fluoroscopy table. A meticulous prep and drapeshould be employed to minimize the risk of any infection.

Imaging

Generally, AP imaging is first used to ensure midline placement of the needle. Then,if necessary, lateral fluoroscopic imaging can confirm whether or not the needle hasentered the sacral hiatus (Fig. 3). Again, initial injections should be with contrastbecause quite often there is vascular uptake with initial placement of needles. The ini-tial volume of injectate will help determine what volume should ultimately be used.There is a fair amount of so-called dead space before any glucocorticoid would reachthe L5-S1 disc space.

282 Reynolds and Kine

Needle Techniques

After the meticulous prep and drape, a 22- or 23-gage spinal needle can be used. Acaudal epidural can often be performed with manual palpation of the bony structurefollowed by confirmation of needle placement with fluoroscopy.

The midline sacral prominences are felt with one hand. As the structures are palpatedcaudally, usually around S4, the sacral promontory is felt. The 22- or 23-gage needle isthen placed slightly caudad to the sacral hiatus. The bevel of the needle is faced downso that it may glide more easily over the osseous structures. As the sacral hiatus isapproached, the bevel of the needle is then placed face up to help guide it within the caudalcanal. Contrast is injected and outlining of the sacral nerve roots is typically seen. Oneshould carefully observe for possible vascular uptake of contrast. An inadvertent duralpuncture should be highly unlikely.

Injectate

After placing the contrast-confirming caudal epidural spread, a solution of glucocorticoidwith anesthetic can be placed. Mixing the glucocorticoid with a small amount of anestheticmay be considered. The anesthetic may be diluted with contrast so that it can be ascer-tained where the medicine travels. After this first aliquot of glucocorticoid with anestheticand contrast (approx 5 mL) is delivered, it may then be followed by either dilute anestheticor additional contrast. The concept is to push the glucocorticoid solution more cephaladup the epidural tree to its desired location.

Other Considerations

Caudal epidurals can also be placed with a catheter. An advantage is that one candeliver the medicine more cephalad and more likely closer to the area of pathology.

Techniques in Pain Management 283

Fig. 3. Lateral view of caudal epidural with 22-gage spinal needle. Contrast is seen in thesacral caudal canal.

Catheters can often be steered toward the dominant site of pathology and also may bedirected toward the ventral or dorsal epidural space. The disadvantage of a catheter is thatit generally requires a larger-bore introducer needle. Care should be taken to minimizescraping the periosteum, because this can be painful.

Sacroiliac Joint BlocksBackground

The idea that the sacroiliac joint could be a potential source of pain has been debatedthroughout the last century. Over the past several decades, technology has advanced toa point where injections into the sacroiliac joints can be performed both safely and con-fidently. The primary reason for performing the injection is to obtain a diagnosis. It isgenerally believed that there is no set of clinical features or manual examinations thatcan accurately predict pain of sacroiliac origin. To date, there is no convincing evidencethat the use of glucocorticoid will consistently produce long-term relief. However, thereis anecdotal evidence that short-term relief may often occur.

Positioning

The patient is prone on the fluoroscopy table. A pillow bolster under the abdomenmay make for a more comfortable experience for the patient and help to correct extremelordosis if this is present.

Imaging

The most consistent way to enter the sacroiliac joint is at its inferior aspect. An APview, perpendicular to the sacrum, often gives the most initial information. It should beremembered that the image is a two-dimensional interpretation of a three-dimensional(3D) joint. The more medial radiolucency represents the dorsal aspect of the joint, andthe more lateral is the more ventral aspect. Often, these two joint surfaces converge 1 to2 cm above the most inferior aspect of the joint. If this bright radiolucency occurs at theconvergence of the joints, this is often an easy access to the joint (Fig. 4).

If this radiolucency at the convergence of the joint is not immediately seen, slightrotation through the C-axis may make it more prominent. Additionally, this allows theoperator to study the joint to develop a 3D picture of the approximate shape of thejoint. Some physicians find it beneficial to rotate through the C-axis to line up the twojoint surfaces. Others find it more useful to separate out the dorsal and ventral surfacesof the joint. When rotating the C-arm contralaterally through its axis, thereby separat-ing out the joints, often the medial cortical line of the joint silhouette becomes sharp. Whenthis is obtained, it represents a direct straight-line access into this medial inferior part of thejoint. If all of these maneuvers through the C-axis do not clearly demonstrate the inferioraspect of the joint, slow cephalocaudad angulation of the C-axis may be useful.

A typical sacroiliac joint holds about 1.5 mL of solution intra-articularly. Whenviewed under real-time fluoroscopy, a fully filled joint can have a “classic lightning bolt”appearance (Fig. 5). The joints often leak with capsular tears that may place additionalmedicine medially in close proximity to the dorsal sacral foramina, superiorly near the L5nerve root, and ventrally in proximity to the lumbosacral plexus.

When leak of contrast is observed, the joint can be visualized in the AP to determineleaks from the superior or inferior ends. In addition, an oblique “en-face” view can beobtained to see the shape of the joint margins.

284 Reynolds and Kine

If all attempts at entering the inferior aspect of the joint 1 to 2 cm above the mostinferior edge fail, injections can be attempted into the superior or inferior capsularrecesses. On rare occasions, it may be necessary to try to enter the midpoint of thejoint. Studying a CT axial image will help guide the practitioner in this endeavor.

Techniques in Pain Management 285

Fig. 4. Placement of 25-guge needle in radiolucency of left sacroiliac joint.

Fig. 5. Arthrogram of sacroiliac joint.

Typically, the needle must traverse medially to laterally as well as somewhat inferiorlyto achieve entry.

Needle Techniques

A 31/2-in., 25-gage needle should be sufficient for most needs. The depth of the inferioraspect of the joint is such that rarely more than a small percentage of the needle mustpenetrate the skin. Care should be taken to strike the sacrum first and then pass the needlemore laterally into the joint. This prevents inadvertent passage of the needle through thesciatic notch and into the pelvic cavity.

If on first injection there is difficulty, one should try turning the bevel of the needle sothat the bevel is facing cephalad toward the plane of the joint. A 3-cc syringe with aLuer-Lok should be used because this joint requires very little volume and occasionallymodest pressurization is needed to inject through a small-bore needle such as a 25 gage.

Injectate

As previously mentioned, the sacroiliac joint on average holds about 1.5 cc of solution.If an optimal quality arthrogram is desired, one should consider injecting just contrastfirst. It is important to remember to keep the volumes low (0.5–1 cc) so as to leave roomfor the anesthetic. Alternatively, the contrast and anesthetic may be mixed together. Asomewhat stronger anesthetic such as 2% lidocaine may be considered in the hope ofcreating a sufficient block. A glucocorticoid may be included in the mixture, withsomewhat guarded expectations.

Other Considerations

Sacroiliac joint blocks are regarded as primarily a diagnostic tool. A diagnosis ofsacroiliac dysfunction should be considered when the most intense pain is found belowthe L5 level, when the patient points directly to the posterior superior iliac spine, and ifthere is tenderness on palpation of the sacral sulcus just medial to the posterior superioriliac spine. These signs should also be considered in a patient who has had a previouslysuccessful arthrodesis of the lumbar spine that then deteriorates over time.

Medial Branch BlocksBackground

Blockade of the medial branch of the dorsal primary ramus has been advocated asthe best method for determining pain of posterior element origin. It has a lesser potentialfor false positive response compared to true intra-articular injection of the facets. It hasalso been used as a screening procedure prior to radiofrequency neurotomy.

Performing the blocks on different occasions using anesthetics of different durations hasbeen advocated to minimize the false positive rate of a single block. Although certainlythis paradigm is advantageous in a scientific study, in clinical practice, it leads to aquandary. A placebo responder may have true pathology. Therefore, a potentially thera-peutic procedure, such as radiofrequency neurotomy, may be denied to an individualowing to the fact that he or she simply exhibits a placebo response. Another school ofthought would then say that in a known placebo responder, quite simply, to see if a neuro-tomy will be useful, the neurotomy should be performed. A neurotomy has been shownto be safe and reproducible in experienced hands.

Perhaps medial branch block testing has as its strongest virtue the ability to rule out asource of pain. A carefully performed, fluoroscopically controlled procedure that does notdecrease the regionalized pain directs the practitioner to search for other potential sources.

286 Reynolds and Kine

The nomenclature and anatomy for the procedure should be well understood. Eachjoint is innervated by two medial branches. For example, to anesthetize the L4-L5zygopophyseal joint, it is necessary to anesthetize the L3 and L4 branches. These L3and L4 branches are found at the base of the superior articular process where it meetsthe transverse processes of L4 and L5. At the L5 level over the sacrum, the anatomy issomewhat different. The medial branch has not yet separated from the intermediate andlateral branches. It is generally agreed on that there is no consistent branch from the S1neural foramina innervating the L5-S1 joint.

Positioning

The patient is prone on the fluoroscopy table. A pillow or bolster may be used toreduce the lordosis.

Imaging

The patient is visualized in the AP view with cephalocaudad angulation used to alignthe end plates with the fluoroscopy beam (Fig. 6). Then, slight oblique C-arm movementis used to best visualize the superior articulating process. It is also necessary to visualizethe superior border of the transverse process.

A small amount of contrast should be mixed with the anesthetic (Fig. 7). Visualizationof the block should occur under real-time fluoroscopy to ensure that there is no vascularuptake. If necessary, lateral imaging will ensure that the medication is sitting in the valleywhere the superior articulating process and transverse process meet. Occasionally themedicine is deposited dorsally, so if visualization is only in the AP or oblique modes,this can be checked by obtaining a more oblique view or a full lateral.

Needle Techniques

Generally, a very small-bore needle is all that is necessary. A 25-gage needle shouldbe sufficient. Occasionally, in an extremely obese patient, it may be necessary to use aslightly greater-diameter needle.

Care should be taken to use minimal anesthetic to reach the target point. A 1-ccsyringe with anesthetic may be used and the anesthetic administered in minute quan-tities as the needle is passed through the skin and tissue levels. This will ensure thattoo much anesthetic is not used, thereby eliminating the possibility of a false positiveresponse. The target point is 2 to 3 mm below the superior edge of the transverse pro-cess at the junction of the superior articulating process. If the needle target point istoo close to the superior edge, there is a chance that the medicine will track mediallyalong the nerve root path toward the neural foramen. If the needle is placed too cau-dad along the transverse process, the chance of not properly anesthetizing the nervebecomes greater.

Injectate

Generally, 0.2–0.4 cc of total injectate is needed. Commonly used is 2% lidocaine,mixed in a 50/50 mixture with contrast or 0.75% bupivacaine diluted with contrast.There is no generally accepted use of glucocorticoid for this injection.

Other Considerations

Medial branch blocks are a diagnostic tool. Meticulous pre- and postblock evaluationsare mandatory. Optimally, the patient is blinded as to the duration of the anesthetic.Pain assessment should be guided by a trained professional.

Techniques in Pain Management 287

288 Reynolds and Kine

Fig. 6. AP view of needle placement for left L4 and L5 medial branch blocks.

Fig. 7. Oblique view of contrast and anesthetic injected for L4 and L5 medial branch block.

Occasionally, it will be impossible to block the medial branch accurately. One exampleof this is in the individual who has had a previous arthrodesis. In such an individual,there may exist an intertransverse fusion mass preventing access to the expected locationof the medial branch. In this instance, it may prove useful to perform an intra-articular

zygopophyseal joint block at this first mobile segment. Naturally, this fusion masswould also interfere with the performance of a medial branch neurotomy.

Facet (Zygapophyseal Joint) BlocksBackground

Facet joint injections have been performed for many decades. Many clinical signsand symptoms are suggestive of posterior element pain, but none are 100% diagnostic.Currently, medial branch blockade is thought to be the better diagnostic tool as comparedto intra-articular injections. There is always the chance of spillover of the joints intoother structures, leading to a false positive. However, there are occasions when medialbranch block testing cannot be physically performed.

Positioning

The patient is generally prone on the fluoroscopy table. However, an oblique positioningis sometimes useful, particularly if the plane of the joint is especially coronal in orientation.Additionally, with oblique positioning, it is possible to rotate the patient’s shouldersback while maintaining the position of the hips, thereby partially subluxing the jointsand allowing easier access.

Medicine may be deposited into the joints through either the superior or inferiorrecesses or by directly going to the joint itself. Of the superior and inferior recesses,typically it is the inferior that is much easier to enter. For this method of blockade, thepatient is best in the prone position with several pillows under the abdomen to inducemild flexion of the spine. When the attempt is to enter the joint itself, often it is necessaryto put the patient in a more oblique position.

Imaging

Studying any previously obtained axial images of the spine can prove useful. Byknowing the concavity of the joint in question, it is possible to then plan the entry intoeach joint. When visualizing a joint under the process of fluoroscopy, cephalocaudadangulation of the C-arm is accomplished to line up the beam with the end plates of thevertebral body. The C-arm next goes through the maximum obliquity to best visualizethe joint. The C-arm is then brought to a more AP view until the joint just disappears.Because of the concave nature of the surfaces of the joint, this most likely is the easiestentrance point (Fig. 8).

If access of the inferior recess of the joint is anticipated, then the patient should beprone on the fluoroscopy table (Fig. 9). The C-arm is adjusted so that the inferior borderof the lamina is best visualized. Tracing this line out laterally should then help theobserver appreciate the approximate location of the inferior articulating process.

With any method used, correct placement is confirmed only through the use of contrast.Extravasation of contrast can then be studied, which will lead to some degree of confi-dence in the selection of the intra-articular injection.

Needle Techniques

A smaller-bore needle is preferred. In most patients, a 25 gage needle should be suffi-cient. Additionally, with this smaller-bore needle, it is very easy to introduce the tip fullywithin the capsule of the joint. This smaller needle also has less chance of introducing anytrauma. Occasionally, in larger patients, a somewhat larger-bore needle may be necessary.

Techniques in Pain Management 289

As in the case with medial branch blocks, it is necessary to use minimal anesthetic toreach the target point. This minimizes the chance of a false positive response. If noanesthetic is used, increases the chance of a false negative block; the patient perceives

290 Reynolds and Kine

Fig. 8. Oblique view of interarticular arthrograms of right L4-L5 and L5-S1. The needletargets the midportion of the joint of L4-L5 and the inferior recess of the L5-S1 joint.

Fig. 9. AP view of arthrograms of L4-L5 and L5-S1 showing different approaches whentargeting midportion of joint vs inferior recess.

Techniques in Pain Management 291

postprocedural pain from the needling itself and cannot determine whether or not theregionalized pain has been removed.

Injectate

The injectate may consist of anesthetic, contrast, and glucocorticoid. Generally, verysmall volumes are necessary. A typical joint will hold approx 1.5 cc or less. A joint shouldnot be overpressurized because this could potentially rupture the capsule.

Other Considerations

Although not scientifically validated with double-blind controlled studies, numerousobservational studies and anecdotal experience suggest, in certain individuals, that glu-cocorticoid placed intra-articularly may reduce discomfort. The duration of this reliefmay be more likely on the order of weeks or months, rather than years. Although notproved with absolute scientific scrutiny, anecdotal evidence suggests that duration ofrelief on the order of days most likely is nondiagnostic and may be from systemicuptake of glucocorticoid or other factors. Durations of relief in excess of 2 wk, how-ever, are suggestive, although not conclusive, of the medicine having been placed in ornear the pain generator.

Sacrococcygeal Joint InjectionsBackground

The sacrococcygeal joint is a potential source of localized pain. Trauma to this jointmay occur during a fall and is occasionally seen after childbirth. There may be damageto the joint itself or there may be a contusion to the sacrococcygeal nerves supplying thejoint. Occasionally, discomfort in the area of the coccyx is a referred pain from a structurehigher up.

Positioning

The patient is prone on the fluoroscopy table. The area is meticulously prepped anddraped.

Imaging

Initial imaging should be in AP view. This will ensure that needle placement is in themidline, which will help prevent sliding alongside the joint into the deeper structures.

A lateral fluoroscopic image should be obtained to assess the plane of the joint(Fig. 10). Occasionally, more than one radiolucency is observed. There is sometimescomplete separation of the joint, and this should be ascertained prior to placement ofthe needle.

Needle Techniques

A short 25- or 27-gage needle is all that is typically necessary. After the joint architec-ture is assessed by biplanar fluoroscopy, the needle should be placed with AP visualization.The needle generally travels only a few millimeters to reach the joint. The C-arm can thenshow lateral imaging to assess whether the joint can indeed be entered. Often, collimationof the image intensifier will aid in proper visualization of the joint.

Injectate

Only an extremely small volume of injectate is necessary. Usually, this is <0.5 cc.The mixture should consist of anesthetic, contrast material, and glucocorticoid. Largervolumes will also anesthetize the sacrococcygeal nerve.

Other Considerations

Use of glucocorticoid in the sacrococcygeal joint is based on anecdotal evidenceand observational analysis. Initial relief may be on the order of weeks or months.Repeat injections have the potential to give more prolonged relief. Ultimately, surgicalcoccygectomy may be necessary in the most refractory and distressing cases.

REFERENCES

1. Moran R, O’Connell D, Walsh MG. The diagnostic value of facet joint injections. Spine1988;13(12):1407–1410.

2. Schwarzer AC, Wang SC, O’Driscoll D, Harrington T, Bogduk N, Laurent R. The ability ofcomputed tomography to identify a painful zygapophysial joint in patients with chroniclow back pain. Spine 1995;20(8):907–912.

3. Schwarzer AC, Aprill CN, Derby R, Fortin J, Kine G, Bogduk N. Clinical features of patientswith pain stemming from the lumbar zygapophysial joints: is the lumbar facet syndrome aclinical entity? Spine 1994;19(10):1132–1137.

4. Revel M, Poiraudeau S, Auleley GR, et al. Capacity of the clinical picture to characterizelow back pain relieved by facet joint anesthesia: proposed criteria to identify patients withpainful facet joints. Spine 1998;23(18):1972–1976; discussion, 1977.

5. Kaplan M, Dreyfuss P, Halbrook B, Bogduk N. The ability of lumbar medial branchblocks to anesthetize the zygapophysial joint: a physiologic challenge. Spine 1998;23(17):1847–1852.

6. Dreyfuss P, Schwarzer AC, Lau P, Bogduk N. Specificity of lumbar medial branch and L5dorsal ramus blocks: a computed tomography study. Spine 1997;22(8):895–902.

7. Schwarzer AC, Aprill CN, Derby R, Fortin J, Kine G, Bogduk N. The relative contribu-tions of the disc and zygapophyseal joint in chronic low back pain. Spine 1994;19(7):801-806.

292 Reynolds and Kine

Fig. 10. Lateral image of sacrococcygeal joint injection with contrast and anesthetic.

8. Butler D, Trafimow JH, Andersson GB, Mcneill TW, Huckman MS. Discs degeneratebefore facets. Spine 1990;15(2):111–113.

9. Marks RC, Houston T, Thulbourne T. Facet joint injection and facet nerve block: a randomisedcomparison in 86 patients with chronic low back pain. Pain 1992;49(3):325–328.

10. Carette S, Marcoux S, Truchon R, et al. A controlled trial of corticosteroid injections intofacet joints for chronic low back pain. N Engl J Med 1991;325(14):1002–1007.

11. Schwarzer AC, Derby R, Aprill CN, Fortin J, Kine G, Bogduk N. The value of the provocationresponse in lumbar zygapophyseal joint injections. Clin J Pain 1994;10(4):309–313.

12. Niemisto L, Kalso E, Malmivaara A, Seitsalo S, Hurri H. Radiofrequency denervation forneck and back pain: a systematic review within the framework of the cochrane collaborationback review group. Spine 2003;28(16):1877–1888.

13. Slipman CW, Bhat AL, Gilchrist RV, Issac Z, Chou L, Lenrow DA. A critical review of theevidence for the use of zygapophysial injections and radiofrequency denervation in thetreatment of low back pain. Spine J 2003;3(4):310–316.

14. Leclaire R, Fortin L, Lambert R, Bergeron YM, Rossignol M. Radiofrequency facet jointdenervation in the treatment of low back pain: a placebo-controlled clinical trial to assessefficacy. Spine 2001;26(13):1411–1416; discussion, 1417.

15. van Kleef M, Barendse GA, Kessels A, Voets HM, Weber WE, de Lange S. Randomizedtrial of radiofrequency lumbar facet denervation for chronic low back pain. Spine 1999;24(18):1937–1942.

16. Dreyfuss P, Halbrook B, Pauza K, Joshi A, Mclarty J, Bogduk N. Efficacy and validityof radiofrequency neurotomy for chronic lumbar zygapophysial joint pain. Spine2000;25(10):1270–1277.

17. Schofferman J, Kine G. The effectiveness of repeated radiofrequency neurotomy for lumbarfacet pain. Spine 2004;29:2471–2473.

18. Stitz MY, Sommer HM. Accuracy of blind versus fluoroscopically guided caudal epiduralinjection. Spine 1999;24(13):1371–1376.

19. Renfrew DL, Moore TE, Kathol MH, el-Khoury GY, Lemke JH, Walker CW. Correct place-ment of epidural steroid injections: fluoroscopic guidance and contrast administration. Am JNeuroradiol 1991;12(5):1003–1007.

20. White AH, Derby R, Wynne G. Epidural injections for the diagnosis and treatment of low-back pain. Spine 1980;5(1):78–86.

21. Bogduk N, April C, Derby R. Epidural steroid injection, in Spine Care Diagnosis and Con-servative Treatment White A, Schofferman J, eds. (AH W, eds.), Mosby, St. Louis, 1995,pp. 322–343.

22. Derby R, Lee SH, Chen Y. Injection therapies, nerve ablation, and intradiscal electrothermalannuloplasty for degenerative lumbar conditions. Semin Spine Surg 2003;15(4): 393–410.

23. Buchner M, Zeifang F, Brocai DR, Schiltenwolf M. Epidural corticosteroid injection in theconservative management of sciatica. Clin Orthop 2000(375):149–156.

24. Carette S, Leclaire R, Marcoux S, et al. Epidural corticosteroid injections for sciatica dueto herniated nucleus pulposus. N Engl J Med 1997;336(23):1634–1640.

25. Karppinen J, Malmivaara A, Kurunlahti M, et al. Periradicular infiltration for sciatica: arandomized controlled trial. Spine 2001;26(9):1059–1067.

26. Vad VB, Bhat AL, Lutz GE, Cammisa F. Transforaminal epidural steroid injections in lum-bosacral radiculopathy: a prospective randomized study. Spine 2002;27(1):11–16.

27. Wang JC, Lin E, Brodke DS, Youssef JA. Epidural injections for the treatment of symptomaticlumbar herniated discs. J Spinal Disord Tech 2002;15(4):269–272.

28. Karppinen J, Ohinmaa A, Malmivaara A, et el. Cost effectiveness of periradicular infil-tration for sciatica: subgroup analysis of a randomized controlled trial. Spine 2001;26(23):2587–2595.

29. Riew KD, Yin Y, Gilula L, et al. The effect of nerve-root injections on the need for operativetreatment of lumbar radicular pain: a prospective, randomized, controlled, double-blindstudy. J Bone Joint Surg Am 2000;82-A(11):1589–1593.

Techniques in Pain Management 293

30. Dooley JF, Mcbroom RJ, Taguchi T, Macnab I. Nerve root infiltration in the diagnosis ofradicular pain. Spine 1988;13(1):79–83.

31. Derby R, Kine G, Saal JA, et al. Response to steroid and duration of radicular pain as pre-dictors of surgical outcome. Spine 1992;17(6 Suppl):S176–S183.

32. Chen Y, Derby R, Kim BJ, Lee SH. Epidural steroid injections: past, present and future.Spineline 2003;4:9–20.

33. Schwarzer AC, Aprill CN, Bogduk N. The Sacroiliac joint in chronic low back pain. Spine1995;20(1):31–37.

34. Slipman CW, Jackson HB, Lipetz JS, Chan KT, Lenrow D, Vresilovic EJ. Sacroiliac joint painreferral zones. Arch Phys Med Rehabil 2000;81(3):334–338.

35. Murata Y, Takahashi K, Yamagata M, Takahashi Y, Shimada Y, Moriya H. Origin and path-way of sensory nerve fibers to the ventral and dorsal sides of the sacroiliac joint in rats. J OrthopRes 2001;19(3):379–383.

36. Atlihan D, Tekdemir I, Ates Y, Elhan A. Anatomy of the anterior sacroiliac joint with referenceto lumbosacral nerves. Clin Orthop 2000(376):236–241.

37. Maigne JY, Aivaliklis A, Pfefer F. Results of sacroiliac joint double block and value of sacroil-iac pain provocation tests in 54 patients with low back pain. Spine 1996;21(16):1889–1892.

38. Dreyfuss P, Michaelsen M, Pauza K, Mclarty J, Bogduk N. The value of medical history andphysical examination in diagnosing sacroiliac joint pain. Spine 1996;21(22):2594–2602.

39. Slipman CW, Sterenfeld EB, Chou LH, Herzog R, Vresilovic E. The predictive value ofprovocative sacroiliac joint stress maneuvers in the diagnosis of sacroiliac joint syndrome.Arch Phys Med Rehabil 1998;79(3):288–292.

40. Elgafy H, Semaan HB, Ebraheim NA, Coombs RJ. Computed tomography findings inpatients with sacroiliac pain. Clin Orthop 2001(382):112–118.

41. Miller JA, Schultz AB, Andersson GB. Load-displacement behavior of sacroiliac joints.J Orthop Res 1987;5(1):92–101.

42. Sturesson B, Uden A, Vleeming A. A radiostereometric analysis of the movements of thesacroiliac joints in the reciprocal straddle position. Spine 2000;25(2):214–217.

43. Katz V, Schofferman J, Reynolds J. The sacroiliac joint: a potential cause of pain after lumbarfusion to the sacrum. J Spinal Disord Tech 2003;16(1):96–99.

44. Rosenberg JM, Quint TJ, De Rosayro AM. Computerized tomographic localization ofclinically-guided sacroiliac joint injections. Clin J Pain 2000;16(1):18–21.

45. Hanly JG, Mitchell M, Macmillan L, Mosher D, Sutton E. Efficacy of sacroiliac corticos-teroid injections in patients with inflammatory spondyloarthropathy: results of a 6 monthcontrolled study. J Rheumatol 2000;27(3):719–722.

46. Luukkainen RK, Wennerstrand PV, Kautiainen HH, Sanila MT, Asikainen EL. Efficacy ofperiarticular corticosteroid treatment of the sacroiliac joint in non-spondylarthropathicpatients with chronic low back pain in the region of the sacroiliac joint. Clin Exp Rheumatol2002;20(1):52–54.

47. Slipman CW, Lipetz JS, Plastaras CT, et al. Fluoroscopically guided therapeutic sacroiliacjoint injections for sacroiliac joint syndrome. Am J Phys Med Rehabil 2001;80(6):425–432.

48. Ferrante FM, King LF, Roche EA, et al. Radiofrequency sacroiliac joint denervation forsacroiliac syndrome. Reg Anesth Pain Med 2001;26(2):137–142.

49. Wray CC, Easom S, Hoskinson J. Coccydynia. Aetiology and treatment. J Bone Joint Surg Br1991;73(2):335–338.

50. Perkins R, Schofferman J, Reynolds J. Coccygectomy for severe refractory sacrococcygealjoint pain. J Spinal Disord Tech 2003;16(1):100–103.

294 Reynolds and Kine

15Experience With Minimally Invasive

Nucleus Replacement

Michele Marcolongo, PhD, Parviz Kambin, MD,Anthony Lowman, PhD, and Andrew Karduna, PhD

BACKGROUND

Scope of the Problem

More than 5 million Americans suffer from chronic low-back pain, making it the lead-ing cause of lost workdays in the United States and one of the most expensive healthcare issues today (1). Although the causes of low-back pain remain unclear, it is believedthat approx 75% of cases are associated with degenerative disc disease (1). It is generallybelieved that dehydration of the degenerated nucleus pulposus (NP) leads to a reductionin hydrostatic pressure on the internal surface of the annulus, resulting in an abnormalstress state in the tissue and, consequently, a breakdown of the annular tissue seenmacroscopically as fissures and tears. This manifests as chronic and debilitating painowing to tissue impingement on nerve roots and presents as a herniated or ruptured disc.Current treatment options, such as discectomy and fusion, are fairly successful in reduc-ing pain but do not restore normal biomechanical function to the disc. The likely out-come of these procedures is further degeneration of either the initially affected disc (fordiscectomy) or adjacent segments (for fusion). Degenerative disease of the spine is oneof if not the leading musculoskeletal disorder confronting our health system. The spineprovides the major structural element of the neck and trunk while protecting the spinalcord. Spinal degeneration is an irreversible process leading to loss of mechanicalintegrity with the potential for neurological compromise. Clinical manifestations ofdegenerative spine disease are variable and graded and are categorized into a variety ofdiseases. These diseases include mechanical cervical and lumbar pain such as internaldisc disruption, acute spinal instability, herniated NP, degenerative spondylolisthesis,degenerative scoliosis, and spinal stenosis.

Spine degeneration is ubiquitous. Cadaveric examination of lumbar disc showed a97% incidence of disc degeneration by age 50 (2). Approximately one-third of the pop-ulation has evidence of degenerative disease on radiographic study by age 40 (3,4).Sixty to 80% of all working adults lose time from work owing to back pain (5).Tremendous pain and suffering are associated with the degenerative diseases of thespine, in addition to the societal costs.

295

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

Treatment of Degenerative Disc Disease

Current treatment for degenerative disc disease focuses primarily on relieving backand leg pain. The inflammatory response secondary to tissue injury and nerve rootimpingement are frequently observed in association with degenerative disc changes andare thought to mediate a pain response. Although the exact mechanism of pain genera-tion is debated, >75% of low-back pain is associated with degenerative disc disease (1).Moreover, sciatica has an even stronger association with disc herniation and nerve rootimpingement. In the absence of neurological impairment, treatment begins with conser-vative care, activity modification, and anti-inflammatory medication. Under this regi-men, 85–90% of patients are treated successfully in 3 mo (10,11). However, theremaining 10–15% of patients incur >75% of the treatment costs, often requiring highlyinvasive surgical interventions. The most common surgical treatments—discectomyand spinal fusion—are performed to reduce pain, not to restore disc function. Discec-tomy is employed when the disc has herniated and is impinging on nerve roots, causingpain. In this surgery, the impinging region of the annulus fibrosus (AF) and NP isexcised, hence alleviating pressure on the nerves and eliminating pain. Pain is elimi-nated in 90–95% of cases (12). However, as previously noted, this approach does noth-ing to restore normal biomechanics of the vertebral segment (13): the NP is stilldehydrated and the annulus fibers are still likely operating in compression. Therefore,the patient continues along the path of disc degeneration over ensuing years.

Surgical fusion—inducing bone growth across the functional spinal unit to eliminatedisc loading and motion—is reserved for patients with chronic severely disabling pain.Generally, discs treated with fusion are farther along the path of degeneration. Withoutdelineating the specific indications for fusion of the functional spinal unit (which arevaried depending on the signs and symptoms of the degenerative disease), suffice it tosay that approx 150,000 spinal fusions are performed per year in the United Statesalone. The numbers are growing exponentially. Regardless of the extent to which thisprocedure is performed, the results of spinal fusion vary extensively (14). Failure offusion can be common in 10–30% of patients depending on disease. More perplexing isthe clinical outcome, which may not improve with increased rates of fusion. Significantlong-term limitations are associated with a spinal fusion. Spinal fusion does nothing torestore the normal biomechanics of the vertebral segment. In fact, the lack of motionwithin the segment can lead to further degeneration of the adjacent intervertebral discs(15). Lehmann et al. (16) pursued a long-term follow-up of lumbar fusions in patientsfrom 21 to 52 yr of age and found that 44% of the patients were currently still experi-encing low-back pain, 50% had back pain within the previous year, 53% were on medi-cations, 5% had late sequelae secondary to surgery, and 15% had repeat lumbar surgery.This suggests the need for alternatives to fusion.

Intervertebral Disc Replacement

An alternate approach to the treatment of degenerative disc disease is to remove thediseased disc in its entirety and replace it with a synthetic implant (17–33). Discreplacement may serve to eliminate pain while restoring physiological motion. Thisapproach for total knee and hip replacement has been highly successful. However, discreplacement has not enjoyed the same clinical success. More than 56 reports onmechanical replacement have been described (28), most from the standpoint of design

296 Marcolongo et al.

concept, with very few clinical reports. In general, these solutions provide dynamicload bearing with passive restraint but mechanically fail in long-term application. Oneof the most promising designs is the LINK disc replacement developed by Buttner-Janzet al. (30,31) and Zippel (32), which has had clinically success. This sandwich designconsists of cobalt chromium alloy end plates and a polyethylene core. The results ofclinical trial in 93 patients showed back pain relief in 20% of patients and leg pain reliefin 40–50% of patients after an average implantation time of 1 yr (33). This device iscurrently an investigational device in the United States.

NP Replacement

Rather than replacing the entire disc, several investigators have attempted to replacethe NP alone. This is intended to provide a surgical technique that would offer a less inva-sive approach to pain relief while potentially restoring the functional biomechanics to thesystem. This approach could be most effective in patients with early diagnosis of disc dis-ease, before the annulus has suffered significant degeneration. The concept of nucleusreplacement was developed by Nachemson in the early 1960s, who attempted injection ofa self-curing silicone into the disc space in cadavers (34). Further research into siliconereplacement of the nucleus continued into the early 1990s (35–40). The silicone prosthe-ses have been promising as far as mechanical properties and ease of insertion into thenucleus; however, silicone synovitis and its associated complications may play a signifi-cant role in limiting the clinical success of this material, as it has in other orthopedic joints(41). Gan et al. (42) investigated nucleus tissue engineering as a way of regenerating thedegenerated tissue. Although the cells clearly adhered to the glass substrate and primarilyheld their phenotype after 3 wk in vitro, it was not clear that the matrix was that of ahealthy NP. This approach is reasonable in an era of tissue engineering solutions, but celland molecular biologists are still struggling to determine the nature of the NP cell. There-fore, setting and meeting the requirements of regenerating the tissue, although promising,has many challenges to overcome before adaptation as a clinical treatment.

Ray and colleagues have developed and internationally commercialized a polyacryloni-trile hydrogel nucleus replacement covered with a polyethylene fiber jacket. The device,generally implanted in pairs, is intended to improve disc height, restore motion, and relievepain owing to disc herniation (43). The current surgical success rate for patients implantedfrom 1999 through 2001 is 88%, with the primary failure being dislocation of the implantfrom the nucleus. The open procedure through the annulus can allow the hydrogels (whichare implanted in the hydrated state) to exit through the incision site. Although the surgicaltechnique is still evolving to the level of having a truly satisfactory procedure, patient painin this short-term follow-up study was reduced 86% (Oswestry and visual analog scalepain levels) and spinal flexibility increased 67%. This clinical trial has provided the firstevidence of improvement in the treatment of degenerative disc disease via nucleus replace-ment. This nucleus replacement device is currently completing a phase I clinical investiga-tion in the United States and is being sold commercially in Europe.

Bao and Higham (44,45) have also approached nucleus replacement with a hydrogelpolymer. This material selection has resulted in an implant that has similar mechanicalproperties to those of the nucleus as well as similar physiological properties, maintain-ing about 70% water content under physiological loading conditions. The particularhydrogel employed by these researchess comprises semicrystalline polyvinyl alcohol

Minimally Invasive Nucleus Replacement 297

(PVA) (44,45). PVA is a biocompatible polymer (46) that has the ability to absorb wateror physiological fluid and survive mechanical loading as would exist in the nucleusregion of the intervertebral disc. However, PVA is not entirely stable within the physio-logical environment of the body, showing degradation through the melting out of smallcrystallites over time, which can result in a reduction of mechanical properties andleaching of molecules into the physiological environment (47).

Intervertebral Disc Mechanics

Earlier work was performed for nucleus replacement with a synthetic material incadaveric FSUs (48–51) and in animals (52,53). The data reported were primarily forend-plate strains (48) and segmental mobility in combined loading modes (49–51).Parameters such as rotational displacement, disc height, and range of motion wereobserved before and after nucleus implantation to assess the restoration ability of thenucleus implant device used. However, in all of the cases, nucleotomy was facilitatedby making a small incision through the AF. This may not be the ideal approach forassessing the effect of nucleus implant because the surrounding annulus was damagedat least partially. No human cadaver studies have reported the effect of nucleus implantreplacement on the pure compressive behavior of the FSU. However, Meakin et al. (52)used sheep discs to assess the effect of nucleus implant on bulging direction of theannulus fibers in pure compression. They observed that a nucleus implant with a modu-lus in the range of 0.2–40 MPa prevented the inward bulging of the annulus, seen in thecase of a denucleated specimen. However, their numerical modeling (52) showed thatthe stresses were restored to those of the intact FSU only with an implant in the modu-lus range of 3–5 MPa. The idea of nucleus replacement by a synthetic material wasproven feasible in all of the studies just descnbed. Mechanisms of interactions betweenthe nucleus implant and surrounding tissue have not been thoroughly explored. In addi-tion, numerical modeling of the human lumbar FSU with a nucleus replacement has notbeen reported in the peer-reviewed literature.

Stable, Solid Hydrogel Polymer NP Implant

Previous work in our laboratory focused on developing and characterizing a highlychemically stable hydrogel polymer system (54–56). The motivation for materials selec-tion for a device to replace the nucleus is fourfold: (1) the individual polymers are biocom-patible; (2) the polymers interact to form physical crosslinks that serve to stabilize thehydrogel material; (3) the material can be processed in a variety of ways that allow the tai-loring of mechanical properties without modification of chemistry; and (4) the material dis-plays shape memory properties with hydration level, which may facilitate minimallyinvasive implantation of the device. This polymer is based on PVA but incorporates a sec-ond polymer, polyvinylpyrrolidone (PVP), that serves to stabilize the hydrogel throughinterpolymer complexes, which serve as secondary, physical crosslinks and provide thenetworks with additional stability, as demonstrated in vitro (55,56).

One major concern about the use of hydrogels is the potential leaching of unreactedmonomers or crosslinking agents from the insoluble, chemically crosslinked structures.A significant benefit of the proposed hydrogels is that the structure of the proposedimplants relies on physical crosslinking, rather than covalent chemical crosslinking, tohold polymer chains together. The gels are prepared by blending the two polymers.

298 Marcolongo et al.

The polymers proposed for use are soluble and can thus be purified easily prior toblending. Because PVA (46) and PVP (57) have been shown to be compatible in thebody, the blend of the two is also likely to be compatible.

Aside from chemical considerations in hydrogel stablility in vivo, there are mechanicalconsiderations for this structural application: (1) the hydrogel must be able to withstandthe repetitive loading environment of movement; (2) the hydrogel must have a viscoelas-tic behavior that matches fairly closely that of the AF so that there is no excessive defor-mation of the device owing to creep; and (3) the hydrogel must display a modulus andPoisson’s ratio that, on loading, will provide an interfacial stress with the annulus thatmimics the normal intradiscal pressure to restore mechanical function to the implantedvertebral segment. Our preliminary data have demonstrated that the family of PVA/PVPhydrogels is promising in meeting these material requirements. However, if furthermechanical analysis shows a limitation to the mechanical behavior of the gels, the mechan-ical behavior of the material may be altered through processing changes, or through varia-tion of the polymer composition and/or polymer concentration, making this family ofmaterials very desirable when the need arises to tailor mechanical properties.

DESIGN AND CHARACTERIZATION OF HYDROGEL

Characterization and Surgical Feasibility of PVA/PVP Hydrogel Material

Previous research in our laboratories led to the development and characterization ofnovel hydrogels prepared from blends of PVA and PVP (55,56). As already stated, themotivation for materials selection for a device to replace the nucleus is fourfold. The goalof this prior work was to establish the in vitro stability of the gels and to determine thesurface chemical changes and mechanical behavior of the gels over time of immersionin vitro. These properties were examined as a function of the composition of PVA/PVPas well as the molecular weight of PVA and PVP.

In Vitro Stability: Swelling and Dissolution Behavior

Blends containing between 0.5 and 75 wt% PVP and the balance PVA were preparedwith four different molecular weight combinations. Solutions of PVA and PVP wereprepared by dissolving various ratios of the polymers in deionized water at 90ºCovernight. The solutions, which contained 10% by weight polymer, were hom*ogenizedfor 30 min using sonication. The solutions were then cast into Plexiglas® trays anddried at 37ºC for 72 h. On drying, residual water was removed from the polymer filmsin a vacuum oven at 35°C with an absolute pressure of 127 mmHg. The polymer blendswere swollen in deionized water for 1 to 2 h to form gels. Circular discs were punchedfrom the films and dried in an oven to evaporate deionized water introduced inswelling. After the dry mass of the circular discs (n = 3) was measured, the discs wereswollen at 37°C in deionized water. The mass of each swollen gel was measured regularlyfor 120 d. Additionally, the deionized water was replaced frequently. Following 120 d ofswelling, the gels were removed from solution and dried under vacuum. The dryweights of the discs were recorded in air and heptane in order to determine the volumeand density. The weight swelling ratio (weight of swollen gel/weight of dry polymer)and volume swelling ratio (volume of swollen gel/volume of dry polymer) for the poly-mers were calculated. Additionally, the dissolution of the polymer was determined bycomparing the initial preswollen weight to the final dry weight.

Minimally Invasive Nucleus Replacement 299

The swelling and dissolution data are presented in Fig. 1 for the average molecularweight combination of 143,000 g/mol (PVA) and 10,000 g/mol (PVP), the most stablemolecular weight combination tested. For all of the samples, the gels rapidly hydrated tomaximum value, followed by some decrease in gel mass. The initial reduction wasowing to the gels relaxing during swelling followed by some dissolution of the polymers.However, for polymers containing <25% PVP, the swelling behavior reached equilib-rium within 2 d. Increasing amounts of PVP, however, led to a material that was not

300 Marcolongo et al.

Fig. 1. (A) Mass swelling ratio over time of immersion for gels prepared with (A) 143KPVA/10K PVP molecular weight combination with (●) 0% PVP, ( ) 0.5% PVP, (■) 0.75%PVP, ( ) 1% PVP, (▲) 5% PVP, (Δ) 10% PVP, (◆) 25% PVP, and ( ) 50% PVP (+) 75% PVPand (B) 50% PVA/50% PVP hydrogel prepared with (●) 143K PVA/10K PVP ( ) 143KPVA/40K PVP (■) 95K PVA/10K PVP, and ( ) 95K PVA/40K PVP. (B) Polymer mass loss isminimized from that of PVA alone with the addition of 0.5–5% PVP.

stable, as indicated by the declining q values over time of immersion in physiologicalsolution. The total polymer mass loss after 120 d resulted in 7% loss for PVA alone, vs3% for PVA with up to 1% PVP. Increasing amounts of PVP in the gel led to increasedpolymer mass loss, indicating reduced stability of the material in vitro.

The dissolution behavior of the gels was examined using swelling studies and atten-uated total reflectance-Fourier transform infrared spectroscopy (ATR-FTIR) (NicoletMagna-IR 560 Spectrometer; Madison, WI). For the studies, thin, flat strips of the driedpolymers were mounted on a zinc-selenium ATR crystal and the IR spectra wereobtained using 2048 scans and a resolution of 2 cm−1. These samples were then weighedin air and heptane and placed in phosphate-buffered saline solutions at 37ºC. At set timesfollowing immersion (1, 2, 4, 8, and 12 h), the strips were removed from solution andweighed. These strips were then dried under vacuum between flat plates for 72 h and theweights were recorded. The percent weight loss for the gels was calculated.

As seen in Table 1, short-term dissolution studies showed that for the gels a majorityof the polymer dissolving out of the gel was lost during the first 24 h. Based on thenature of the blends, we hypothesized that the early time mass loss was owing to disso-lution of the nonhydrogen bonded, amorphous PVP from the semicrystalline PVA. Todemonstrate This hypothesis, ATR-FTIR spectra were obtained for the gels prior toswelling and postswelling in the same time period as the weight loss analysis. The char-acteristic peaks that we chose to represent the two polymers were the secondary alcohol–C-C-O stretch peak at 1075 cm–1 for PVA (58), and a mixed mode of carbonyl groupstretch and –N-C stretch at 1654 cm−1 for PVP (59). The IR spectra of the dried blendfilms prior to any swelling are shown in Fig. 2. The changes in the relative peak heightwith changing composition are clear.

The IR spectra of dried films postswelling were obtained for each set time. A repre-sentative sample is shown in Fig. 2B. For each blend film, the carbonyl peak of PVPlost the initial height after swelling. The height of the 1075-cm−1 peak of PVA remainednearly constant for the entire period, indicating that most of the PVA remained in thehydrogel networks after swelling. On the other hand, for the 1654-cm−1 peak of PVP,the peak height decreased dramatically at the end of the first hour of swelling butremained nearly constant for the rest of the study. These changes are in agreement withthe dissolution experiments and support the hypothesis that PVP initially dissolves out thegel, followed by increased gel stability.

Finally, tensile tests of thin strips of swollen hydrogels were performed after 2 d and1, 2, 4, and 8 wk of immersion in vitro (n = 5). The dimensions of the strips (4.25 mmwide and 75 mm long) were in accordance with ASTM Standard Test Method for Tensile

Minimally Invasive Nucleus Replacement 301

Table 1Weight Loss (%) of PVA/PVP Hydrogels From Dissolution Analysis (n = 3)

Weight loss (%)

100/0 95/5 90/10 75/25 50/50Time (h) PVA/PVP PVA/PVP PVA/PVP PVA/PVP PVA/PVP

1 0.67 ± 0.58 0.77 ± 0.56 4.28 ± 3.80 22.93 ± 1.82 45.46 ± 3.658 0.77 ± 0.33 1.91 ± 0.16 7.02 ± 2.92 23.59 ± 3.11 48.39 ± 2.8924 1.57 ± 1.14 3.60 ± 0.15 7.92 ± 1.26 22.06 ± 2.62 47.00 ± 0.73

Properties of Thin Plastic Sheeting (D882-95A) (60). Tensile modulus measurementsindicated that the modulus was approx 2 MPa and that for the composition with 1%PVP, the modulus was stable over 56 d of immersion (55). A statistically significantreduction in modulus was measured for PVA alone. The reduction in modulus in PVA ismost likely owing to the significant mass loss, whereas the 1% PVP compositions hadno statistically significant reduction in modulus and correspondingly little mass losscompared to PVA alone. The structures of the hydrogels were evaluated and character-ized using rubber elasticity theory based on swelling studies and tensile experiments(55,61). The values of Mc increased for the hydrogels with more PVP initially blendedin the system, indicating the presence of fewer physical crosslinks (Fig. 3).

Mechanical Behavior: Stress Relaxation and Fatigue

Stress relaxation behavior of the PVA/PVP hydrogel (10 w/w% polymer, 99% PVA,1% PVP) is important to understand because of the effect of constant deformation onthe ability of this viscoelastic material to hold or dissipate stress. Pilot experimentswere conducted in our laboratory in which the hydrogel was held in a compressive dis-placement of 15% strain in a hydrated environment at 25°C. The load was monitoredand the load relaxation was examined over time. Figure 4 shows that the materialreached half of its initial load after 4 h of loading and that an equilibrium load level was

302 Marcolongo et al.

Fig. 2 (A) ATR-FTIR spectra prior to swelling: (*) –C-C-O stretch peak of PVA; (**) car-bonyl group stretch peak of PVP; a, 100/0 (PVA/PVP); b, 95/5 (PVA/PVP); c, 90/10(PVA/PVP); d, 75/25 (PVA/PVP); e, 50/50 (PVA/PVP). (B) ATR-FTIR spectra postswelling forinitially 95/5 (PVA/PVP) hydrogels: (*) –C-C-O stretch peak of PVA; (**) carbonyl groupstretch peak of PVP; a, prior to swelling; b, 1 h of swelling; c, 4 h of swelling; d, 8 h of swelling;e, 12 h of swelling; f, 24 h of swelling.

approached after 60 h. This pilot study enabled the test methodology to be developed(although temperature control at 37ºC will be added) and indicates that the viscoelasticresponse is very typical of elastomeric polymers.

Compression–compression fatigue was used to determine the duration of the mate-rial over repeated cycling in vitro. Samples were cycled from 1 to15% strain to repre-sent physiological strains achieved while walking (62) for either 10,000, 100,000, or1,000,000 cycles in a simulated physiological solution (63) held at 37°C for the dura-tion of loading (n = 3). The test was conducted at 5 Hz. A customized fixture wasdesigned and fabricated that included grooved bearing plates so that the hydrogelwould remain in a stable position for the duration of the test. Earlier work showed thatthe hydrogel tended to slip out of the smooth bearing plates. The load was recordedover the cycling period. After testing, the material was permitted to recover in vitro for14 d, and then a single compression test was run to determine whether any permanentmechanical degradation had occurred to the material.

Results showed that the instantaneous compressive modulus of the hydrogel afterany of the cycles up to and including 10 million cycles of fatigue was not appreciably

Minimally Invasive Nucleus Replacement 303

Fig. 3. The molecular weight between crosslinks increased for hydrogels with more PVP ini-tially blended into the system, indicating the presence of fewer physical crosslinks.

Fig. 4. Stress relaxation behavior of PVA/PVP hydrogel.

different from that of the unfatigued samples. (Fig. 5B). In addition, there was no per-manent change in dimensions of the implant up to 1 million cycles. However, after 10million cycles in fatigue, a change in dimensions had occurred. The height was reducedby 22% and the diameter increased by 8% (Fig. 5C). An analysis of the load-VS-number-of-cycles curve (Fig. 5A) shows that the load-carrying capacity of the material wasreduced as the number of cycles increased. This may be attributed in part to the stressrelaxation behavior of the material, as described in Fig. 4. The 1 million-cycle test took54 h to run; this corresponds to the point of equilibrium observed in the preliminarystress relaxation experiment, which may explain the reduced load-carrying capability ofthe material held at constant strain amplitude over time of cycling.

Cadaveric Endoscopic Implantation of Dehydrated Hydrogel Implant

To demonstrate the ability to remove damaged NP material and to insert the dehy-drated hydrogel endoscopically, a human cadaveric model was incorporated. Using aC-arm X-ray machine and instrumentation developed by a colleague (P. Kambin, MD,

304 Marcolongo et al.

Fig. 5. (A) Load-carrying capability of PVA/PVP hydrogel over 1 million cycles of strain-controlled loading. (B) Compression-proof test after hydrogel was loaded for 10 million cyclesand subsequently allowed to recover in vitro. The dark gray bars represent the compressivemodulus data of control samples that did not undergo any mechanical loading, and the light graybars represent the data from samples tested after 10 million cycles of fatigue. (C) Photograph ofcontrol (left) and 10 million cycle-fatigued implants.

Drexel University), we introduced an oval cannula through the triangular working zoneof the annulus (12). The 5 × 8 mm oval cannula provided a space for inserting forceps,which were used to resect a region of the nucleus. Once the nucleus was cleared, thedehydrated hydrogel polymer implant was inserted through the cannula into the nuclearcavity (Fig. 6). A saline drip was used to rehydrate the implanted gel in situ. Figure 6shows the implant and cadaveric model as well as an X-ray of the positioning of the can-nula in the nucleus region of the intervertebral disc. This exercise demonstrated the abilityof the implant to be introduced endoscopically; however, details regarding the dehydra-tion/rehydration of the implant as well as further development of the surgical techniqueare required.

BIOMECHANICAL CONSIDERATIONS OF NUCLEUS REPLACEMENT: COMPRESSIVE STIFFNESS OF IMPLANTED HUMAN CADAVERIC FSUs

In a separate study, we examined the compressive mechanical behavior of the humancadaveric FSU after implantation with a hydrogel nucleus replacement (64). This workfocused on the ability of the gel to re-create the “normal” biomechanics of the intactvertebral segment. The gel was not implanted in a dehydrated state, and the insertionmethod was not intended to represent the case of clinical implantation, which wouldrequire significant modification. However, the aim of this work was to establish theimplant biomechanics with an intact annulus.

A 10% polymer mixture was prepared from a blend of 95% PVA/5% PVP asdescribed previously, and the mixture was processed using six freeze-thawing pro-cesses. Hydrogel implants were mechanically tested in unconfined compression at100% strain/min. Lumbar spines were harvested from eight cadavers (three males, fivefemale) with an average age 65 yr. FSUs (n = 15) from L1 to L5 levels were resectedand prepared by removing the facet joints and posterior elements. Two flat, parallel cutswere made in the bone to ensure alignment of the axial compression load, and the bonewas stored in a freezer until the day of testing. Specimens were thawed for at least 2 h atroom temperature prior to testing. Mechanical testing was performed on an Instronmechanical test machine (Model 1331). Specimens were potted in the test fixture with

Minimally Invasive Nucleus Replacement 305

Fig. 6. Cadaveric endoscopic implantation of dehydrated hydrogel implant using annularapproach.

commercially available bonding material, and all specimens were preconditioned for 50cycles at 3% strain (commonly used for spine preconditioning) (65–67). Specimenswere axially compressed with a loading rate of 1 mm/s up to 15% of total height ofintervertebral disc (IVD) at 0.5 Hz for 5 cycles for each condition. The data of the fifthloading cycle were taken for analysis.

The following series of axial compressive tests was completed on each specimen.The intact specimen was first tested (intact condition). Then, the upper vertebral bodywas drilled using, a 16-mm-diameter Cloward core drill bit. This hollow-core drill waspositioned over the NP on the upper cut surface of the vertebral body and drilled per-pendicular to this surface. The vertebral body was penetrated to the top of the IVD,and the cylindrical bone plug above the IVD was removed (Fig. 7). The bone plug wasreinserted and the specimen was tested with the bone plug placed in its original posi-tion on the FSU (BI condition). Then, the bone plug was again removed and the NPwas incised in line with the core drill and the central portion of the NP was removedusing standard surgical instruments. This procedure allowed the AF to remain intactand the FSU was again tested (DN-1 denucleated condition). Next, a hydrogel implantwith a 16-mm diameter and a height equal to the IVD height was implanted in the cavity(formed by excision of the NP). The cylindrical bone plug was again placed in its posi-tion as before, and the specimen was tested (implanted condition). Finally, the implantwas removed, and the specimen was tested again in “denucleated” condition (DN-2condition).

Figure 8B shows the typical nonlinear nature of the load-displacement curve for oneof the specimens, in axial compression. In all the specimens, curves for DN-1 conditionand for DN-2 condition superimposed, indicating return of the specimen to its originaldenucleated state after removal of the implant. Drilling into the vertebrae (BI) reducedthe stiffness compared to the intact condition and was significantly different (p < 0.05).A more dramatic reduction in the stiffness was observed for denucleated specimens(stiffness value of 48% of BI at 15% strain, p < 0.05). Insertion of the hydrogel implantrestored the stiffness of the FSU to a value of 89% of BI at 15% strain and was signifi-cantly different at higher deformations (p < 0.05).

Calculated stiffness values agreed well with those previously reported in the literature(68,69). Restoration of the stiffness to the denucleated FSU after implantation with a

306 Marcolongo et al.

Fig. 7. Surgical preparation for cadaveric implantation of hydrogel nucleus replacement: (A)removal of NP, (B) implanted FSU.

polymeric hydrogel is evident from Fig. 8. The general premise that the spinal biomechan-ics results from a synergistic effect between the implanted hydrogel and the surround-ing annulus has been shown through this experimentation. A linear summation of thestiffness of the denucleated FSU (776.0 N/mm at 15% strain) and that of only hydrogel(2.0 N/mm at 15% strain) does not equal the stiffness of the implanted FSU (1433.0 N/mmat 15% strain). We hypothesize that this nonlinear increase in stiffness after hydrogelimplantation is the result of a mechanical interaction between the annulus and theimplant, in which the Poisson effect of the implant results in tension developing in thefibers of the annulus as the disc is loaded in compression.

Axisymmetric Finite Element Modeling of Human FSU

To understand better the stress state of the annulus with respect to the normal, intacthuman lumbar intervertebral disc and the effect of incorporating nucleus replacementsof different mechanical properties, we incorporated finite element modeling (64). TheFSU was assumed to be an axisymmetric object with respect to the sagittal plane, with

Minimally Invasive Nucleus Replacement 307

Fig. 8. (A) Stiffness vs strain for FSU under different conditions (n = 15) and (B) typicalload deformation curve for FSU under different conditions.

simplified geometry. To simulate the experimental condition, posterior elements werenot modeled for this analysis because the anterior region provides a major load-bearingarea in axial compression (70). The FSU was modeled as a cylindrical object with threemain components—NP, AF, and vertebrae—with the cartilage end plates neglected forthis preliminary analysis (as demonstrated in our modeling).

The axisymmetric FE model (ABAQUS) contained 14,702 nodes and 10,775 ele-ments. NP and bone were modeled as elastic materials. The AF was modeled as ahyperelastic material. Material properties for the AF were assumed isotropic for thispreliminary study (a limitation of the analysis, in that the actual tissue properties areanisotropic), and bone was assumed to be entirely trabecular. Values were obtainedfrom the literature (71). Mooney-Rivlin strain energy potential of first order was used tocharacterize the properties of the AF. The material properties used are presented inTable 2. Four-node axisymmetric elements were used. The total number of variables inthe model was 32,241.

The FE Model was validated against the experimental results obtained in our labora-tory. The validation criterion was matching of the load-displacement curve of the intactFSU. Using the geometry described and the properties noted in Table 2, we adjusted theMooney-Rivlan constants for the AF until the solution converged on the load-displacementcurve best representing that of the experimentally determined curve for the intact FSU(Fig. 9).

In this study, we were interested in determining the mechanism responsible for therestoration of near-intact compressive stiffness values by the implanted cadaver FSUexperiment. We investigated radial displacements of the annulus in the region of thenucleus interface, and the calculations revealed that the radial displacements of the intactand implanted conditions were quite similar, showing tensile displacements in the annulusnear the nucleus interface, whereas those of the denucleated condition revealed markeddifferences in that compressive displacements were calculated in the interface betweenthe annulus and the nuclear cavity.

To test the effect of nucleus implant modulus on the stiffness of the FSU using thevalidated model, we examined a range of moduli for the nucleus implant ranging from0.01 to 100 MPa. The results, shown in Fig. 10, demonstrated that the implant modulusdid affect compressive stiffness of the FSU and that those moduli in the range of 0.01–1MPa resulted in the closest matching of the implanted to the intact condition. In addi-tion, calculated intradiscal stresses (at the interface between the implant and the annulus)also changed with implant material modulus.

This analysis does correlate with our experimental observations and shows that thereis a relationship among intradiscal pressure, FSU compressive stiffness, and modulus ofthe nucleus implant. The relationship between modulus and annulus radial displace-ments has also been observed in a cadaveric sheep experimental model (52).

308 Marcolongo et al.

Table 2Input Material Properties to Axisymmetric Finite Element Model

Young’s modulus (MPa) Poisson’s ratio (ν)

Bone 12,000 0.30AF Hyperelastic 0.45NP 1 0.49

CONCLUSION

We have developed and characterized a PVA/PVP hydrogel polymer as a candidatefor nucleus replacement of the intervertebral disc. The material was shown to be highlystable in vitro and demonstrated mechanical integrity over 10 million cycles of com-pression–compression fatigue under physiological strain levels. When used as anucleus implant, the PVA/PVP material was able to interact with the annulus in a syner-gistic fashion to restore the compressive stiffness of the vertebral segment to 89% ofthat of the BI condition. Likely, this is a result of the nucleus implant providing a stress

Minimally Invasive Nucleus Replacement 309

Fig. 9. (A) Validation of the FE Model (FEM) to the compressive load-displacement curveof the modeled intact condition to the experimental intact condition also resulted in closematching of the denucleated modeled and experimental conditions. (B) Effect of implantmaterial modulus on load-displacement curves of implanted FSU model. Implant modulibetween 0.01 and 1 MPa showed closest matching to the load-displacement curve of the intactexperimental FSU.

to the annulus that mimics the intradiscal pressure of the normal NP. Although our pre-liminary data support the premise that treatment of degenerative disc disease with ahydrogel nuclear implant may reproduce intact FSU biomechanics from the compres-sive loading perspective, important questions still need to be answered, including staticimplant biomechanics in order in order under different loading regimes as well as time-dependent implant biomechanics, in order to begin to address viscoelastic considera-tions. In addition, a critical design challenge to the success of such a device will be theability of the implant to be inserted in a minimally invasive fashion and to maintain itsposition within the nucleus cavity without expulsing through the defect created oninsertion or through any preexisting tears in the AF. Although the data thus far arepromising for this technique, on going studies are necessary in order to achieve clinicalsuccess in incorporating nucleus replacement.

REFERENCES

1. MedPro, Month, 1998;VIII(1).2. Miller JAA, Schmatz C, Schultz, AB. Lumbar disc degeneration: correlation with age, sex,

and spine level in 600 autopsy specimens. Spine 1988;13(2):173.3. Boden SD, Davis DO, Dina TS, Patronas NJ, Wiesel SW. Abnormal magnetic-resonance

scans of the lumbar spine in asymptomatic subjects. J Bone Joint Surg 1990;72A:403–408.4. Wiesel SA, Tsournas N, Feffer HL, Citrin CM, Patronas NJ. Incidence of positive CAT

scans in an asymptomatic group of patients. Spine 1984;9:549–551.5. Deyo RA, Tsui-Wu YJ. Descriptive epidemiology of low back pain and its related medical

care in the United States. Spine 1987;12:264–268.6. Frymoyer JW. Back pain and sciatica. N Eng J Med 1988;318:291–300.7. Snook SH, Jensen RC. Cost, in Occupational Low Back Pain, (Pope MH, Frymoyer JW,

Andersson C et al.), Praeger, New York 1984, pp. 115–121.8. Biering-Sorensen F. Physical measurements as risk indicators for low back trouble over a

one-year period. Spine 1984;9:106–119.9. Holbrook TL, Grazier KL, Kelsey JL, Stauffer RN (eds.). The Frequency of Occurence,

Impact, and Cost of Musculoskeletal Conditions in the United States American Academy ofOrthopaedic Surgeons, Chicago, 1984.

10. Bush K, Cowan N, Katz DE, Gishen P. The natural history of sciatica associated with discpathology. Spine 1992;17:1205–1212.

310 Marcolongo et al.

Fig. 10. Finite element predictions for radial displacements in (A) intact, (B) denucleated, and(C) PVA/PVP-implanted human lumbar FSU. The radial displacements of the intact and implantedconditions were quite similar, showing tensile displacements in the annulus near the nucleus inter-face, whereas those of the denucleated condition showed marked differences in that compressivedisplacements were calculated in the interface between the annulus and the nuclear cavity.

11. Ahn S-H. Comparison of clinical outcomes and natural morphologic changes betweensequestered and large central extruded disc herniations. Yonsei Med 2002;43(3):283–290.

12. Kambin P, Savitz MH. Arthroscopic microdicectomy: an alternative to open disc surgery.M Sinai J Med 2000;67(4):283–287.

13. Weber H. Lumbar disc herniation: a controlled prospective study with ten years of observa-tion. Spine 1993;8:131–140.

14. Lee C, Langrana NA, Parsons JR, Zimmerman MC Development of a prosthetic interverte-bral disc. Spine 1991;16(6 Suppl): S253-S255.

15. Leong JC, Chun SY, Grange WJ, Fang D. Long-term results of lumbar intervertebral discprolapse. Spine 1983:793–799.

16. Lehmann TR, Spratt KF, Tozzi JE, et al. Long-term follow-up of lower lumbar fusionpatients. Spine 1987; 12:97–104.

17. Salib RM, Pettine KA. Intervertebral disk arthroplasty. 1993: US Patent 5258031.18. Patil A. Artificial intervertebral disc. 1982: US Patent 4309777.19. Baumgartner W. Intervertebral disk prosthesis. 1994: US patent 5320644.20. Hedman TP, Koastiuk JP, Fernie GR, Hellier WG Artificial spinal disc. 1988: US Patent

529548.21. Monson GL. Synthetic intervertebral disc prosthesis. 1969: US.22. Fischer H. Support chamber for holding two adjacent vertebrae. 1987: West Germany

patent.23. Edeland HG. Some additional suggestions for an intervertebral disc prosthesis. J Biomed

Mater Res: Appl Biomater 1989;23:189–194.24. Lee CK, et al. Functional and biocompatible intervertebral disc spacer. 1990: US Patent.25. Parsons JR, et al. Functional and biocompatible intervertebral disc spacer containing

elastomeric material of varying hardness. 1992: US Patent.26. Steffon AD. Artificial disc. 1991: US Patent.27. Enker P, Steffee A, McMillin C, Keppler L, Biscup R, Miller S. Artificial disc replacement.

Spine 1993;18:1067–1070.28. McMillin CR, Steffee AD. Artificial spinal disc with up to five years follow-up, in Transac-

tions of the 20th Annual Society for Biomaterials, 1994, p. 89.29. Steffee AD. The Steffee artificial disc, in Clinical Efficacy and Outcome in the Diagnosis

and Treatment of Low Back Pain (Weinstein JN, ed.) Raven, New York: 1992, pp. 245-257.30. Buttner-Janz K, et al. Intervertebral disc endoprosthesis. 1988: US Patent.31. Buttner-Janz K, Schellmack R, Zippel H. Biomechanics of the SB Charite lumbar interver-

tebral disc endoprosthesis. Int Orthop 1989;13:173–176.32. Zipple H. Charite modular: concept, experience, and results, in The Artificial Disc, (Brock

M, Myer HM, Weigel K, ed.) Springer, Berlin: 1991; pp. 69–77.33. Griffith SL, Shelukov AP, Buttner-Janz K. A multicenter retrospective study of the clinical

results of the LINK SB Charite intervertebral prosthesis. Spine 1994;19:842–1849.34. Nachemson A. Some mechanical properties of the lumbar intervertebral disc. Bull Hosp

Joint Dis 1962;23:130–132.35. Schneider PG, Oven R. Intervertebral disc replacement: experimental studies, clinical con-

sequences. Z Othop 1974;112:791–792.36. Schneider PG, Oven R. Plastic surgery on intervertebral disc. Part 1: Intervertebral disc

replacement in the lumbar region with silicone rubber: theoretical and experimental stud-ies. Z Othop 1974;112:1078–1086.

37. Camille RG, Sailani R. Livaste, Experimental study of lumar disc replacement. Rev ChirOthop Appar Mat 1978; 64(Suppl III):106–107.

38. Fassie B, Ginestle JF. Disc prosthesis made of silicone: experimental study and first clini-cal cases. Nouvelle Presse Med, 1978;21:207.

39. Bost M. Mechanical loading of the vertebral body cover plate: measurement of the directstress distribution at the interface between intervertebral disc and vertebral body, in DieWirbelsuule in Forschung, Praxis, Junulanns H, ed. Hippokrates, Stuttgart, 1982.

Minimally Invasive Nucleus Replacement 311

40. Ashida H, Yolurnvaniva K, Okumulu A. An attempt to develop artificial nuclei pulposi inlumbar intervertebral discs. I. Jpn Orthop Assoc 1990;64:S947.

41. Chan M, Chowchuen P, Workman T, Eilenberg S, Schweitzer M, Resnick D. Silicone syn-ovitis: MR imaging in five patients. Skeletal Radiol 1998;27(1):13–17.

42. Gan JC, duch*eyne P, Vresilovic E, Shapiro IM. Bioactive glass serves as a substrates asa substrate for maintenance of phenotype of nucleus pulposus cells of the intervertebraldisc. J Biomed Mater Res 2000;51(4):596–604.

43. Bertagnoli R, Schonmayr R. Surgical and clinical results with the PDN prosthetic disc-nucleus device. Eur Spine J 2002;11(Suppl 2):S143–S148.

44. Bao QB. Higham PA. Hydrogel intervertebral disc nucleus. 1991: US Patent.45. Bao QB, Higham PA. Hydrogel intervertebral disc nucleus. 1993: US. Patent46. Hara Y, Matsuura T, Taketani F. Biocompatibility of polyvinylalcohol gel as a vitreous sub-

stitute. Nippon Ganka Gakkai Zasshi 1998;102(4):247–255.47. Peppas NA, Merrill EW. Development of semicrystalline PVA networks for biomedical

applications. J Biomed Mater Res. 1977;11:423–434.48. Frei H, Oxland T, Rathonyi G. The effect of nucleus prosthetic device on a lumbar spine

mechanics, in Transactions of the 47th Annual Meeting of the Orthopaedic Research Soci-ety, 2001.

49. Dooris A, Hudgins G, Goel VK. Biomechanics of a multisegmental lumbar spine with aprosthetic nucleus. Spine 2002;2:3S–44S.

50. Bain AC, Sherman T, Norton BK, Hutton WC. A comparison of the viscoelastic behaviorof the lumbar intervertebral disc before and after the implantation of a prosthetic discnucleus. BED 2000;48:203–204.

51. Eysel P, Rompe JD, Schoemayr R. Biomechanical behavior of a prosthetic lumbar nucleus.Acta Neurochir (Wien) 1999;141:1083–1087.

52. Meakin JR, Reid JE, Hukins DW. Replacing the nucleus pulposus of the intervertebraldisc. Clin Biomech 2001;16:560–565.

53. Ledet E, Dirisio D, Tymeson M. The Raymedica PDN prosthetic disc nucleus device in thebaboon lumbar spine. Spine 2002;2:47S–128S.

54. Thomas J, Gomes K, Lowman A, Marcolongo M. Novel associated hydrogels for nucleuspulposus replacement. J Appl Biomater, in press.

55. Thomas J, Lowman A, Marcolongo M. Novel associated hydrogels for nucleus pulposusreplacement. J Biomed Mater Res 2003;67A:1329–1337.

56. Liu X, Marcolongo M, Lowman A. Short term in vitro response of associating hydrogels.Polymers, submitted.

57. Risbud MV, Bhat SV. Properties of polyvinyl pyrrolidone / beta-chitosan hydrogel mem-branes and their biocompatibility evaluation by haemorheological method. J Mater Sci:Mater Med 2001;12:75–79.

58. Lin-Vien D, Colthup NB, Fateley WG, Grasselli J. Alcohols and phenols, in The Handbookof Infrared and Raman Characteristic Frequencies of Organic Molecules, Academic, SanDiego: 1991.

59. Hu Y, et al. In Macromolecular Chemistry and Physics, 200. pp. 705–714.60. American society for Testing and Materials. Standard Test Method for Tensile Properties of

Thin Plastic Sheeting (D882-95a), ASTM in Annual Book of ASTM Standards, New York,1995: pp. 182–187.

61. Melrose J, et al. R D A longitudinal study of the matrix changes induced in the interverte-bral disc by surgical damage to the annulus fibrosis. J Orthop Res: Off Publ Orthop ResSoc, 1992. 10(5): p. 665–676.

62. Bagga CS, Williams P, Higham PA, Bao QB. Development of fatigue test model for aspinal nucleus prosthesis with preliminary results for a hydrogel spinal prosthetic nucleus.BED 1997;35: p. 441–442.

312 Marcolongo et al.

63. Marcolongo M, duch*eyne P, Lacourse W. Surface reaction layer formation in vitro on abioactive glass fiber/polymeric composite. J Biomed Mater Res, 1997;37: 440–448.

64. Joshi A, Karduna A, Marcolongo M. A non-linear finite element model of the lumbar func-tional spinal unit to determine the ideal material properties of a hydrogel nucleus implant.in 29th Annual Proceedings of the Society for Biomaterials, 2003; p. 672.

65. Glazer PA, et al. In vivo evaluation of calcium sulfate as a graft substitute for lumbar spinalfusion. Spine 2001;1:395–401.

66. Goel VK, Weinstein JN. Biomechanics of intact ligamentous spine, in Biomechanics of theSpine: Clinical and Surgical Perspective, CRC Press, Boca Raton FL: 1990; p. 116.

67. Suh JK, et al. Viscoelastic behavior of human cervical FSU in torsion, in NASS 11thAnnual Meeting, 1996.

68. Langrana NA, Edwards T, Sharma M. Biomechanical Analyses of Loads on the LumbarSpine. In The Lumbar Spine (Wiesel SW, Weinstein J, Herkowitz H, Dvorak J, Bell G, eds)W. B. Saunders, Philadelphia: 1996; pp. 163–171.

69. White AA, Panjabi MM. Clinical Biomechanics of the Spine, 2nd ed. Lippincott, Philadel-phia, 1990.

70. Frankel V, Nordin M. Basic Biomechanics of the Skeletal System. Lea & Febinger,Philadelphia: 1980.

71. Goel A. combined finite element and optimization investigation of lumbar spine mechanicswith and without muscles. Spine 1993;18(11):1531–1541.

Minimally Invasive Nucleus Replacement 313

16Interspinous Process Implant for Treatment

of Lateral and Central Spinal StenosisOperative Technique and Results

Douglas Wardlaw, ChM, FRCSEd

INTRODUCTION

The clinical syndrome of spinal stenosis is usually seen in middle-aged or elderlyindividuals who complain of leg pain, weakness, paresthesia, or sensory changes thatgenerally develop on standing or walking and are relieved, typically, by sitting andlying down. Often patients stoop forward to ease the symptoms when standing orwalking. The condition can be extremely disabling, making the sufferer, in many cases,housebound.

The earliest description of spinal stenosis in the English literature was in 1899 bySachs and Frankl, who described patients with lumbar or lower-extremity pain whowalked bent forward and whose symptoms were relieved by laminectomy (1). In 1927,there was a similar description by Putti (2), and he and Hirsch, in 1948 (3), describedforaminal compression. Both recommended facetectomy as treatment. Harris and MacNab(4) described the basic pathology of disc degeneration and its consequent effects onother spinal structures and pathological anatomy. MacNab (5) described lateral recessstenosis, foraminal encroachment by the posterior articular process, and foraminal andlateral disc herniation. Verbiest from Holland is generally accredited with recognizingthe clinical syndrome and relating it to spinal stenosis. The story typically told is thathe noted that patients had difficulty walking, but less or no difficulty cycling, whichcould be accounted for by the fact that the patients’ lumbar spines are in the flexed posi-tion while cycling but in the erect position while walking. His treatment was by widelaminectomy (6–9).

PATHOLOGICAL ANATOMY

The syndrome of adult spinal stenosis is almost certainly brought about by a combi-nation of factors. Verbiest, in his early work, demonstrated in his patients with spinalstenosis that the interpedicular distance of the lumbar spine vertebrae was close to theexpected normal range, but that the anteroposterior (AP) distance was significantly less

315

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

than expected (6). He suggested that the narrowing was owing to encroachment on thespinal canal by the articular processes.

The size and shape of the spinal canal vary enormously from a relatively openrounded triangular configuration to a trefoil shape (10). Eisenstein (11) studied the mor-phometry and pathological anatomy of the lumbar spine in South African Negro andCaucasoid skeletons with specific reference to spinal stenosis, He measured the inter-pedicular and AP diameter and found the canals narrowest from L2 to L4, and the trefoilconfiguration was usually at L5, making L4/L5 potentially the narrowest level. Bonydegenerative changes increased in the lower level and tended to be in a situation likelyto cause foraminal impingement of a nerve root (11).

Porter used ultrasound to measure the oblique diameter of the spinal canal in youngmen and women to assess the degree and extent of bony stenosis and showed that therewas a wide range representing a Gaussian curve (12). Kadziolka et al. (13) confirmedhis findings in a cadaver study. He subsequently went on to show that patients who sub-sequently developed symptomatic disc lesions all had canals within the narrow range,strongly suggesting that available space in the spinal canal is significant in the symp-tomatology of cauda equina and nerve root entrapment syndromes and in the responseto treatment (14). He and Eisenstein surveyed the lumbar spines of skeletons and showedthat the interpedicular distance at L5 is the narrowest point in the lumbar spine canal.Poor nutrition in early childhood may be a factor (15). Papp et al. (16) demonstratedthat the trefoil configuration occurs in about 25% of the population but that shape is notgenerally apparent until adulthood and that the midsagittal diameter of trefoil canals wassignificantly smaller than in the unaffected canals. The AP diameter is affected by thelength of the pedicles (17–19). Postacchini showed that loss of disc height leads toreduction in foraminal height but not AP distance, which is affected by pedicle length(20,21). In addition, degenerative spondylolisthesis is most common at the L4/L5 level.It is not surprising, therefore, that L4/L5 is the level most commonly affected in spinaland foraminal stenosis (22–25).

The main risk factors, therefore, for the development of symptoms of spinal stenosisare undoubtedly hereditary and developmental. The subsequent development of symp-toms will ultimately depend on the degree of herniation or degenerative change thatdevelops in the discs and segments of the lumbar spine throughout life.

As degenerative changes develop, the general structure of the nucleus of the discbreaks down with narrowing of the disc space and bulging outward of the annulus. At thesame time, the root canals narrow because of overriding of the facet joints (4,5,17,26).This in itself narrows the spinal canal, but also the subsequent bulging and hypertrophyof the facet joint capsule and infolding of the ligamentum flavum narrow the spinal canaland nerve root exit foramen further (26–29). Degenerative spinal stenosis narrows thespinal and root canals even further because as the cephalad vertebra moves forward, itslamina and the inferior facet move closer to the posterosuperior edge of the vertebralbody below, continuing to narrow the spinal and nerve root exit canal. Degenerative sco-liosis occurs when there is asymmetrical collapse of a disc in the process of discdegeneration and can affect several segments in the lumbar spine. Postacchini suggestedclassification of degenerative spinal stenosis as follows: central degenerative spinal stenosisalone, spinal stenosis owing to or in association with degenerative spondylolisthesis, andspinal stenosis in association with degenerative scoliosis (30).

316 Wardlaw

The symptoms produced do not seem to have any direct relationship to the degreeof apparent stenosis seen on radiological, computed tomography (CT), or magneticresonance imaging (MRI) images (31). It would appear that a process like this hap-pening slowly in adult life does allow the neural tissue to accommodate to the chang-ing environment because often when patients become symptomatic, the actual degreeof stenosis seen on MRI and CT scan is surprisingly severe and a point comes atwhich the blood supply to the nerve roots cannot accommodate for increased demandon walking. The leg pain produced is neurogenic pain and not muscle ischemia (32).Whether the symptoms are owing primarily to reduced arterial flow or venousobstruction or a combination of both needs to be ascertained (14,33,34). Of course,these symptoms are often associated with a feeling of weakness in the leg, as well asparesthesia and other sensory changes. Positional changes occur in the normal spineresulting in variations in the sizes of the spinal and root canals (27,28,35). Classically,these symptoms occur on standing, when the upright posture brings about relativeextension of the lumbar spine compared to the sitting position, compressing the neu-ral tissues and narrowing the root canals. The pain then increases with exercise, whenthere is increased demand on the neural tissues, and is relieved at rest, particularly bysitting down. In the sitting position, the lumbar spine is relatively flexed comparedwith the standing position and, therefore, the spinal canal and nerve root exit fora-mena are opened up, allowing improved blood supply to neural tissue and relief ofsymptoms (6,14,36–41). Patients often stand or walk stooped forward to relievesymptoms (42).

TREATMENT

When symptoms are mild, conservative treatment has been shown to be successful inapproximately half of patients (22). When conservative treatment fails, surgery isalways a subsequent option. If surgical treatment is the treatment of choice, this mayconsist of unilateral or bilateral laminectomy, medial or undercutting facetectomy, lim-ited laminotomy procedures with partial or undercutting facetectomy, or removal of discherniations and osteophytes from the vertebral margins or from the facets. Traditionally,the thick or “hypertrophic” ligamenta flava are removed (43–57). There is evidence ofbone regrowth following decompression procedures, which may lead to a recurrence ofsymptoms (30,58). Spinal fusion is often performed, particularly if back pain is a domi-nant feature of the symptom complex or if bony decompression is likely to result in anunstable situation such as degenerative spondylolisthesis, when posterolateral or cagefusion combined with pedicle screw fixation is often advocated (46). Spinal fusion hasbeen shown to alter the biomechanics of the segment above or adjacent to the fusion(59–66) and potentially carries the risk of future degenerative change in that segment(67–80).

My own surgical treatment of choice has been mobilization of the ligamentum flavumfrom bone, leaving it attached medially; preservation of the flavum; and bony decom-pression followed by reposition of the flavum. Preserving the ligamentum flavum untilthe bony decompression has been carried out significantly reduces the complication rate interms of root injury and dural tears, and the ligamentum flavum can continue to perform itsnormal function in protecting and separating the dura from the more superficial muscula-ture. It is certainly true that the ligamentum flavum may be thickened and hypertrophied,

Interspinous Implant for Spinal Stenosis 317

tending to get in the way. It has two layers. The outer interlaminar layer is the one thatbecomes thickened, and it can be separated from the inner layer and removed if necessary,leaving the inner protective layer. The results from this procedure compare very favor-ably with those of other published series of decompression (81).

Lumbar decompression is a major operative procedure, particularly for elderly people,many of whom have comorbidities and are poor risks for general anesthesia (82). Conse-quently, although they may have severe symptoms, they may be deemed poor operativerisks and not offered surgical decompression. Following surgery, patients require a sig-nificant period of in-hospital stay followed by a rehabilitative process of recovery from afairly long general anesthetic, and appropriate support in the community if they are elderly.Furthermore, it is estimated that in the United States the population over the age of 60will almost double by the year 2025 (82). The development of less invasive proceduresto treat this disabling condition in such individuals is therefore important.

Percutaneous endoscopic decompression by means of rongeurs, burrs, and lasercannot adequately deal with lateral recess or central canal stenosis. It can deal withforaminal stenosis, but there is a significant incidence of flaring up of root pain owingto manipulation or heating of the nerve root. Fortunately, this usually dissipates after afew weeks.

A novel approach to deal with this problem is the X-Stop interspinous process distrac-tion device (Fig. 1A). The instrumentation is very simple and effective (Fig. 1B). Manyother similar devices are now coming on the market. In addition, other devices that couldbe used to have a similar effect are on the market. However, for their insertion, theyrequire excision of the supra- and interspinous ligaments, which are essential to ensurethat overdistraction cannot occur and, in my opinion, are essential to maintain stability ofthe motion segment. Furthermore, some devices cut into the base of the spinous process,which must inevitably induce a stress riser, risking stress fracture of the process.

RESULTS OF INTERSPINOUS PROCESS DISTRACTION DEVICE

The X-Stop has been trialed in a prospective randomized study comparing surgicaltreatment by insertion of the device to nonoperative, conservative management (83).Two hundred patients were initially enrolled in the study, and 100 were treated by theX-Stop and 91 by nonoperative therapy as the control group. The Zurich ClaudicationQuestionnaire (ZCQ), validated for lumbar spinal stenosis, was the primary outcomemeasure. It measures physical function, symptom severity, and patient satisfaction. Inser-tion of the X-Stop has quite clearly shown a significant difference from conservativetreatment in terms of clinical improvement, satisfaction rate, and success rate. Two ofthe X-Stop patients had the implants removed at 1 yr, three withdrew from the study,and three went on to have a laminectomy. In the nonoperative group, 12 patients with-drew from the study and 17 went on to have a laminectomy. There was no significantdifference in the ZCQ between the groups preoperatively. At 6 wk, the success rate was52% for the X-Stop group, and 10% for the control group. At 6 mo, the success rateswere 52 and 9%, and at 1 yr, 59 and 12%, respectively. The success rate is comparablewith that for other published series for surgical decompression, but with considerablylower morbidity.

In vitro biomechanical testing in cadaver spines has demonstrated that insertion ofthe interspinous process distraction device increased the dimensions of the spinal and

318 Wardlaw

neural canal at the implanted level and tended to hold the segment in flexion (84). It didnot alter the dimensions of the adjacent intact levels. Distraction also brought abouta reduction in pressure on the posterior annulus and nucleus at the implanted levelswithout altering the pressures at adjacent level. In effect, the device places the segmentin the sitting position, which is the position in which patients typically obtain relief ofsymptoms, by increasing the space within the spinal and nerve root canals.

Radiological studies of the sagittal alignment have demonstrated subtle changes of<1° in patients implanted at one level and a change of just over 1° in patients implanted attwo levels (85).

Interspinous Implant for Spinal Stenosis 319

Fig. 1. (A) X-Stop device; (B) instrumentation for the device (from left to right): small dila-tor, intermediate dilator, distraction and measuring instrument, holder for main body of device,holder for wing, torque screwdriver.

OPERATIVE TECHNIQUE

It is my experience that all one-level and many two-level implantation procedurescan be carried out under local anesthetic block. Sedation or light general anesthetic maybe given in addition. The limiting factor is the safe amount of local anesthetic one cangive to individual patients. The procedures are carried out with an anesthetist presentand quite often some sedation is added. In obese patients, in whom it is very difficult toobtain adequate anesthesia with local anesthetic, and for three-level procedures, gener-ally a light general anesthetic is used in addition.

Following local anesthetic or light general anesthetics, a few hours after the procedurepatients are allowed to go to the toilet or are taken for a short walk. If back pain is a signif-icant component of patients’ symptoms prior to the procedure, then it tends to be a problemin the immediate postoperative period but generally rapidly subsides. Patients are givenappropriate analgesia and reassured that being mobile is the best thing to do. In the first fewweeks, sitting for short periods is permitted, but generally patients are advised to rest lyingdown, rather than sitting, and otherwise to be up and about and to go for walks.

For the implantation procedure, the patient is placed in the right lateral position. Thelevel to be operated on is checked by an image intensifier. Local anesthetic is used in theform of 1% lignocaine or Xylocaine® in 1:200,000 adrenaline. This is used to anesthetizethe skin, the superficial tissues, and the muscles and midline structures. In addition, 0.5%bupivacaine in 1:200,000 adrenaline is used to anesthetize the muscles and deeper struc-tures. This is best done by injecting local anesthetic directly into the muscle bulk and thenadjacent to the facet joint below the pars at the level of, and the facet joints above, thelevel for surgery. It generally takes the local anesthetic 5–10 min to work effectively, andthe long-acting bupivacaine provides good postoperative pain relief for up to 8 h.

The procedure is then performed by centering the skin incision over the inter-spinous process space. The lumbar fascia is incised longitudinally on either side of thesupraspinous ligament, leaving the supraspinous ligament and interspinous ligamentscompletely intact. The muscles are separated from the spinous processes down to thelaminae on either side. A small dilator is then inserted through the interspinous liga-ment at the base of the spinous processes and the hole dilated. An intermediate dilatoris inserted followed by the distraction device. The distraction device is then employed,with the surgeon using his or her hand to feel the tension developing. Initially, distrac-tion is continued until the tissues become tight. The distraction device is then locked inthat position and left for a few minutes. At the same time, the supraspinous ligament ispalpated with the index finger of the surgeon’s other hand to feel the tension in it. Usuallyafter a few minutes, owing to the creep in the tissues, the surgeon finds that further dis-traction is possible and continues distraction until he or she can feel that the tissues arecompletely distracted and taut. It is recommended that in patients who are perhapsslightly osteoporotic, such distraction be carried out very carefully.

At this point, it is possible to read off from the distraction device the size of the X-Stop device to be inserted (Fig. 2). The largest possible size should be inserted. The twoparts of the device are then attached to their introducers (Fig. 3), and then the patient isasked to curl up in the fully flexed position, bending knees up to the chest and head andchin over the top of the knees. In a patient relaxed by a general anesthetic, this is notnecessary. This position opens up the interspace to the optimum while the distraction

320 Wardlaw

Interspinous Implant for Spinal Stenosis 321

Fig. 2. Distraction and measuring instrument showing size 14.

Fig. 3. X-Stop device attached to introducers.

device is removed and the x-stop device inserted. It is usually a little difficult to insertthe device in the first instance, but it will slip into place with gentle, persistent pressure(Fig. 4). The retaining wing is then screwed onto the device and tightened off with thetorque screwdriver (Fig. 5).

322 Wardlaw

Fig. 4. Main body of X-Stop device inserted.

Fig. 5. X-Stop device in position.

Interspinous Implant for Spinal Stenosis 323

Fig. 6. (A) Lateral X-ray of device at L4/L5; (B) AP X-ray of device at L4/L5.

Owing to the use of local anesthetic in 1:200,000 adrenaline, there is very littlebleeding. The wound is then closed by resuturing the lumbar fascia to the supraspinousligaments on either side with one or two sutures, as necessary. The skin is then closed.To date, 30 patients have been operated on, many of whom had significant comorbidi-ties, without significant perioperative complications.

X-ray image intensification is used throughout the procedure to ensure that a correctlevel is inserted. It is also reassuring to the surgeon to be able to view the distractionbeing obtained prior to inserting the device. X-rays are taken to demonstrate the finalposition of the device (Fig. 6A,B).

MOST RECENT STUDIES

In Aberdeen, Scotland, we are carrying out studies using the Fonar positional MRIscan to measure a range of parameters before and after insertion of the device. Theseare total flexion and extension of the lumbar spine (Fig. 7A,B), the AP distance andcross-sectional area of the spinal canal (Fig. 8), the cross-sectional area of the nerveroot foramen (Fig. 9), and the effect on the flexion and extension movement of the seg-ments above and below the treated segment. There also appears to be an effect on thecoronal alignment in patients who have degenerative scoliosis in that insertion of thedevice seems to reduce the degree of lateral bend at the treated level.

The number of postoperative measurements made is small; however, there is a cleartrend showing that there is no significant difference in the total range of flexion and

324 Wardlaw

Fig. 6. (Continued)

extension of the lumbar spine. There is a highly significant increase in the AP distanceand cross-sectional area of the spinal canal and the cross-sectional area of the nerve rootexit foramen at the treated level. There is no evidence of any difference in flexion andextension movement at the levels above and below the treated level. For the positionalMRI scan study, the plan is to treat some 45 patients who have been treated at one, two,and three levels.

Figure 7A,B shows MRIs of a patient (patient 1) with two-level disease in flexionand extension prior to insertion of the device. The patient was a 75-yr-old female with ahistory of bilateral leg pain, worse on the right side, of increasing severity over an 18-moperiod unrelieved by conservative therapies including nonsteroidal anti-inflammatory andanalgesic therapy, physiotherapy, and sacral epidural injection. She had no pain on sittingand lying on her side but had some pain lying on her back. Pain occurred immediatelyon standing, and she preferred to stand and walk stooped forward using a walking stick.Her walking distance was 20 yd maximum, and she developed weakness and pares-thesia in her lower legs and feet. She had surgery under local anesthesia and a lightsedation. Six hours afterward, she was up walking without pain and able to standupright. At 6-mo follow up, there was no restriction in her walking distance; she wasable to shop, do gardening, and play golf; and her positional MRI scan in the vertical

Interspinous Implant for Spinal Stenosis 325

Fig. 7. (A) Positional MRI scan of patient 1 in flexion; (B) positional MRI scan of patient 1in extension.

position (Fig. 10) demonstrated the X-Stop in position and its effect on the spinal canal.The midsagittal AP distance at L3/L4 and L4/L5 were clearly improved with relief ofthe stenosis.

Patient (patient 2) is an extremely fit 85-yr-old man who was unable to walk morethan 50 yards owing to left leg pain radiating from the buttock down the posterolateralthigh to the foot. X-ray demonstrated a degenerative spondylolisthesis at L4/L5, andpositional MRI scans confirmed stenosis (Fig. 11A,B). The treatment options were

326 Wardlaw

Fig. 8. Spinal canal measurement.

Fig. 9. Foraminal measurement.

discussed with him and he preferred to pursue nonoperative treatments. He was offeredan epidural injection, with the option of an X-Stop if it failed to relieve his symptoms.He chose to seek a second opinion. He was offered and treated with a left L5 nerve rootinfiltration, which relieved pain for only 2 wk, again with the option of an X-Stop if it

Interspinous Implant for Spinal Stenosis 327

Fig. 10. Positional MRI scan of patient no. 1 with X-Stop in position at L3/L4 and L4/L5 at6-mo follow-up.

failed to relieve his symptoms. He decided to opt for the X-Stop treatment, which imme-diately relieved his symptoms. At 6 wk, he was walking 3 mi/d, and at 6 mo, 5 mi/d,enjoying fishing and shooting, his favorite pastimes. His positional MRI scan at 6 modemonstrates the current position (Fig. 12A,B).

So far, I have treated 30 patients. Clinical improvement has taken place in all of thesepatients. Perhaps the most striking effect in most patients is the immediate pain relief andability to stand up straight, rather than stooped forward. Walking distance is also improvedimmediately. There have been no significant complications. However, follow-up is shortat the present time, with 10 patients followed to 6 mo, and the long-term results will bepublished in due course.

CONCLUSION

Treatment of spinal stenosis by means of an interspinous process distraction devicesuch as the X-Stop appears to be a significant improvement in the treatment of this con-dition. The procedure can be carried out under local anesthetic as a day case or, inelderly patients, with one overnight stay. Mobilization is quick and patients do notrequire any specific postoperative rehabilitation or extra social support after dischargehome. A small number of patients do have increased back pain, but this does dissipatewithin a few days, or certainly within a few weeks. The procedure is extremely safeand, so far, no significant complications have been encountered. It is accepted that if the

328 Wardlaw

Fig. 11. (A) Positional MRI scan of patient 2 showing stenosis at L4/L5; (B) positional MRIscan of patient 2 showing stenosis at L4/L5 axial view.

Interspinous Implant for Spinal Stenosis 329

Fig. 12. (A) Positional MRI scan of patient 2 showing X-Stop at L4/L5 at 6 mo; (B) pos-tional MRI scan of patient 2 following X-Stop at L4/L5 at 6 mo.

device fails, the patient still has the option of surgical decompression in the traditionalmanner. The long-term results of the original prospective study demonstrate that thereis no deterioration in symptoms over time. This is clearly a minimal intervention proce-dure that is safe and effective, particularly for elderly patients who are otherwise infirmand unfit for a full general anesthetic. They can quickly return home and resume normalactivities with an immediate improvement in symptoms.

REFERENCES

1. Sachs B, Frankel V. Progressive and kyphotic rigidity of the spine. J Nerv Ment Dis 1900;27:l.2. Putti V. Lady Jones lecture: on new concepts on the pathogenesis of sciatic pain. Lancet

1927;210:(5491)53–60.3. Hirsch C. On lumbar pacetectomies. Acta Orthop Scand 1948;17:240–252.4. Harris RI, MacNab I. Structural changes in the lumbar intervertebral discs: their relation-

ship to low back pain and sciatica. J Bone Joint Surg 1954;386:304–322.5. MacNab I. Negative disc exploration. J Bone Joint Surg 1971;53A:891–903.6. Verbiest H. Neurogenic intermittent claudication in cases with absolute and relative steno-

sis of the lumbar vertebral canal (ASLC and RSLC), in cases with narrow lumbar interver-tebral foramina, and in cases with both entities. Clin Neurosurg 1973;20:204–214.

7. Verbiest H. Results of surgical treatment of idiopathic developmental stenosis of thelumbar vertebral canal: a review of twenty-seven years’ experience. J Bone Joint Surg1977;59B(2).

8. Verbiest H. Stenosis of the lumbar vertebral canal and sciatica. Neurosurg Rev 1980;3:75–89.9. Verbiest H. The treatment of lumbar spondyloptosis or impending lumbar spondyloptosis

accompanied by neurologic deficit and/or neurogenic intermittent claudication. Spine1979;4:68–77.

10. Porter RW. Variations in shape and size of the lumbar spinal canal. Paleopathol Newsle1980;32:9–11.

11. Eisenstein S. The morphometry and pathological anatomy of the lumbar spine in SouthAfrican Negroes and Caucasoids with special reference to spinal stenosis. J Bone JointSurg 1977;59B:172–180.

12. Porter RW, Wicks M, Ottwell D. Measurement of the spinal canal by diagnostic ultrasound.J Bone Joint Surg (Br) 1978;60B(4):481–484.

13. Kadziolka R, AszteIy M, Hanai K, Hansson T, Nachemson A. Ultrasonic measurement ofthe lumbar spinal canal: the origin and precision of the recorded echoes. J Bone Joint Surg1981;63B(4):504–507.

14. Porter RW, Hibbert CS, Wicks M. The spinal canal in symptomatic lumbar disc lesions. JBone Joint Surg (Br) 1978;60B(4):485–487.

15. Papp T, Porter RW, Aspden RM. The growth of the lumbar vertebral canal. Spine1994;19(24):2770–2773.

16. Papp T, Porter RW, Aspden RM. Trefoil configuration and developmental stenosis of thelumbar vertebral canal. J Bone Joint Surg (Br) 1995;77B:469–472.

17. Arnoldi CC, Brodsky AE, Cauchoix J, et al. Lumbar spinal stenosis and nerve rootentrapment syndromes: definition and classification. Clin Orthop 1976; (115):4–5.

18. Hawkes CH, Roberts GM. Neurogenic and vascular claudication. J Neurol Sci1978;38:337–345.

19. Salibi BS. Neurogenic intermittent claudication and stenosis of the lumbar spinal canal.Surg Neurol 1976;5:269–272.

20. Cinotti, G, De Santis P, Nofroni I, Postacchini F. Stenosis of lumbar intervertebral fora-men: anatomic study on predisposing factors. Spine 2002;27(3):223–229.

21. Panjabi MM, Takata K, Goel VK. Kinematics of lumbar intervertebral foramen. Spine1983;8:348–357.

330 Wardlaw

22. Araksinen O, Herno A, Turunen V, Saari T, Suomlainen O. Degenerative spondylolisthesis:surgical outcome of 438 patients treated surgically for lumbar spinal stenosis. Spine 1997;22:2278–2282.

23. Amundsen T, Weber H, Nordal HJ, Hele J, Magnaes B, Adelnoor MMPH, Lilleas F. Lumbarspinal stenosis: conservative or surgical management? a prospective 10 year study. Spine2000;25:1424–1436.

24. Herkowitz HN, Kurz LT. Degenerative lumbar spondylolisthesis with spinal stenosis. JBone Joint Surg (Am) 1991;73:802–807.

25. Mardjetko SM, Connolly PJ, Short S. Degenerative lumbar spondylolisthesis: a meta-analysis of literature 1970–1993. Spine 1994;19:2257S–2265S.

26. Crock HV. Normal and pathological anatomy of the lumbar spinal nerve root canals. J BoneJoint Surg (Br) 1981;63:487–490.

27. Adams MA, May S, Freeman BJ, et al. Effects of backward bending on lumbar interverte-bral discs: relevance to physical therapy treatments for low back pain. Spine2000;25:431–437: discussion, 438.

28. Chung SS, Lee CS, Kim SH, et al. Effect of low back posture on the morphology of thespinal canal. Skeletal Radiol 2000;29:217–223.

29. Rauschning W. Normal and pathologic anatomy of the lumbar root canals. Spine1987;12:1008–1019.

30. Postacchini F, Cinotti G. Bone regrowth after surgical decompression for lumbar spinalstenosis. J Bone Joint Surg (Br) 1992;74:862–869.

31. Amundsen T, Weber H, Lilleas F, et al. Lumbar spinal stenosis: clinical and radiologicfeatures. Spine 1995;20:1178–1186.

32. Keenan GF, Ashcroft GP, Roditi GH, et al. Measurement of lower limb blood flow inpatients with neurogenic claudication using positron emission tomography. Spine1995;20(4):408–411.

33. Naylor A. Factors in the development of the spinal stenosis syndrome. J Bone Joint Surg(Br) 1979;61B(3):306–309.

34. Porter RW. Spinal stenosis and neurogenic claudication. Spine 1996;21:2046–2052.35. Schmid MR, Stucki G, Duewell S, et al. Changes in cross-sectional measurements of the

spinal canal and intervertebral foramina as a function of body position: in vivo studies onan open-configuration MR system. Am J Roentgenol 1999;172:1095–1102.

36. Fujiwara A, An HS, Lim TH, et al. Morphologic changes in the lumbar intervertebral fora-men due to flexion-extension, lateral bending, and axial rotation an in vitro anatomic andbiomechanical study. Spine 2001;26:876–882.

37. Inufusa A, An HS, Lim TH, et al. Anatomic changes of the spinal canal and intervertebralforamen associated with flexion-extension movement. Spine 1996;21:2412–2420.

38. Penning L, Wilmink JT. Posture-dependent bilateral compression of L4 or L5 nerve rootsin facet hypertrophy: a dynamic CT-myelographic study. Spine 1987;12:488–500.

39. Rothman SL. Dynamic effect on the lumbar spinal canal. Spine 1998;23:1506, 1507.40. Vitzthum HE, Konig A, Seifert V. Dynamic examination of the lumbar spine by using vertical,

open magnetic resonance imaging. J Neurosurg 2000;93:58–64.41. Wildermuth S, Zanetti M, Duewell S, et al. Lumbar spine: quantitative and qualitative

assessment of positional (upright flexion and extension) MR imaging and myelography.Radiology 1998;207:391–398.

42. Dyck P. The Stoop test in lumbar entrapment radiculopathy. Spine 1979;4:89–92.43. Aryanpur J, Ducker T. Multilevel lumbar laminotomies: an alternative to laminectomy in

the treatment of lumbar stenosis. Neurosurgery 1990;26:429–433.44. Cinotti G, Postacchini F, Weinstein JN. Lumbar spinal, stenosis and diabetes: outcome of

surgical decompression. J Bone Joint Surg (Br) 1994;76:215–219.45. Getty CJM. Lumbar spinal stenosis: the clinical spectrum and results of operations. J Bone

Joint Surg (Br) 1980;62:481–485.

Interspinous Implant for Spinal Stenosis 331

46. Grob D, Humke T, Dvorak J. Degenerative lumbar spinal stenosis: decompression with andwithout arthrodesis. J Bone Joint Surg (Am) 1995;77:1036–1041.

47. Herno A, Airaksinen O, Saari T. Long-term results of surgical treatment of lumbar spinalstenosis. Spine 1993;18:147l–1474.

48. Herron LD, Mangelsdorf C. Lumbar spinal stenosis. J Spinal Disord 1991;4:26–33.49. Hilibrand AS, Rand N. Degenerative lumbar stenosis: diagnosis and management. J Am

Acad Orthop Surg 1999;7:239–249.50. Jonsson B, Annertz M, Sioberg C, et al. A prospective and consecutive study of surgically

treated lumbar spinal stenosis. Part II: five-year follow-up by an independent observer.Spine 1997;22:2938–2944.

51. Jonsson B, Stromqvist B, Decompression for lateral lumbar stenosis: results and impact onsick leave and working conditions. Spine 1994;19:2381–2386.

52. Katz JN, Lipson SJ, Chang LC, et al. Seven- to 10-year outcome of decompressive surgeryfor degenerative lumbar spinal stenosis. Spine 1996;21:92–98.

53. Katz JN, Lipson SJ, Larson MG, et al. The outcome of decompressive laminectomy fordegenerative lumbar stenosis. J Bone Joint Surg (Am) 1991;73:809–816.

54. Katz JN, Lipson SJ, Larson MG, Mclnnes JM, Fossel AH, Liang MH. The outcome ofdecompressive laminectomy for degenerative lumbar spinal stenosis. J Bone Joint Surg(Am) 1991;73:809–813.

55. Postacchini F, Cinotti G, Gumina S, et al. Long-term results of surgery in lumbar stenosis:8-year review of 64 patients. Acta Orthop Scand 1993;64 78–80.

56. Postacchini F, Cinotti G, Perugia D, et al. The surgical treatment of central lumbar stenosis:multiple laminotomy compared with total laminectomy. J Bone Joint Surg (Br) 1993;75:386–392.

57. Postacchini F. Surgical management of lumbar spinal stenosis. Spine 1999;24:1043–1047.58. Chen O, Baba H, Kamitani K, et al. Postoperative bone re-growth in lumbar spinal stenosis:

a multivariate analysis of 48 patients. Spine 1994;19:2144–2149.59. Chow DH, Luk KD, Evans JH, et al. Effects of short anterior lumbar interbody fusion on

biomechanics of neighboring unfused segments. Spine 1996;21:549–555.60. Cunningham BW, Kotani Y, McNulty PS, et al. The effect of spinal destabilization and

instrumentation on lumbar intradiscal pressure: an in vitro biomechanical analysis. Spine1997;22:2655–2663.

61. Ha KY, Schendel MJ, Lewis JL, et al. Effect of immobilization and configuration on lum-bar adjacent-segment biomechanics. J Spinal Disord 1993;6:99–105.

62. Kim YE, Goel VK, Weinstein JN, et al. Effect of disc degeneration at one level on the adja-cent level in axial mode. Spine 1991;16:331–335.

63. Marchesi DG, Aebi M. Pedicle fixation devices in the treatment of adult lumbar scoliosis.Spine 1992;17:S304–S309.

64. Rohlmann A, Calisse J, Bergmann G, et al. Internal spinal fixator stiffness has only a minorinfluence on stresses in the adjacent discs. Spine 1999;24:1192–1195.

65. Rohlmann A, Neller S, Bergmann G, et al. Effect of an internal fixator and a bone graft onintersegmental spinal motion and intradiscal pressure in the adjacent regions. Eur Spine J2001;10:301–308.

66. Weinhoffer SL, Guyer RD, Herbert M, et al. Intradiscal pressure measurements above aninstrumented fusion: a cadaveric study. Spine 1995;20:526–53l.

67. Aota Y, Kumano K, Hirabayashi S. Postfusion instability at the adjacent segments afterrigid pedicle screw fixation for degenerative lumbar spinal disorders. J Spinal Disord1995;8:464–473.

68. Etebar S, Cahill DW. Risk factors for adjacent-segment failure following lumbar fixationwith rigid instrumentation for degenerative instability. J Neurosurg 1999;90:163–169.

69. Frymoyer JW, Hanley EN Jr, Howe J, et al. A comparison of radiographic findings infusion and nonfusion patients ten or more years following lumbar disc surgery. Spine1979;4:435–440.

332 Wardlaw

70. Guigui P, Lambert P, Lassale B, et al. Long-term outcome at adjacent levels of lumbararthrodesis. Rev Chir Orthop Reparatrice Appar Mot 1997;83:685–696.

71. Kumar MN, Baklanov A, Chopin D, Correlation between sagittal plane changes and adja-cent segment degeneration following lumbar spine fusion. Eur Spine J 2001;10:314–319.

72. Kumar MN, Jacquot F, Hall H. Long-term follow-up of functional outcomes and radio-graphic changes at adjacent levels following lumbar spine fusion for degenerative disc dis-ease. Eur Spine J 2001;10:309–313.

73. Lee CK. Accelerated degeneration of the segment adjacent to a lumbar fusion. Spine1988;13:375–377.

74. Lehmann TR, Spratt KF, Tozzi JE, et al. Long-term follow-up of lower lumbar fusionpatients. Spine 1987;12:97–104.

75. Miyakoshi N, Abe E, Shimada Y, et al. Outcome of one-level posterior lumbar interbodyfusion for spondylolisthesis and postoperative intervertebral disc degeneration adjacent tothe fusion. Spine 2000;25:1837–1842.

76. Nagata H, Schendel MJ, Transfeldt EE, et al. The effects of immobilization of long seg-ments of the spine on the adjacent and distal facet force and lumbosacral motion. Spine1993;18:2471–2479.

77. Schlegel JD, Smith JA, Schleusener RL. Lumbar motion segment pathology adjacent tothoracolumbar, lumbar, and lumbosacral fusions. Spine 1996;21:970–981.

78. Shono Y, Kaneda K, Abumi K, et al. Stability of posterior spinal instrumentation and itseffects on adjacent motion segments in the lumbosacral spine. Spine 1998;23:1550–1558.

79. Van Horn JR, Bohnen LM. The development of discopathy in lumbar discs adjacent to alumbar anterior interbody spondylodesis: a retrospective matched-pair study with a postop-erative follow-up of 16 years. Acta Orthop Belg 1992;58:280–286.

80. Whitecloud TS III, Davis JM, Olive PM. Operative treatment of the degenerated segmentadjacent to a lumbar fusion. Spine 1994;19:531–536.

81. Askar Z, Wardlaw D, Choudhary S, Rege A. A ligamentum flavum-preserving approach tothe lumbar spinal canal. Spine 2003;28(19):E385–E390.

82. Reeg SE. A review of comorbidities and spinal surgery. Clin Orthop 2001;384:101–109.83. Zucherman JF, Hsu KY, Hartjen CA, et al. A prospective randomized multi-center study for

the treatment of lumbar spinal stenosis with the X STOP interspinous implant: 1 yearresults. Eur Spine J 2004;13:22–31.

84. Swanson KE, Lindsey DP, Hsu KY, Zucherman JF, Yerby SA. The effects of an inter-spinous implant on intervertebral disc pressure. Spine 2003;28(l):26–32.

85. Lindsey DP, Swanson KF, Fuchs P, Hsu KY, Zucherman JF, Yerby SA. (2003) The effectsof an interspinous implant on the kinematics of the instrumented and adjacent levels in thelumbar spine. Spine 28:2192–2197.

Interspinous Implant for Spinal Stenosis 333

17Frameless Stereotactic Imaging Techniques

in Minimally Invasive Spinal Surgery

Kern Singh, MD, Laurence N. Fitzhenry, MD,and Alexander R. Vaccaro, MD

INTRODUCTION

Frameless stereotactic imaging is widely used in intracranial neurosurgical proceduresto locate pathology precisely and decrease the morbidity of traditional open exploratoryprocedures. However, the clinical application of computer-assisted image-guided spinalsurgery has not been widely accepted owing to the time-consuming and arduous natureof maintaining accurate registration coordinates. Frameless spinal stereotaxy is anevolving adjunct to the surgeon’s anatomical knowledge and skill. Newer technologyinvolving intraoperative real-time image acquisition and user-friendly instruments willundoubtedly improve the popularity of this technology in complex spinal procedures.Currently, computer image-guided spinal surgery is used for the placement of pediclescrews; transarticular C1-C2 fixation; transoral odontoid resection; anterior cervicalcorpectomies; and, most recently, in endoscopic spinal surgery.

STEREOTAXIS

Computer-assisted navigation systems are based on the principle of stereotaxis. Thebasic components of the original “frame-based” stereotaxis included the surgical object(e.g., the cervical spine), the virtual object (an image of the cervical spine obtainedfrom computed tomography (CT) or magnetic resonance imaging (MRI) and a frame(navigator) fixed to the surgical object (cervical spine) for guidance of the surgicalinstruments according to the coordinates extrapolated from the preoperative images.“Frameless” sterotaxis is possible now owing to further development in medical imagingand computer science. The exact coordinates of the surgical object can be defined in athree-dimensional (3D) data set (CT or MRI) of the region of interest. The surgeon navi-gates, the instruments via a motion analysis system without any mechanical frame. Optical(infrared light), magnetic, or acoustic (ultrasound) signals are utilized to localize theposition of the instrument or implant in space.

A major problem in applying frameless stereotaxy to the spine is the relative mobilitybetween adjacent vertebrae and the overlying soft-tissue structures. As a result, frame-less stereotaxy in the spine requires registration that is based on the exposed anatomy of

335

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

the vertebrae. Navigation is then reliable only on that registered anatomical structure.To navigate on adjacent vertebrae, the registration process has to be repeated. As onecan see, the process can become time-consuming, especially when the registration pro-cess is not proceeding fluidly.

IMAGE ACQUISITION

Image-guided spinal surgery typically begins with preoperative acquisition of 3Dimages using CT or MRI. A CT scanner is an effective tool for collecting preoperativeimages with thin-slice (approx 1 mm) imaging. The image data are then transmitted viaDICOM interfaces to the picture archiving and communication system (PACS) and intothe image-guided spinal surgery local area network. The connectivity of the scannersand the PACS interfaces are the most important aspects of successfully orchestratingimage-guided therapy. Alternatively, a surgery-dedicated server can be used, in conjunc-tion with the existing PACS system, for presurgical planning and delivery of postprocessedimages to the operating room.

The main limitation in existing stereotactic systems is that images are acquired beforesurgery and are not updated to reflect intraoperative changes. In addition, existing framelessstereotactic systems for spinal surgery do not provide localization of the spinal segment,typically obtained using a lateral radiograph, and necessitate manual registration of fiduciallandmarks in the operative field to the image data set acquired before surgery. However,recently, Woodward et al. (1) have described the use of intraoperative MRI (IMRI) as anadjunct in minimally invasive spinal surgery.

The IMRI involves a novel, open-configuration, cryogenless superconducting magnetwith a vertical gap between the two magnetic components that allows direct access tothe patient during imaging (Fig. 1). IMRI provides all the advantages of conventional

336 Singh et al.

Fig. 1. Photograph of open-configuration intraoperative magnetic resonance imager. Themagnet is constructed with two interconnected cryostats with a 56-cm vertical opening permittingaccess to the patient during imaging. (Courtesy of General Electric, Milwaukee, WI.)

MRI including multiplanar image acquisition and excellent definition of soft-tissueanatomy. Using IMRI, 12 patients underwent successful spinal surgery involving threelumbar discectomies, three anterior cervical discectomies and fusions, three cervical ver-tebrectomies with allograft fusions, two cervical foraminotomies, and one decompres-sive cervical laminectomy. The primary advantage of IMRI is that it provides accuratereal-time imaging that can be used in surgical planning and assessment. However, aswith conventional MRI, IMRI affords limited depiction of osseous anatomy, which maylimit its role in spinal surgery. Nevertheless, the improved real-time imaging afforded byIMRI makes this an attractive intraoperative adjunct to spinal surgery in the future.

PRESURGICAL PLANNING

Preparation for the use of computer-assisted image guidance depends on the specificrequirements of the surgical procedure. All procedures, however, require special softwaretools to manipulate the acquired images in order to create a graphic model of the surgicalanatomy. Prior to surgery, CT or MRI images of the spine are obtained using a 3D imagingprotocol. The scan is transferred to the computer workstation via a dedicated imagingnetwork. A 3D image of the spine is segmented from the slice images obtained prior tosurgery (Fig. 2). Morphological operations, such as thresholding and voxel connectivity,can be used on the images to create a 3D image of the bony surface of the spine. Inaddition, 3D CT images can be used for standard orthogonal views as well as oblique

Frameless Stereotactic Imaging Techniques 337

Fig. 2. Typical 3D reconstruction image obtained from thin-slice (1-mm) CT of lumbosacraljunction.

cuts that display the planned trajectory of the surgical probe. For example, in a lumbarpedicle screw case, these visualizations can help in determining the trajectory anddiameter of a pedicle screw that is needed, as well as the depths of screw insertion. Surgicalplanning is done in the virtual space of the 3D image model, and relationships to soft tis-sues, such as spinal cord and critical arteries, can be defined during presurgical planning.

SURGICAL NAVIGATION

Components of a surgical navigation system include a computer workstation with anetwork interface, surgical instruments that may be tracked via light-emitting diodes, areference arc or other method for tracking patient anatomy, and an optical camera thattracks the reference arc and the surgical probes (Figs. 3 and 4). The technical require-ments for performing image-guided surgery may vary among different systems, but con-ceptually the methods are very similar.

The surgical navigation system enables free-hand navigation of the surgical space andframeless stereotactic surgery through image guidance. It allows a surgeon to track bothpatient anatomy and instruments in three dimensions with improved accuracy. Error in

338 Singh et al.

Fig. 3. Typical optoelectric navigational system. The Stealth system (courtesy of MedtronicSofamor Danek, Memphis, TN) is depicted with the optical tracking apparatus and the intraoper-ative computer imaging display.

localizing the target in a phantom is typically 2.5 mm or less. Real-time manipulation ofthe computer model allows a surgeon to localize surgical position on preoperativeimages. The optical tracking technology permits real-time, dynamically referencedlocalization of patient anatomy and surgical instrumentation. By equipping standardsurgical instruments with this technology, a surgeon can operate with familiar instru-mentation that is coupled with the patient’s diagnostic images.

Typically the reference arc is attached to the spinous process of the involved vertebra(Fig. 5). Without the reference frame, even slight movement by the patient wouldadversely affect navigational accuracy. Therefore, the patient’s anatomy must be definedin conjunction with the position of the reference frame. It is logical to assume that rigidfixation of the reference frame must be performed for each vertebral segment requiringsurgical intervention. The smallest movement of the reference frame in relationship to thespinous process may result in drastic changes in navigational accuracy. The frame isattached to the patient’s spine via either a clamp or pin. The bony landmarks of the spineare exposed and identified by touching the surgical probe to identifiable points on the sur-face anatomy. The camera array dynamically tracks the reference arc and motion of thevertebral segments to which the arc is attached.

Frameless Stereotactic Imaging Techniques 339

Fig. 4. Photograph showing close-up of wireless dynamic reference frame with attachedclamp that connects to StealthStation. (Courtesy of Medtronic Sofamor Danek, Memphis, TN.)

Registration accuracy is a critical variable in the use of image-guided spinal surgery.Accurate and reliable identification of anatomical landmarks in the spine is relativelydifficult. Misregistration of anatomical landmarks with preacquired imaging landmarksis a major factor in potential inaccuracy of these systems. Although alternative methodsusing internal fiducial markers have been tested (2–9), the reliance on the identificationof anatomical landmarks remains the most widely used technique for coregistration ofimage and patient coordinate spaces. Minimizing registration error involves centeringthe target over the anatomical landmark and increasing the number of and distancebetween anatomical landmarks.

The average misalignment, or gap, between corresponding landmarks may bereported by the surgical navigation workstation as the mean fiduciary error. The meanfiduciary error describes how closely points in the image correspond to points on theanatomy of the patient. The surgeon should note that the mean fiduciary error is not atrue estimate of the accuracy of the system and that a perfect registration of points maystill not define precisely the anatomical space. For example, equally spaced points identifiedon a sphere could be perfectly registered to equidistant points on an image of thesphere, yet the two spheres (the actual sphere and its virtual image) could be rotated180° relative to each other. For this reason, noncollinear landmarks that are widelyspaced around the target anatomy are recommended. It should be emphasized that thesurgeon should always perform a global assessment of the registration accuracy byidentifying independent surface points and by visually confirming how well the imagespace and surgical space correlate.

340 Singh et al.

Fig. 5. Photograph showing close-up of dynamic reference frame rigidly attached to spinousprocess of a cadaveric spine. (Courtesy of Medtronic Sofamor Danek, Memphis, TN.)

ALTERNATIVE IMAGING MODALITIES: REAL-TIME FLUOROSCOPY

Compared with current 3D computer image-guided surgery systems, virtual fluo-roscopy is readily applicable to a wide variety of spinal procedures such as pediclescrew insertion, anterior odontoid screw fixation, and interbody cage placement (10–15).Virtual fluoroscopy offers several distinct advantages over computer-assisted imaging,such as avoiding the time and cost of obtaining specially formatted preoperative CTor MRI images. The often time-consuming and frustrating step of image-to-patientregistration is unnecessary. The image calibration process for virtual fluoroscopy is fullyautomated. In addition, real-time image updating for positional changes of the patientafter manipulation of a given spinal segment is easily achieved during surgery by simplyacquiring a new fluoroscopic image. Likewise, real-time intraoperative fluoroscopicvalidation of the virtually displayed instrument position can be obtained at any time,providing a “safety check” (Figs. 6 and 7).

Despite these many advantages, virtual fluoroscopy is a two-dimensional (2D) navi-gational system (Figs. 8 and 9). It does not provide the detailed multiplanar imaginggenerated by 3D systems. Errors in the clinical interpretation of 2D images and theextrapolation of 2D information to 3D anatomy are still dependent on the expertise ofthe spine surgeon. Additionally, virtual fluoroscopy cannot compensate for factorsresulting in poor image quality such as obesity, the presence of bowel contrast aftertrauma assessment, or the inadvertent positioning of radioopaque structures. The effectsof parallax also must be considered when using a virtual fluoroscopy system. Despitethese limitations, computer enhancement of fluoroscopically assisted procedures pro-vides a broad-based and practical application of surgical navigational technology.

CURRENT APPLICATIONS OF IMAGE-GUIDED SURGERY

Because of the complex and varied nature of cervical spine anatomy, instrumentationin the spine can be extremely challenging, even in very experienced hands. Placementof C2-C1 transarticular screws exemplifies the deficiencies in current techniques forspinal navigation. Up to 15–20% of patients lack the required bony volume in theregion of the C2 isthmus to accept safely a screw with a 3.5 mm diameter. Although

Frameless Stereotactic Imaging Techniques 341

Fig. 6. Photograph of FluoroNav system including StealthStation, C-arm with imaging target,and dual-lens camera. (Courtesy of Medtronic Sofamor Danek, Memphis, TN.)

cervical lateral mass screw placement has been considered extremely safe regardingpotential injury to the vertebral artery or spinal cord, Heller et al. (16,17), in a cadavericstudy, noted a 7.3% rate of cervical root injury using the Magerl technique for lateralmass screw placement. They also reported on a clinical series of 78 patients of whomfour developed a symptomatic radiculopathy owing to misplaced screws (5%) (16,17).Clearly, these studies indicate the potential benefit of better intraoperative neurovascularsurveillance and guidance when spinal implants are placed.

The value of spinal imaging becomes more obvious when one realizes the difficultyassociated with so-called hidden spinal structures. Accurate knowledge of spinal anatomyis totally dependent on the surgeon’s experience and the ability to visualize anatomicalstructures three dimensionally (18–22). Weinstein et al. (18) reported that 21% ofinstrumented thoracolumbar pedicles placed using standard fluoroscopy showed evi-dence of cortical perforation. Even more shocking was that 92% of these failuresoccurred medially within the spinal canal. Clearly, the potential for serious neurologicalsequelae exists as a result of misdirected spinal implants using current conventionalimaging techniques.

Several researchers have detailed their experiences with computer-assisted imagingin the cervical spine. Welch et al. (23) performed 11 upper cervical spine proceduresincluding transoral odontoid resection, posterior atlantoaxial fusion with transarticularC2-C1 screw fixation, and spinal tumor resection utilizing CT-guided stereotaxis. Ineach case, frameless stereotaxy was used to plan the incision, delineate the resectionmargins, and determine the appropriate trajection of implant placement. They noted no

342 Singh et al.

Fig. 7. Photograph showing close-up of C-arm with target guide mounted on image intensi-fier. (Courtesy of Medtronic Sofamor Danek, Memphis, TN.)

intraoperative complications, and each patient successfully underwent adequate resec-tion of the pathological lesion and satisfactory placement of cervical instrumentation.

Bolger and Wigfield (24,25) retrospectively reviewed 120 cases in which imaged-guidedsurgery involving the cervical and thoracic spine was utilized. Of these, 34 casesinvolved transarticular atlantoaxial screw placement, 22 cases involved cervical lateralmass plating, 40 cases involved anterior cervical decompressions, and 18 cases involvedvertebrectomies. The authors concluded that the use of image-guided technology providedan enormous amount of anatomical information, particularly in aberrant anatomicalcases, and that it was easier to determine the best screw trajectory on an individualbasis rather then having to rely on population-based, predetermined implant placementrecommendations.

More recently, Richter et al. (26) evaluated whether C2–C1 transarticular anatom-ical screws as well as transpedicular screws in C3 and C4 could be applied safelywith computer-assisted surgery. Hole positions were evaluated by palpation, CT, anddissections. The investigators noted that 48 (92%) of the 52 drilled pedicles were cor-rectly positioned. The vertebral artery was not injured in any specimen. Finally, all of the26 C2–C1 transarticular k-wires were placed properly with no injury to the vascular struc-tures noted. Richter et al. (26) concluded that image guidance greatly improved theaccuracy of implant placement and reduced potential complications.

Frameless Stereotactic Imaging Techniques 343

Fig. 8. Anteropasterior (AP) (open-mouth odontoid)/lateral virtual fluoroscopic image of C1lateral mass and spinous process. (Courtesy of Medtronic Sofamor Danek, Memphis, TN.)

Laine et al. (27) prospectively evaluated 30 adult patients who had undergone place-ment of 174 thoracolumbar pedicle screws. Of those screws, 139 were placed using a CT-guided imaging system, and the remaining 35 were placed without computer assistance.The investigators noted a significantly higher rate of cortical wall perforation withoutthe use of image-guided assistance (14.3 vs 4.3%). In fact, of the screws placed conven-tionally, three perforated the medial wall of the pedicle, with one screw requiring repo-sitioning secondary to nerve root irritation (Fig. 10).

Foley and colleagues presented their experience with the FluoroNav (Medtronic Sur-gical Navigation Technologies, Louisville, CO) image guidance system in the cervicalspine (10,11,28,29). The FluoroNav system was used for three separate cervical spineapplications: odontoid screw placement, lateral mass screw placement, and anterior cer-vical corpectomy and plate placement. By using (virtual) preacquired AP, open-mouth,and lateral views, the investigators were able to successfully place all three odontoidscrews, as well as match the intended midsagittal plane screw angulation in the 24 cervi-cal lateral mass screws placed. The intended angle of projection was 30 in the sagittalplane. The actual measured screw placement measured approx 30.5 , with a range of27–35 . All corpectomy troughs had a symmetric boundary regarding the midline, withan average trough diameter of 16.8 mm, which was 0.8 mm larger than the intendeddiameter of 16 mm. All cervical plates were placed in the correct midline orientation.

344 Singh et al.

Fig. 9. AP/lateral C-spine fluoroscopic image with image marker indicating midline of C5vertebral body. (Courtesy of Medtronic Sofamor Danek, Memphis, TN.)

FUTURE APPLICATIONS

Successful applications of image-guided spinal surgery have been reported to achievecomplex spinal procedures such as pedicle screw placement, transarticular C1–C2 fixa-tion, transoral odontoid resection, and anterior vertebral corpectomies. The futureundoubtedly will fuse the precision of computer assisted image technology with the cel-ebrated techniques of minimally invasive spinal applications. The greatest challenge forthis merger are the difficulties encountered with accurate percutaneous registration of thespine using standard endoscopy techniques. The registration process at this time requiresthe physical exposure of a part of the vertebra, whereas percutaneous endoscopic exposuresavoid unncessary soft-tissue dissections. Unfortunately, at this time, external landmarkscannot be reliably used as fiducials for registration because of tissue deformation.

A potential solution to this problem is the development of a percutaneously placedreference frame recently described by Assaker et al (30). They confirmed the efficacy ofthis strategy in a cadaveric feasibility study and then reported successful clinical resultsin two patients treated endoscopically with image-guided assistance.

Assaker et al. (30) utilized the Stealth computer-assisted image-guided system(Medtronic). The reference frame was designed to be inserted percutaneously into thepedicle of the affected vertebra under CT guidance with local sedation (Fig. 11). An

Frameless Stereotactic Imaging Techniques 345

Fig. 10. Intraoperative depiction of pedicle screw placement utilizing an image-guided CTnavigational system. (Courtesy of Medtronic Sofamor Danek, Memphis, TN.)

additional antirotational tube locked the pedicle implant to the spine, creating a singlerigid body for subsequent image acquisition.

The purpose of using an external reference frame in endoscopic surgery is twofold.First, the frame acts as a reference system (fiducial) necessary to assimilate the imagedata with the surgical space. Second, the frame is a dynamic 3D localizer. As mentionedpreviously, fiducials are necessary for registration. For endoscopic use, fiducials have tobe independent of the vertebrae. To resolve this problem, six points were chosen on theCT images of a frame (Figs. 12 and 13). The corresponding points were identified onthe frame and touched with the calibrated pointer for registration. This enabled a frame-based registration and navigation with image guidance. Two patients were subsequentlytreated using this computer-assisted endoscopic spinal technology (Fig. 14). Onepatient, a 70-yr-old male, underwent a thoracoscopic decompression of a calcified centraldisc herniation at T8-T9. The second patient, a 32-yr-old man, underwent an endo-scopic decompression of a sacroiliac osteophyte causing intermittent low-back pain.Both patients had postoperative CT scans that showed complete decompression of theoffending pathology, which resulted in complete resolution of symptoms.

The combination of endoscopic surgery and computer-assisted imaging should be arevolutionary advancement for both sciences in their quest to become more broadlyaccepted by the spine community. The benefits of endoscopic surgery, such as limitedsoft-tissue dissection, faster postoperative recovery period, and decreased hospitaliza-tion costs, may be safely balanced by the improved visualization afforded by framelessstereotaxy.

346 Singh et al.

Fig. 11. Patient positioned in lateral decubitus position for thoracoscopic surgical approach.The reference frame has been percutaneously placed under CT guidance into the pedicle of theT9 vertebra.

CONCLUSION

Despite the advances made in spinal imaging and computer-assisted spinal surgery,many limitations still exist impeding its widespread acceptance. This includes primarily itslack of user-friendly applications and its cost. The current cost of computer-guided systemsand their lack of documented efficacy at this time in terms of cost savings and reducedpatient morbidity limits their acceptance at many spine centers. However, dismissingcomputer image-guided surgery as too expensive to be practical is by far premature. Acost–benefit analysis should take into account the infancy of the technology and theenormous potential of the synergy of frameless stereotaxis and minimally invasivespinal techniques.

The technological limitations of image-assisted surgery are clearly evident. The lackof real-time image acquisition, the difficulties with optical apprecation of light-emittingdiode (LED) sensors, and the time-consuming nature of preoperative image acquisitionlimit the current application of frameless stereotaxy in general spinal applications.

In an elegantly performed study by Rampersaud et al. (31), a geometric model wasdeveloped relating spinal pedicle anatomy to accuracy requirements for image-guidedsurgery. The investigators noted a very small maximum permissible translational errorof <1 mm and rotational error of <5° at the midcervical spine, midthoracic spine, and

Frameless Stereotactic Imaging Techniques 347

Fig. 12. View of monitor during endoscopic navigation showing tip of pointer on preplannedtarget.

thoracolumbar junction (31). These accuracy requirements far exceeded the overallclinical utility error of current image-guided systems. This raises the question, Why doimage-guidance systems improve the clinical accuracy of pedicle screw placement?What may actually happen intraoperatively is that the image-guided system may bringthe surgeon much closer to the actual starting point and trajectory. Further refinementsof the screw path result from the surgeon’s appropriate response to visual and tactilefeedback, and the self-centering, mechanical constraint provided by the pedicle wall.Thus, an image-guided surgical system functions as a synergistic tool along with thesurgeon’s judgment of the intraoperative anatomy.

Accordingly, users of image-guided surgery systems must be cognizant of the limi-tations of these current systems. The progressive development of real-time imagingtechnology, coupled with refinements in minimally invasive spinal techniques, appearsto be very promising in improving patient outcomes and reducing the morbidity ofinvasive spinal procedures. In conclusion, image-guided surgical systems can aid thesurgeon in navigating complex spinal anatomy; however, these systems are only toolsthat must be combined with good surgical judgment and skill to achieve successfulsurgical outcomes.

348 Singh et al.

Fig. 13. CT image of geometry of frame used as a fiducial for registration. Paired pointmatching and surface mapping are systematically performed.

REFERENCES

1. Woodward E, Leon S, Moriarty T, et al. Initial experience with intraoperative magneticresonance imaging in spine surgery. Spine 2001;26(4):410–407.

2. Salehi SA, Ondra SL. Use of internal fiducial markers in frameless stereotactic navigationalsystems during spinal surgery: technical note. Neurosurgery 2000;47(6):1460–1462.

3. Albert T, Klein G, Vaccaro A. Image-guided anterior cervical corpectomy: a feasibilitystudy. Spine 1999;24:826–830.

4. Dyer P, Patel N, Pell G. The ISG viewing wand: an application to altanto-axial cervicalsurgery using the LeFort I maxillary osteotomy. Br J Oral Maxillofac Surg 1995;33:370–374.

5. Glossop ND, Hu RW, Randle JA. Computer-aided pedicle screw placement using frame-less stereotaxis. Spine 1996;21:2026–2034.

6. Kalfas I, Kormos D, Murphy M. Application of frameless stereotaxy to pedicle screw fixationof the spine. J Neurosurg 1995;83:641–647.

7. Kamimura M, Ebara S, Itoh H, et al. Accurate pedicle screw insertion under the control ofa computer assisted image guiding system: laboratory test and clinical study. J Orthop Sci1999;4:197–206.

8. Kato A, Yoshimine T, Hayakawa T. A frameless, armless navigational system for computer-assisted neurosurgery. J Neurosurg 1991;74:845–849.

9. Kramer DL, Ludwig SC, Balderston RA, Vaccaro AR, Foley KT, Albert TJ. Placement ofpedicle screws in the cervical spine: comparative accuracy of three techniques. Spine2000;25:1655–1667.

10. Foley K, Silveri C, Vaccaro AR, et al. Atlantoaxial transarticular screw fixation: risk assessmentand bone morphology using an image guidance system. J Bone Joint Sury (Br) 1998;80:245.

11. Foley K, Smith M. Image-guided spine surgery. Neurosurg Clin North Am 1996;7:171–186.12. Ludwig SC, Klein GR, Vaccaro AR, Rushton SA, Lazar RD, Albert TJ. The efficacy of using

an image-guided Kerrison in performing an anterior cervical foraminotomy: an anatomicanalysis of the spine. Spine 1999;24:1358–1362.

Frameless Stereotactic Imaging Techniques 349

Fig. 14. Schematic of typical operative setup for image-guided thoracoscopy.

13. Ludwig SC, Kramer DL, Balderston RA, Vaccaro AR, Foley KT, Albert TJ. Placement ofpedicle screws in the human cadaveric cervical spine: comparative accuracy of three tech-niques. Spine 2000;25(13):1665–1667.

14. Notle L, Slomczykowski M, Berlemann U, et al. A new approach to computer aided spinesurgery: fluoroscopy-based surgical navigation. Eur Spine J 2000;9(Suppl 1):S78–S88.

15. Pollack I, Welch W, Jacobs G. A frameless stereotactic guidance: an intraoperative adjunctin the transoral approach for ventral cervicomedullary junction decompression. Spine1995;20:216–220.

16. Heller JG, Carlson GD, Abitbol JJ, et al. Anatomic comparison of the Roy-Camille and Magerltechniques for screw placement in the lower cervical spine. Spine 1991;16:S552–S557.

17. Heller JG, Silcox H, Sutterlin C. Complications of posterior cervical plating. Spine1995;20(22):2442–2448.

18. Weinstein JN, Spratt KF, Spengler D, et al. Spinal pedicle fixation: reliability and validityof roentgenogram-based assessment and surgical factors on successful screw placement.Spine 1988;13:1012–1018.

19. Roessler K, Ungerboeck K, Dietrich E. Frameless stereotactic guided neurosurgery: clinicalexperience with an infrared based pointer device navigational system. Acta Neurochir1997;139:551–559.

20. Sandeman D, Gill S. The impact of interactive image guided surgery: the Bristol experiencewith the ISG/Elekta viewing wand. Acta Neurochir Suppl 1995;64:54–58.

21. Silveri CP, Foley KT, Vaccaro AR, Shah S, Garfin SR. Atlantoaxial transarticular screw fixa-tion: risk assessment and bone morphology using an image guidance system. Annual meeting,25th CSRS, Dec. 7–7, 1997, Rancho Mirage, CA.

22. Smith K, Frank K, Bucholz R. The neurostation––a highly accurate, minimally invasive solu-tion to frameless stereotactic neurosurgery. Comput Med Imaging Graph 1994;18:247–256.

23. Welch E, Sunach B, Pollack I, et al. Frameless stereotactic guidance for surgery of theupper cervical spine. Neurosurgery 1997;40:958–964.

24. Bolger C, Wigfield C. Frameless stereotaxy and anterior cervical surgery. Computer AidedSurg 1999;4:322–327.

25. Bolger C, Wigfield C. Image-guided surgery: applications to the cervical and thoracic spineand review of the first 120 procedures. J Neurosurg (Spine 2) 2000;92:175–180.

26. Richter M, Amiot LP, Neller S, Kluger P, Puhl W. Computer assisted surgery in posteriorinstrumentation of the cervical spine: an in-vitro feasibility study. Eur Spine J2000;(Suppl 1):S65–S70.

27. Laine T, Schlenzka D, Makitalo K, et al. Improved accuracy of pedicle screw insertionwith computer-assisted surgery: a prospective clinical trial of 30 patients. Spine1997;22:1254–1258.

28. Foley K, Rampersaud Y. Frameless spinal stereotaxis. Curr Orthop 1998;12:104–110.29. Foley K, Rampersaud YR, Simone DA, Jansen TH. Virtual fluoroscopy for cervical spine

surgery, annual meeting, 27th CSRS, Dec 16–18, 1999, Seattle, WA.30. Assaker R, Cinquin P, Cotton A, Lejeune J. Image-guided endoscopic spine surgery: part I

and II. Spine 2001;26(15):1705–1718.31. Rampersaud Y, Simon D, Foley KT. Accuracy requirements for image-guided spinal pedicle

screw placement. Spine 2001;26(4):352–359.

350 Singh et al.

18The Rise and Fall of Chemonucleolysis

James W. Simmons, Jr., MD and Robert D. Fraser, MD

INTRODUCTION

The past four decades have witnessed the arrival of numerous interventional proceduresfor the treatment of sciatica and back pain. Some have failed to gain general acceptanceand have simply faded away, and others continue to be used without evidence of effi-cacy. What is unique about chymopapain is that it has fallen out of favor despite con-vincing evidence of its utility and efficacy in the treatment of sciatica from lumbar discprotrusion. It has been shown to be superior to placebo in no less than three double-blind,randomized controlled trials (1,2), and to be cost-effective with low morbidity whencompared with open surgery (3).

This chapter describes the circ*mstances and events leading to the rise and fall ofchymopapain, exploring how commercial interests and emotional influences have over-ridden substantial scientific evidence of efficacy and safety (4–7). Much of this accountdescribes information that has already been published extensively, but some of itemphasizes details not well known but considered important in the overall context,particularly from a historical perspective.

The attraction for the use of chymopapain as a proteolytic agent for the treatment ofherniated nucleus pulposus (NP), is readily understandable and its introduction and riseto prominence are covered in some detail in this chapter. On the other hand, the factorsleading to the demise of chemonucleolysis are less well documented in the literaturebut are equally apparent in light of events to be described.

The use of chymopapain B in the treatment of problems related to the disc is probablyone of the most critically appraised invasive procedures involving a pharmaceuticaldrug that we as clinicians have dealt with. However, in spite of extensive evidence thatsupported safety and efficacy leading to worldwide use, the product suffered a demisethat occurred without a satisfactory official explanation. What follows is an account ofthe history of chymopapain followed by an assessment of the factors leading to itsremoval from clinical use.

Eugene Jansen and Arnold Balls isolated chymopapain in 1941 from the crude latexderived from the fruit of carica papaya by “milking” the green papaya fruit while on theplant prior to harvest (8).

351

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

Lewis Thomas was seeking an enzyme that would reduce circulating protein inblood that clogged the renal tubules resulting in renal shutdown, as enzymatic activitycannot be produced by known chemical methods and has to be obtained by way of nat-urally occurring biologicals. In 1956, he injected rabbits intravenously with crudepapain and noticed that their ears drooped. Forty-eight hours later, the rabbits’earsresumed their erect posture. This indicated a reversible action on the chondral intracel-lular substance of the ears. It was also noted that the trachea softened, but there was noeffect on any other tissue in the rabbit. It was particularly of interest that

apart from the unusual cosmetic effect, the animals exhibited no evidence of systemic illnessor discomfort. The ears had replenished their basophilic chondroid matrix allowing themto regain their original shape. Larger doses of papain were injected and had an impact onjoint cartilage, epiphyseal growth plates, and tracheal and bronchial cartilage; however, nosystemic problems were found to be present (9).

Lyman Smith first thought that could papain might be of value in treating chondrob-lastic tumors. Although this did not prove worthy, he found that intradiscal injections inrabbits removed the NP, leaving the annulus largely intact (10).

Prior to this, Carl Hirsch had the concept of injecting a specific enzyme into theintervertebral disc but not specifically for the treatment of disc herniation (11). He rea-soned that an intradiscal injection of a chondrolytic enzyme would cause the disc tobecome stable and asymptomatic by accelerating the process of disc degeneration. Hehad envisioned an enzyme similar to that produced by bacteria as is seen in infectiousprocesses. Other proteolytic enzymes were investigated and found to have a similareffect on the disc tissue, but toxicological studies revealed chymopapain to be the leasttoxic and to have the most specific action on the mucopolysaccharide of the intervertebraldisc. Other proteolytic enzymes such as collagenase were advocated (12). Extensiveresearch with collagenases proved it not to be “safe and effective.”

Subsequently, Smith sold the patent for chymopapain to Baxter-Travenol for $1. Thecompany then formulated a product called Discase, which was a combination of chy-mopapain B, cysteine sodium sulfite, and EDTA in lyophilized form. During the periodof the first phase of investigation, it was used in 10 patients in Switzerland in 1963, andby 1975, 35 of us as investigators in a phase 3 trial had injected approx 17,000 patients.

A controversial study done at Walter Reed Army Medical Center in 1975 triggeredthe withdrawal of the New Drug Application that had been filed with the Food andDrug Administration (FDA) for the use and treatment of intervertebral disc disease withDiscase (13). The study reported no statistical difference in instance or quality ofimprovement between the placebo group (29% success rate) and the group treated withDiscase (58% success rate). Baxter-Travenol voluntarily withdrew the drug, rather thantake the chance of rejection by the FDA. Although Brown and Daroff (14) criticized theWalter Reed Army Medical Center study because of the early code break, the lack ofinert placebo, the insufficient dose of Discase, and the lack of technical experience, thedrug continued to be unavailable for use in indicated patients. Physicians in the UnitedStates who had been using the drug with excellent results were disheartened that no effortappeared to be being made toward an FDA approval. Investigational use continuedthroughout most of the world, particularly in Australia and in England. Yugoslavia had pro-duced a product (Lekopain) that was basically a chymopapain B and was widely used inthe Eastern bloc countries and, to a lesser extent, in France and Italy with favorable results.

352 Simmons and Fraser

On October 15, 1975, Baxter-Travenol withdrew its new drug application for chymopa-pain from the FDA, effectively removing the drug from use in the United States(15). Patients desiring chemonucleolysis with chymopapain were forced to seek treat-ment in Canada, where the drug was commercially available. In early 1977, a coalitionof American physicians interested in making chymopapain available to patients wasformed and chartered the Committee Advocating the Development and Use of Chymopa-pain to Eliminate Unnecessary Surgery (CADUCEUS). This group pursued legislativeapproaches to reactivate the withdrawn new drug application and even considered sub-mitting its own application. For several years, the group worked diligently to gainnational approval. However, in the late 1970s, other groups seeing little progress in thenational trust began to explore state legislative approval to allow the production and useof chymopapain at the suggestions of CADUCEUS. Illinois, Indiana, and Texas all hadlegal vehicles to allow such interstate use, but of these states only Texas had the envi-ronment that would support growth of the raw materials. This was of critical importance,because according to FDA regulations all materials had to be available within the stateand not cross-state boundaries. Subsequently, the State of Texas was granted approvalfor local manufacture of chymopapain preparation under the Food, Drug & CosmeticAct. The product Chemolase produced in Texas was used from January 1980 untilapproval by the FDA of Chymodiactin® in 1982. Although the Texas group attemptedto get FDA approval of Chemolase, it was apparent that only Chymodiactin would beapproved. Without FDA approval for national distribution, the production of Chemolasewas discontinued.

During the use of Chemolase (16,17), orthopedic surgeons and neurosurgeons and919 patients participated in an open-label uncontrolled clinical trial involving the useof chymopapain (Chemolase) in Texas from 1980 to 1982. Patients admitted to thestudy had persistent low-back pain or sciatica owing to protrusion, extrusion, ordegeneration of the lumbar intervertebral disc that was not responsive to conservativemanagement. Although the study was not specifically designated to assess safety andefficacy, it did allow retrospective analysis of these entities following intradiscaladministration of chymopapain. Patients were considered candidates for the study if amyelogram or computed tomography scan suggested lumbar disc disease and specificclinical signs of lumbar disc disease were present. The dose of chymopapain most fre-quently used ranged from 3000 to 4000 U/disc. Responses to chemonucleolysis wereassessed at 1, 3, or 6 mo after injection, depending on patient and physician adherenceto the protocol. Patients’responses were assessed by point reduction systems thatwere deducted from an initial patient score of 10, based on the presence of variouslevels of discomfort or limitations of daily activities. Based on the point scores,patients’responses were categorized as poor, fair, good, or excellent. Fair, good, andexcellent responses were considered treatment successes. Of the 919 patients whounderwent chemonucleolysis with chemolase (chymopapain B), 408 were evaluated1, 3, or 6 mo after injection by a physician. An independent biostatistician reviewedall of the data and performed the statistical analysis. Fifty-five percent of patientsreceived injections in a single intervertebral disc, and the remainder in two to fourdiscs. Success rates were 93% 1 mo after injection, 92% at 3 mo, and 93% at 6 mo.An unusual finding was observed when the effect of different variables on treatmentefficacy was analyzed. Significantly lower response rates were found for Hispanics,

Rise and Fall of Chemonucleolysis 353

blue-collar workers, and patients covered by workers’compensation insurance inthe categories of race, occupation, and type of insurance coverage, respectively. In theother 511 patients, either the patient did not report for follow-up examinations or thefollow-up evaluation was not recorded appropriately for evaluation of efficacy. Forthe 919 patients who received chymopapain by intradiscal injection, 70 adverse reac-tions were noted in 46 patients (5.0%). Erythema was the most common side effect,occurring in 1.8% of patients. The most serious reaction was anaphylaxis, whichoccurred in 1.1% of patients; however, based on the individual physician assessment,severe anaphylactic reactions were reported in only 0.54% of patients. All patientswere managed medically without lasting effects. Giant urticaria, hypotension, andparaspinal muscle spasm occurred at similar frequency. Back pain was reported inonly 0.4% of the treated patients. No deaths occurred. Although more sophisticatedstudies have since been done, the Texas study provides additional support for the safetyand efficacy of chymopapain chemonucleolysis in the treatment of low-back pain andsciatica of discogenic origin that do not respond to more conservative management.

In 1979, Baxter-Travenol began a blinded study of chymopapain vs cysteine-edetate-iothalamate (CEI) or saline control, which allowed patients to have a laminectomy inthe United States or go to Canada as a control. Patients during that period had achoice of having a laminectomy in the United States, going to Canada for chemonu-cleolysis, or going to Canada for chemonucleolysis if they had a placebo failure. Itwas apparent that significant difficulty was encountered in accumulating adequatenumbers of trial participants. The results were finally published in 1998. One hundredseventy three participants at 25 locations reported 71% success with Discase comparedwith 45% with CEI (13). More sophisticated studies were being done in Australia.Immunological, vascular, and neurological complications as well as discitis have allbeen discussed scientifically in detail, along with the risk of mortality (18).

Was the fall from favor of chymopapain simply the result of adverse reports aboutcomplications, even though as reported in this chapter such reports were not well founded?If the answer to this question is “yes” it is a sad misconception on the part of the treatingphysicians as well as the patient population. Could it have been an administrative decision?We know that the FDA license to produce and sell Chymodiactin went from Smith toBaxter-Travenol and was separated off later into a company called Omnis. Boots Pharma-ceutical acquired Omnis, thereby obtaining the license from Baxter-Travenol, and the nextstep in this procession was for Knoll Pharmaceuticals to purchase the license from Boots(19). The license to manufacture and sell chymodiactin was subsequently sold to Abbottalong with other pharmaceutical products. It was Abbott’decision, in about 1999, to dis-continue the manufacture of Chymodiactin, thereby making it unavailable for use.

The official explanation for the discontinuation of the manufacture of Chymodiactinhas yet to be obtained. However, the factors influencing the downfall can be summa-rized as follows:

• Income: Chemonucleolysis with chymopapain competed with discectomy, a surgical proce-dure that was the main source of income for many surgeons. This led, unfortunately, tobiased and unfavorable comments being made by those with a vested interest in its demise.

• Inappropriate patient selection: It is apparent that chymopapain was being used to treatpatients who were not good candidates for discectomy. To quote Ian Macnab (personalcommunication, 1977), a leading authority on the treatment of lumbar disc disease in the

354 Simmons and Fraser

1970s, “If a surgeon cannot get good results with chymopapain he should not be operatingon the spine!”

• Poor technique leading to complications: In 1984, during the first few months followingthe release of Chymodiactin in the United States, approx 8000 orthopedic surgeons andneurosurgeons attended 1-d training courses. It is apparent that this alone was inadequatetraining for a percutaneous interventional technique that demanded precision for its safetyand efficacy.

• Fear of litigation: This was generated to a large degree by the much-publicized complicationsthat resulted in part from poor training in technique.

• Competition from the introduction of the automated nucleotome: This is a device thatallowed removal of disc tissue; hence, the procedure achieved greater initial acceptance bysurgeons.

• Change in attitude to early rehabilitation following disc surgery: The practice of earlyambulation and early hospital discharge after disc surgery, first introduced in the early1990s, reduced the advantage in cost-effectiveness of chemonucleolysis.

• Use of targeted epidural steroids: Posterior epidural steroids were of lesser therapeuticvalue in the treatment of lumbar disc prolapse and did not greatly compete with chemonu-cleolysis. The increased success of foraminal epidural steroids (20), whereby the solutionwas delivered to the affected root canal with image intensifier guidance, greatly lessened thenumber of patients being considered for chemonucleolysis.

In all probability these factors combined to reduce dramatically the use of Chymodi-actin to the point where the manufacturer made a commercial decision to remove theproduct from the market.

Dr. Lyman Smith addressed most of the questions and doubts regarding the use,safety, and efficacy in the following three personal communications:

In April 1987, a report on chemonucleolysis was aired on the ABC program 20/20.Timothy Johnson, MD, ABC News medical editor, conducted this report, and the follow-ing is quoted from a letter written to Dr. Johnson by Lyman Smith, MD, dated August27, 1990, in response to his report.

Chemonucleolysis is widely used in Europe. The explanation for this is that surgeons inEurope are commonly on salaries and contingency fees for plaintiff lawyers are illegal.Clearly, the health care crisis in this country today has reached extraordinary proportions.We cannot afford to overlook effective alternatives to dangerous and costly invasivesurgery.

Your report on Chemonucleolysis, which aired on the “20/20” program, in April 1987,condemned thousands of individuals suffering from disc disease in the United States tounnecessary major surgery. Due to your blasted and inaccurate report, very little of whichwas substantiated by facts, use of Chymopapain fell precipitously (40% in the two monthsfollowing the program!).

The result was that either more invasive laminectomy or the percutaneous automateddiscectomy procedure, touted on the program (and since proven a therapeutic failure),were utilized far more often than the less dangerous and less costly Chemonucleolysis.

Further, your reference to Chemonucleolysis causing acute transverse myelitis (ATM) hasbeen proven to be unfounded. The appearance of these reported six cases of alleged acutetransverse myelitis was devastating. As with other severe conditions following any procedurein the United States, the patient records were sealed and access to them denied. Rumorswere the natural consequences. Symptoms in all but one of the cases appeared two to threeweeks after apparently successful treatments with Chymopapain. In that one case, symptomsappeared shortly after treatment and got worse as time went on. All had flaccid paralysis andnone developed signs of spastic paralysis, a characteristic of transverse myelitis. At the time

Rise and Fall of Chemonucleolysis 355

of diagnosis, one patient had a suggestive history of multiple sclerosis and another laterdeveloped clear evidence of this disease. One had a history of viral infection before treatmentand another was diabetic. The other, Mr. Devletsah, who appeared on “20/20”, only devel-oped paralysis after laminectomy following Chemonucleolysis. This, the causes of paralysisin these six patients are far from certain. Some could have been due to inept injections andothers to causes other than the enzyme or the procedure. Finally, it may interest you to knowthe ATM has been reported in Australia after intradiscal injection of saline and another, in thiscountry, following removal of an intramedullary nail. Other causative agents were sought inthese and other comparable cases. Why should this not have been done in the case of a usefuland effective therapeutic agent such as Chymopapain?

Statistics show that Chemonucleolysis is a valid and effective alternative to surgery. Iwould be delighted to review these data with you. These findings are certainly newswor-thy and something your audience will be interested in.”

On August 27, 1990, a telephone conversation was held between Dr. Lyman Smith andDr. Helene M. Cole, senior editor of JAMA. After their conversation, Dr. Smith sent Dr.Cole his rebuttal to the Diagnostic and Therapeutic Technology Assessment (D.A.T.T.A.)(21) on chemonucleolysis, published in Clinical Orthopaedics and Related Research.

D.A.T.T.A. Statement #3: “Even small intrathecal leaks of Chymopapain have a potentialfor damage to the central nervous system.”

REBUTTAL: Chymopapain will cause bleeding from the capillaries of the pia arach-noid, which are not protected by collagen, as are larger blood vessels. Large doses willcause enough bleeding to raise C.S.F. pressure to unacceptable levels and the animal willdie. The toxic effect is purely a pressure phenomenon. If a spinal tap is done early torelieve the pressure, the animal will survive without a following neurological deficit.Small amounts do not damage the central nervous system. The L.D. 50 in Rhesus mon-keys, for example, is 1000 units per kilogram. Fifty units are therapeutic. (22)

The following is a quote from a letter dated September 4, 1990, to Prof. Alf Nachemson,Department of Orthopaedics, Gothenburg University, Sweden, from Dr. Lyman Smith, inresponse to statements made by Prof. Nachemson during a symposium published in Con-temporary Orthopaedics (23).

I have just been saddled with a sever diagnosis leaving a poor prognosis; so I wish to clearthe air with you. You are basically a great scientist, however, some of your biased statementson Chemonucleolysis throughout the years have bordered on the ridiculous.

In 1961, I began work with Baxter Laboratories on the ceramics and after awhile, I toldthe Director of Research about the rabbits. He was interested and we formed a researchteam––the million-dollar question to be pursued: “Was papain the best enzyme and was itsafe?” You know the rest of the story!

In the British Journal of Bone and Joint Surgery, February 1959, Carl Hirsch theorizedas follows: “Sooner or later a substance may be found by which a degenerated disc could betransformed. It might be possible to create a chemolytic enzyme that, injected into a disc,would cause a connective tissue reaction.” As you know, Chymopapain does not act in thatsclerosing fashion. Carl came to my hospital in Elgin on November 6, 1968, with JorgeGalante and we examined some patients and viewed the x-rays of the first 80 patients I hadinjected. The only other investigators at the time were Lee Ford and Leon Wiltse. Carlseemed favourably impressed; it’s a shame that impression did not wash off upon you!

You continue to quote your poor results with Chymopapain obtained with partiallyinactivated DISCASE, “Surgery versus Chemonucleolysis for Herniated Lumbar Discs,”Clinical Orthopaedics, No. 174, April 1983. I pointed out to you then that the Chymopa-pain had been shipped to you from the United States without being refrigerated. DIS-CASE, unlike the present version, CHYMODIACTIN, was sensitive to heat. Mulholland

356 Simmons and Fraser

of Nottingham received the same shipment and complained that laminectomies on thefailed patients showed little or no action of the enzyme. Active enzyme would have led toan accurate study!

Chemonucleolysis is neater, quicker, safer, cheaper, and if performed on an idealpatient, as effective as any invasive mechanical procedure. Prove me wrong!”

A citizens’petition was made to the FDA February 8, 2002, to determine whether thedrug was withdrawn from sale for reasons of safety and effectiveness. A response to thecitizens’petition was received January 27, 2003, stating that

the FDA has reviewed its records and has determined that Chymopapain 10,000 units/vialinjections (Chymodiactin), NDA 18-663, was not withdrawn from sale for reasons ofsafety or effectiveness. FDA will maintain Chymopapain 10,000 units/vial injections in the“discontinued drug product list” of approved drug products with therapeutic equivalenceevaluations (the orange book). (24)

Whatever the reasons for discontinuation of the manufacture of Chymodiactin, itwas not for safety and effectiveness. “The results of Chemonucleolysis depend notonly on the enzyme, but more importantly on the diagnostic acumen of clinician. Ifyou can’t get a good result from Chemonucleolysis you shouldn’t really be allowed tooperate on backs.” Chymopapain continues to be unavailable for use in the UnitedStates (25).

REFERENCES

1. Fraser RD. Chymopapain for the treatment of intervertebral disc herniation: a preliminaryreport of a double-blind study. Spine 1982;7:608–612.

2. Javid MD, Norby EJ, Ford LT, et al. Safety and efficacy of chymopapain (Chymodiactin®)in herniated nucleus pulposus with sciatica: results of a randomized, double-blind study.JAMA 1983;249:2489–2494.

3. Nordby EJ, Wright PH, Schofield SR. Safety of chemonucleolysis: adverse effects reportedin the USA 1982 to 1989. Clin Orthop 1993;(293):122–134.

4. Bradford DS, Oegema TR Jr, Cooper M, et al. Chymopapain, chemonucleolysis andnucleus pulposus regeneration: biological and biochemical study. Spine 1984;9:135–147.

5. Brown M, Daroff RB. The double-blind study comparing disease to placebo: an editorialcomment Spine 1977;2:33.

6. Darakjian HE, Wiltse LL. Low-dose chymopapain as a safe alternative to laminectomy,paper presented at the 3rd annual meeting of the International Intradiscal Therapy Society(IITS), Marbella, Spain. March 1990;7–11.

7. Garvin PG, Jennings RB, Smith L, Gesler RM. Chymopapain: a pharmacologic and toxico-logic evaluation in experimental animals. Clin Orthop 1965;41:204–223.

8. Jansen EF, Balls AK. Chymopapain: a new crystalline proteinase from papaya latex. J BiolChem 1941;137:459–460.

9. Thomas LB. Reversible collapse of rabbit ears after intravenous papain, and prevention ofrecovery by cortisone. J Exp Med 1956;104:245–251.

10. Smith L, Garvin PJ, Jennings RB, Gesler RM. Enzyme dissolution of the nucleus pulposus.Nature 1963;198:1311–1312.

11. Hirsch C. Studies on the pathology of low back pain. J Bone Joint Surg 1959;41B:237–243.12. Brown M. Intradiscal Therapy: Chymopapain or Collagenase. Year Book Medical, Chicago,

1983, pp. 120–127.13. Schwetschenau PR, Ramirez A, Johnston J, Wiggs C, Martins AN. Double-blind evaluation

of intradiscal chymopapain for herniated lumbar discs. J Neurosurg 1976;45:622–627.

Rise and Fall of Chemonucleolysis 357

14. Brown MD, Daroff RB. The double blind study comparing discase to placebo: an editorialcomment. Spine 1977;2:233–236.

15. FDA Drug Bull 1984;14(2).16. Simmons JW, Stavinoha WB, Knodel LC. Update and review of chemonucleolysis. Clin

Orthop 1964;183:51–60.17. Simmons JW, Upman PJ, Stavinoha WB. Pharmacologic and toxicologic profile of chy-

mopapain B (chemolase). Drug Chem Toxicol 1984;7:299–314.18. Nordby EJ, Fraser RD. Chemonucleolysis, in The Adult Spine, Principals and Practices,

2nd ed, 1997.19. Data on file, Boots Pharmaceuticals, Lincolnshire, IL.20. Weiner B, Fraser RD. Foraminal injection for lateral lumbar disc herniation. J Bone Joint

Surg (Br) 1997;79-B:804–807.21. Cole HM. A Diagnostic and Therapeutic Technology Assessment (D.A.T.T.A). JAMA

1989;262(7).22. Gesler RM. Pharmacologic properties of chymopapain. CORR 1969;67:47–51.23. Mooney V, Brown A, Nachemson JA, McCulloch, Schroeder FW, Simmons JW. Symposium:

chemonucleolysis in the management of herniated lumbar disc. Contemp Orthop.1984;9(5):97–124.

24. Simmons JW Jr. Citizen petition (docket no. 02P-0068/CP1). Federal Register Vol. 68, No.17, Monday, January 27, 2003, Notices.

25. Macnab I. Chemonucleolysis Seminar, Montreal, Quebec, November 1977.

358 Simmons and Fraser

19Lumbar Microendoscopic Discectomy

Trent L. Tredway, MD and Richard G. Fessler, MD, PhD

INTRODUCTION

Low-back pain is one of the most common complaints that lead patients to seekmedical attention. It is estimated that there are nearly 15 million physician office visitseach year for low-back pain (1). Additionally, back pain causes nearly 80% of work-ers to miss at least 8 wk of work following a back injury (2). Estimates of lifetimeprevalence and annual incidence has been reported to be 60–90 and 5%, respectively(3). With this prevalence, it has been estimated that the total health care expendi-tures incurred by individuals with back pain in the United States reached $90.7 bil-lion. The total incremental expenditures attributable to back pain among theseindividuals are estimated to be approx $23.6 billion. On average, individuals withback pain encumbered health care expenditures about 60% higher than individualswithout back pain (4).

Although many individuals present with the common complaint of low-back pain,the etiology is often multifactorial. One cause of back pain is related to soft-tissueinjury secondary to back strain. This type of injury involves musculoskeletal and liga-mentous injury, is usually self-limiting, and resolves with nonoperative management.Other causes of back pain may include varying degrees of arthritic changes of the lum-bosacral spine; compression fractures secondary to trauma or osteoporosis; and, lesscommonly, metastases. However, one of the more common causes of back pain isrelated to the changes that occur within the intervertebral disc. Herniated discs, bulgingdiscs, ruptured discs, and radial tears in the annulus have all been used to describe thepathological alterations of the intervertebral disc.

Vesalius (5) first described anatomically the intervertebral disc in 1555. Since then,the intervertebral disc has been eloquently studied and determined to be composed ofan outer annulus consisting of dense connective tissue and an inner nucleus pulposus(NP) that is largely composed of collagen and glycosaminoglycans (6–8). This func-tional design contributes to the dynamic properties of the disc that permit flexion,extension, lateral bending, and rotation to occur along the spinal axis. However, as apatient ages, the NP, a remnant of the notochord, gradually loses its water content as wellas its elastic properties, leading to degeneration of the functional disc unit. These changesmay best be exemplified on magnetic resonance imaging (MRI), with the degenerativedisc appearing dark compared with nondegenerative discs (Fig. 1).

359

From: Arthroscopic and Endoscopic Spinal Surgery: Text and Atlas: Second EditionEdited by: P. Kambin © Humana Press Inc., Totowa, NJ

CLINICAL PRESENTATION

Patients with herniated disc fragments often present with radicular pain, back pain,or both (Fig. 2). The exact nature of the pain associated with a lumbar herniated disc iscontroversial. It is thought that direct compression of the neural elements at or near thelevel of the foramen contributes to the radicular pattern of pain often experienced.Therefore, removal of the offending fragment and inspection of the neural foramen toensure decompression may alleviate the radicular pain. However, patients may presentwith tremendous radicular pain and radiological imaging that does not substantiate acompressive syndrome. It has been surmised that the composition of the disc itself mayinduce an inflammatory response after herniation. The release of local substancesderived from the inflammatory response may be perceived as a painful stimulus. Forexample, increased levels of glutamate are purported to be involved in the perception ofpain secondary to intervertebral disc herniation (9).

360 Tredway and Fessler

Fig. 1. Sagittal T2WI MRI of patient with degenerative disc disease at L5/S1 level. The signalintensity is decreased compared with the normal levels.

Compressive syndromes secondary to disc herniation without associated pain mayoccur. Rarely, a patient may present with an acute neurological deficit that isattributable to neural compression secondary to disc herniation. This may best be exem-plified in patients with L4 or L5 nerve root involvement that notice weakness in dorsi-flexion (tibialis anterior) and extension of the great toe (extensor hallicus longus),respectively. The neurological deficit may be so pronounced on presentation that thepatient may demonstrate a “foot drop” on examination. By contrast, some patients maypresent only with sensory loss along a dermatomal pattern. Finally, patients with large,centrally located lumbar disc herniations may present with cauda equina syndrome. Inthis syndrome, bowel, bladder, and even sexual dysfunction can occur and should beconsidered a neurosurgical urgency. In review of patients presenting with a caudaequina syndrome secondary to a herniated disc, patients undergoing decompressionearly in the time course (less than 24 h after onset of bladder paralysis) have an excellent

Lumber Microendoscopic Discectomy 361

Fig. 2. (A) Sagittal T2WI MRI of patient with herniated NP at L4/L5 disc level; (B) axialT2WI MRI of same patient with herniated fragment located to right of midline and causing com-pression of right exiting nerve root.

chance of recovery compared with patients undergoing surgery after 24 h of symptoms(10). Therefore, a careful history and physical examination should always be performedat the earliest possible time in order to improve functional outcome.

PATIENT EVALUATION

Examination of a patient presenting with symptoms relating to a herniated lumbardisc should include a complete history focusing on the exact location and pattern of thepain; onset and duration of the complaint, factors that aggravate or relieve the symp-toms; and, of course, any changes in bowel, bladder, or sexual function. Many of theseindividuals will report coinciding activities including lifting, straining, twisting, or bend-ing at the time of the acute onset. Other risk factors for disc herniation include smoking;having a sedentary occupation; driving a motor vehicle for prolonged periods; and operat-ing heavy, vibrating machinery (11–14).

After a history of the symptoms is elicited, a complete physical examination is per-formed focusing on the area involved. A thorough neurological examination includestesting the strength and tone of isolated muscle groups. The affected side is comparedwith the contralateral muscle groups and any existing discrepancies are carefully docu-mented. A thorough sensory examination including testing all modalities including lighttouch, two-point discrimination, pinprick, and vibratory sensation is also performed.With this organized approach to sensory evaluation, differences in a dermatomal patternmay be observed. In addition, deep tendon reflex testing may also provide the physicianwith information concerning which nerve root or roots are involved. Furthermore, thestraight leg raise test (also known as Lasègue’s sign) should always be performed inpatients suspected of having a radiculopathy secondary to an intervertebral disc hernia-tion. This maneuver, which comprises raising the supine patient’s straightened leg withpain elicited at <60°, is present in approx 83% of patients with nerve root compressionsecondary to disc herniation (15). Finally, a thorough neurological examination is not

362 Tredway and Fessler

Fig. 2. (Continued)

complete unless evaluation of the perineal region is conducted. Sensory examination ofthe sacral and genital regions may demonstrate a cauda equina syndrome. If any neuro-logical deficit exists, correlation with radiographic imaging is recommended.

In general, radiographs of the lumbar spine are obtained and the osseous anatomy isevaluated. Fractures, scoliotic changes, congenital abnormalities, degenerativechanges, as well as pathological processes may be revealed with careful review of theimages. The etiology of the complaint may be evident on these routine radiographs. Ifthere is no evidence of pathology on the radiographs of the lumbar spine, further evalu-ation may be necessary. An MRI of the lumbar spine, if possible, is the imaging modal-ity of choice. This will enable the physician to evaluate the intervertebral discs as wellas demonstrate other pathological features of the spine. The only limitation to MRI isthat it will not clearly demonstrate osseous pathology. If significant osseous pathologyis suspected, a computed tomography (CT) examination of the lumbar spine may provemore diagnostic. However, the findings on imaging should be cautiously interpretedbecause many asymptomatic individuals will have an abnormality on routine MRI (16).For this reason, the results of the MRI should be correlated with a thorough history andphysical examination. If the symptoms and radiographic results are congruent, thentreatment outcome results are improved (17).

TREATMENT OPTIONS

We routinely recommend conservative, nonoperative treatment for at least 6–8 wkafter initial onset of symptoms secondary to an intervertebral disc herniation. Therapygenerally consists of the use of nonsteroidal anti-inflammatory drugs, muscle relaxants,physical therapy, and occasionally epidural steroid injections for acute exacerbations. Ithas been reported that approx 85% of patients with acute disc herniation will improvewithout surgical intervention in an average of 6 wk (18). Therefore, only patients thatdo not respond to a trial of nonoperative therapy are considered for surgical interven-tion. However, if a neurological deficit is present, as may be observed in a patient witha foot drop, early intervention may be entertained. Once again, if a cauda equina syn-drome exists, urgent surgical intervention is recommended.

Virchow noted pathological changes related to the intervertebral disc when hedescribed traumatic disc disease in 1857; however, surgical intervention for treatmentof degenerative disc disease did not occur until the early twentieth century. Oppenheim(19) first described a posterior lumbar surgical approach for removal of an “enchon-droma” in 1909, and 4 yr later, Elsberg (20) described a laminectomy for removal of a“chondroma.” Dandy (21) also described removal of a “disc,” which he described asmimicking a tumor. However, it was not until 1934, when Mixter and Barr (22) com-piled numerous case reports from the literature, that a relationship between disc hernia-tion and sciatica was finally established. Since then, the surgical treatment ofdegenerative disc disease has been well established as well as controversial.

Traditional discectomies involved performing a laminectomy and removal of theoffending disc fragment through an intradural approach. Later, through improvements insurgical technique, the discectomy procedure evolved into a hemilaminotomy andmicrodiscectomy after Caspar (23) and Yasargil (24) introduced the operating microscopeinto the spine surgeon’s armamentarium. This technological advancement allowed sur-geons to perform arguably a more complete discectomy with better visualization and set

Lumber Microendoscopic Discectomy 363

the “gold standard” for clinical outcomes in which other surgical interventions for discdisease are compared.

Although lumbar microdiscectomy is the most common surgical intervention for thetreatment of lumbar herniated discs, a variety of treatment options have been performedas an alternative. One alternative treatment for herniated lumbar discs is chemonucleol-ysis. This procedure, first performed by Chicago orthopedic surgeon Lyman Smith,consists of percutaneously injecting chymopapain into the symptomatic intervertebraldisc. The enzymatic activity of the chymopapain hydrolyzes the mucoprotein of theNP (25). More than 75,000 patients underwent chemonucleolysis treatment in a 12-moperiod after Chymodiactin® (Boots Pharmaceuticals, Nottingham, UK), a new formulationof chymopapain, was introduced in 1983 (26). Despite its safety performance in clinicaltrials, complications occurred, including fatal anaphylactic reactions, cauda equina syn-dromes, and acute transverse myelitis, from inadvertent intradural injections (27,28).With these reports, the use of Chymodiactin decreased and is no longer manufacturedor distributed in the United States.

Another alternative to surgical intervention for disc disease is intradiscal electrother-mal coagulation (IDET). This therapeutic option is designed to treat internal disc dis-ruption in which the disc is painful secondary to the development of radial fissuresextending into the outer region of the annulus fibrosus. IDET consists of introducing aheating electrode into the NP percutaneously. A recent study has demonstrated thatIDET may provide relief in a small subset of patients with intractable back pain; how-ever, the results cannot be generalized to patients who did not fit the strict inclusion cri-teria for the study (29). Furthermore, IDET is not recommended for large,free-fragment disc herniations observed on radiographic imaging.

Despite the popularity, low morbidity, and clinical outcomes of microsurgical discec-tomy, many surgeons continued to develop innovative ways to treat lumbar herniateddiscs. In 1975, Hijikata et al. (30) introduced the “percutaneous nucleotomy,” whichwas performed through a cannula placed percutaneously using fluoroscopic guidance.Similarly, Kambin and Gellman (31) reported their technique of percutaneous posterolat-eral discectomy via a 6.4-mm–outer diameter cannula. Kambin advocated intraoperativevisualization and introduced the concept of arthroscopic microdiscectomy. Furthermore,he outlined the advantages of posterolateral access (32) (see Chapters 1 and 4).

Smith and Foley were instrumental in the development of the microendoscopic dis-cectomy (MED) procedure, in which specially designed microendoscopic instrumentsare used through a working channel with a 30° endoscope attached to a MetRx™ tubu-lar retractor system (Medtronics, Memphis, TN). The purported advantages of this tech-nique include less iatrogenic soft-tissue injury, improved visualization, less blood loss,less postoperative pain, and earlier hospital discharge (33,34). This surgical procedureis technically challenging at first, with a steep learning curve. However, after experi-ence and mastery, this technique is an extremely attractive alternative to traditionallumbar microdiscectomy.

THE MED PROCEDURE

Anesthesia

General endotracheal intubation with adequate intravenous access is obtained prior topositioning for an MED procedure. An induction dose of Anectine® (succinylcholine

364 Tredway and Fessler

chloride; GlaxoSmithKline, Research Triangle Park, NC) is used for intubation purposes;however, no other paralytic agent, muscle relaxant, or nitrous oxide is administeredthroughout the rest of the procedure. This enables the surgeon to have direct feedback if thenerve root is irritated during the procedure. We do not routinely employ the use ofintraoperative somatosensory evoked potentials and electromyography during MED becausevisualization of the neural elements with the 30° endoscope is excellent. Finally, because aroutine MED procedure lasts approx 45–90 min, we do not place a Foley catheter.

Positioning

The MED procedure is performed with the patient in the prone position and utilizes aWilson frame (O.S.I., Union City, CA) on a Jackson table (O.S.I.). This setup will allowa fluoroscopic unit (C-arm) to be draped into the field and is less cumbersome duringthe surgery (Fig. 3). The arms are positioned with the elbows at right angles to lessentraction on the brachial plexus. Care is taken to ensure that all osseous prominences arewell padded, especially with respect to the ulnar and peroneal nerves. Accordingly, pil-lows are placed under the lower extremities to reduce traction on the sciatic nerves.Sequential compression devices are also placed on the lower extremities to reduce theincidence of deep venous thromboses. Finally, the endotracheal tube is checked afterpositioning to ensure that there is no compression or obstruction.

Surgical Approach

The use of fluoroscopic guidance is pertinent in the preoperative setup. An indeliblemarker is used to mark the skin after a lateral radiograph has identified the correct sur-

Lumber Microendoscopic Discectomy 365

Fig. 3. Operative room setup for MED procedure. The patient is prone on a Wilson frame andJackson table. The video tower and fluoroscopic unit is placed on the opposite side of the surgicalapproach and surgeon.

gical level. The incision is approx 1.5 cm paramedian to the symptomatic side. This dis-tance will enable access to the intervertebral disc space and nerve root.

The skin is cleansed with a three-step technique. We routinely employ a povidine/iodinescrub, a 70% isopropyl alcohol wash, and a final DuraprepTM (3M, St. Paul, MN) paint.The skin and subcutaneous tissue is infiltrated with 10 cc of Marcaine® (0.5% bupiva-caine and 1:200,000 epinephrine solution; Sanofi Winthrop, New York, NY). An approx2-cm incision is then carried down through the skin and soft tissue. A stab incision ismade in the lumbosacral fascia, and attention is then turned toward the placement ofthe MetRx

TM dilator and tubular retractor system (Medtronics) (Fig. 4). With fluoro-

366 Tredway and Fessler

Fig. 4. (A) MetRx dilators for access to surgical site; (B) MetRx 18-mm tubular retractorworking channel; (C) MetRx table clamp and flexible arm; (D) 30° endoscope utilized for MEDprocedure.

scopic guidance, a Steinman pin is docked on the ipsilateral laminofacet junction.Serial fluoroscopic images are obtained while the dilators are used to spread the musclefibers apart and advance the dilators to rest on the laminofacet junction. Once the largestdilator is placed, the 18-mm working channel is placed over the dilator. The workingchannel should be docked on the laminofacet region overlying the surgical disc space.After correct positioning is obtained, the working channel is locked into place using theflexible arm attached firmly to the table clamp (Fig. 5).

Removal of Soft Tissue

The soft tissue overlying the facet and lamina is removed with the use of Bovie elec-trocauterization (Fig. 6A). After identification of the facet and lamina, a straight orangled curette is used to define the sublaminar space from the ligamentum flavum (Fig.6B). Next, a 45° angled Kerrison rongeur facilitates removal of the laminofacet junc-

Lumber Microendoscopic Discectomy 367

Fig. 4. (Continued)

368 Tredway and Fessler

Fig. 5. (A) After placement of the first dilator, a gentle sweeping motion over the lamina willclear soft tissue and allow for easier docking of the subsequent dilators. This maneuver shouldbe performed carefully because the dilator can inadvertently traverse the interlaminar space(image courtesy of Medtronics). (B) Image depicting MetRx dilators docked on laminofacetjunction (image courtesy of Medtronics); (C) image depicting 18-mm working channel attachedto flexible arm (image courtesy of Medtronics); (D) intraoperative image of MED setup showing30° endoscope attached to working channel.

tion (Fig. 6C). The laminotomy is continued cephalad to the point of attachment of theligamentum flavum. With this degree of osseous removal, an angled curette can easilybe placed in the subligamentous/epidural space. Using a gentle sweeping motion withthe angled curette, the ligamentum flavum can be removed safely with a Kerrisonrongeur (Fig. 6D). The ligamentum flavum is removed to identify adequately the lateraledge of the thecal sac as well as the nerve root (Fig. 6E). Often, a partial medialfacetectomy is performed to enable access to the disc space and prevent significant

retraction of the nerve root. Care must be taken to avoid removal of more than half ofthe facet because this could lead to iatrogenic destabilization of the spine at the surgicallevel. The complication of an overaggressive facetectomy could inevitably require afusion procedure for stabilization.

Discectomy

Once the lateral edge of the dura and nerve root has been identified, a Love nerveroot retractor is placed into the working channel, and the nerve root and thecal sac aregently retracted. Often, the epidural venous plexus will need to be coagulated with abipolar electrocautery in order to decrease blood loss and improve visualization whileperforming the discectomy (Fig. 7A). The disc space is identified and an angled dissectoris employed to separate the anterior aspect of the thecal sac from the posterior annulus ofthe disc. Once free, the retractor can be placed in a secure position enabling access to theherniated disc fragment. An annulotomy is performed with a retractable scalpel (Fig. 7B).A straight curette is placed within the disc space and the fragment is loosened (Fig. 7C). Adown-angled curette may also be helpful in dislodging the herniated fragment. A pituitaryrongeur facilitates removal of the offending disc fragment (Fig. 7D). This technique isrepeated until the fragment is removed and the nerve root appears to be decompressed(Fig. 7E). It is imperative to review the MRI or CT scan prior to closure because oftentimesthe radiological imaging will demonstrate the exact location of the herniated fragment.It is also critical to explore the disc space laterally because the disc may be sequesteredwithin the foramen. Once hemostasis is achieved with a combination of bipolar elec-trocautery and thrombin-soaked Gelfoam® (Pharmacia, Kalamazoo, MI), the wound isirrigated with an antibiotic-containing irrigation solution.

Wound Closure

The working channel is removed under endoscopic visualization, and the wound is irri-gated again with an antibiotic-containing irrigation solution. Attention is then turned towardclosure of the incision. A 2-0 absorbable suture is used to reapproximate the lumbosacralfascia if possible. The subcutaneous and dermal layers are closed with a 3-0 absorbable

Lumber Microendoscopic Discectomy 369

Fig. 5. (Continued)

370 Tredway and Fessler

Lumber Microendoscopic Discectomy 371

suture. The skin is closed with a 4-0 absorbable suture in a subcuticular fashion and thencovered with Dermabond® (2-octyl cyanoacrylate; Ethicon, Somerville, NJ) (Fig. 8).

Avoidance of Complications

The complications of MED are similar to those observed in lumbar microdiscectomy.The overall mortality in lumbar microdiscectomy has been reported to be 0.06% and isusually owing to septicemia, myocardial infarction, or pulmonary embolus.(35,36). Toour knowledge, no deaths have been reported utilizing the MED technique, but other risksinclude infection, unintended durotomy with leakage of cerebrospinal fluid (CSF), injuryto the nerve root, and recurrent disc herniation. Superficial wound infections have been

Fig. 6. Intraopertive endoscopic image demonstrating (A) removal of soft tissue overlying lam-ina with Bovie electrocauteriztion, (B) angle curette defining sublaminar space, (C) removal oflamina with a Kerrison punch, (D) removal of ligamentum flavum with a Kerrison punch, and(E) lateral edge of dural and nerve root after removal of ligament.

372 Tredway and Fessler

Lumber Microendoscopic Discectomy 373

reported to occur in 0.9–5% of microdiscectomy cases, with a <1% occurrence of deepinfections. In a retrospective series of patients undergoing the MED procedure, Perez-Cruet et al. (34) reported only one superficial wound infection in 150 patients. Theoreti-cally, the MED procedure provides a smaller portal for entry of infectious agents incomparison to open microdiscectomy. The incidence of durotomy has been reported to be0.3–13% and has been reported to be higher (18%) in reoperative microdiscectomies (37).Interestingly, leakage of CSF was observed in 8 of the 150 patients undergoing the MEDprocedure (34). This complication is usually treated by placing Dura-Gen® dural graftmatrix (Integra NeuroSciences, Plainsboro, N J) over the durotomy and carefully rein-forcing with a layer of fibrin glue. Lumbar drainage for 48 h is also recommended in

Fig. 7. Intraoperative endoscopic image demonstrating (A) bipolar electrocauterization ofepidural plexus overlying disc space, (B) opening of annulus with a retractable scalpel, (C) loosen-ing of NP with a straight curette, (D) removal of offending disc fragment with a pituitary rongeur,and (E) nerve roots after disc removal (note the conjoined roots in this particular case).

larger dural tears. Nerve root injuries or cauda equina syndrome was not observed in thesmall series of 150 patients reported by Perez-Cruet et al. (34). Finally, recurrent disc her-niation has been reported to be 4% with microdiscectomy (38). Similarly, recurrent discherniation observed in patients undergoing the MED procedure was approx 2.7% (4 of150). These 4 patients were subsequently treated with repeat MED (34).

Results

No randomized, prospective study has been performed comparing MED to microdis-cectomy. The largest retrospective series reported in the literature to date involved 150patients ranging in age from 18 to 76 yr (34). In this series, a mean operative time of 97min (110 min for the first 30 cases compared with 75 min for the last 30 cases), a meanhospital stay of 7.7 h, and a mean time to return to work of 17 d were reported. Clinicaloutcome as determined by the modified McNabb criteria, with an average follow-up of12 mo (range: 3–24 mo), demonstrated a 77% excellent outcome, 17% good outcome,3% fair outcome, and 3% poor outcome. These data support the claim that MED is asafe, effective alternative to traditional lumbar microdiscectomy.

CONCLUSION

The treatment of herniated lumbar discs continues to provide intellectual and technicalchallenges for the spine surgeon. Although recent advancements in the area of optics,microendoscopic instruments, and minimally invasive access has allowed the surgeon toperform the MED with less iatrogenic injury to surrounding soft tissue, patient selectionremains one of the most critical factors in achieving excellent clinical outcomes. Therefore,with a concerted effort to train young spine surgeons in patient selection as well as thistechnique, perhaps one day MED will supplant microdiscectomy and become the preferredprocedure for herniated lumbar discs refractory to conservative nonoperative management.

374 Tredway and Fessler

Fig. 8. MED incision with Dermabond skin closure.

REFERENCES

1. Hart LG, Deyo RA, Cherkin DC. Physician office visits for low-back pain: frequency, clinicalevaluation, and treatment patterns from a U.S. national survey. Spine 1995;20(1):11–19.

2. Mayer TG, et al. A prospective two-year study of functional restoration in industrial low-back injury: an objective assessment procedure. JAMA 1987;258(13):1763–1767.

3. Frymoyer J. Back Pain and sciatica. N Engl J Med 1988;318:291–300.4. Luo X, et al. Estimates and patterns of direct health care expenditures among individuals

with back pain in the United States. Spine 2004;29(1):79–86.5. Vesalius A. De Humani Corporis Fabrica Libri Septem, 1555, pp. 71–73.6. Kitano T, et al. Biochemical changes associated with the symptomatic human interverte-

bral disk. Clin Orthop 1993(293):372–377.7. Inerot S, Axelsson I. Structure and composition of proteoglycans from human annulus

fibrosus. Connect Tissue Res 1991;26(1-2):47–63.8. Olczyk K. Age-related changes in proteoglycans of human intervertebral discs. Z Rheumatol

1994;53(1):19–25.9. Harrington JF, et al. Herniated lumbar disc material as a source of free glutamate available

to affect pain signals through the dorsal root ganglion. Spine 2000;25(8):929–936.10. Dinning TA, Schaeffer HR. Discogenic compression of the cauda equina: a surgical emer-

gency. Aust NZ J Surg 1993;63(12):927–934.11. Damkot DK, PM, Lord J, Frymoyer JA. The relationship between work history, work

environment and low-back pain in men. Spine 1984;9(4):395–399.12. Kelsey JL. An epidemiological study of the relationship between occupations and acute

herniated lumbar intervertebral discs. Int J Epidemiol 1975;4(3):197–205.13. Kelsey JL. An epidemiological study of acute herniated lumbar intervertebral discs.

Rheumatol Rehabil 1975;14(3):144–159.14. Kelsey JL, Hardy RJ. Driving of motor vehicles as a risk factor for acute herniated lumbar

intervertebral disc. Am J Epidemiol 1975;102(1):63–73.15. Lewis PJ, et al. Long-term prospective study of lumbosacral discectomy. J Neurosurg

1987;67(1):49–53.16. Jensen MC, et al. Magnetic resonance imaging of the lumbar spine in people without back

pain. N Engl J Med 1994;331(2):69–73.17. fa*ger CA, Freidberg SR. Analysis of failures and poor results of lumbar spine surgery.

Spine 1980;5(1):87–94.18. fa*ger CA. Observations on spontaneous recovery from intervertebral disc herniation. Surg

Neurol 1994;42(4):282–286.19. Oppenheim H, KF. Uber Einklemmung bzw: strangulation der cauda equina. Dtsch Med

Wochenschr 1909;35:697–700.20. Elsberg C. Experiences in spinal surgery: observations upon 60 laminectomies for spinal

disease. Surg Gynecol Obstet 1913;16:117–120.21. Dandy W. Loose cartilage from intervertebral disk simulating tumor of the spinal cord.

Arch Surg 1929;19:660–672.22. Mixter WJ, Barr JS. Rupture of intervertebral disc with involvement of the spinal canal.

N Engl J Med 1934;211:210–215.23. Caspar W. A new surgical procedure for lumbar disc herniation causing less tissue damage

through a microsurgical approach, in Advances in Neurosurgery (Wullenweber R, BrockM, Hamer J, Klinger M, Spoerri O, eds.), Springer-Verlag: Berlin, 1977, pp. 74–77.

24. Yasargil M. Microsurgical operation for herniated disc, in Advances in Neurosurgery(Wullenweber R, Brock M, Hamer J, Klinger M, Spoerri O, eds.), Springer-Verlag: Berlin,1977 p. 81.

25. Smith L. Enzyme dissolution of the nucleus pulposus in humans. JAMA 1964;187:137–140.

Lumber Microendoscopic Discectomy 375

26. Agre K, et al. Chymodiactin postmarketing surveillance; demographic and adverse experiencedata in 29,075 patients. Spine 1984;9(5):479–485.

27. Dubuc FL, et al. Four brief reports on long-term results of intradiscal chymopapain. ClinOrthop 1986(206):42–44.

28. Javid MJ, Nordby EJ. Current status of chymopapain for herniated nucleus pulposus. Neur-surg Q 1994;4:92–101.

29. Pauza KJ, et al. A randomized, placebo-controlled trial of intradiscal electrothermal therapyfor the treatment of discogenic low-back pain. Spine J 2004;4(1):27–35.

30. Hijikata SA, Yamagishi M, Nakayama T, Oomori K. Percutaneous discectomy: a newtreatment method for lumbar disc herniation. J Toden Hosp 1975;5:5–13.

31. Kambin P, Gellman H. Percutaneous posterolateral discectomy of the lumbar spine. ClinOrthop 1983;174:127–132.

32. Kambin P, Savitz MH. Arthroscopic microdiscectomy: an alternative to open disc surgery.Mt Sinai J Med, 2000;67(4):283–287.

33. Smith MW, FK. MED: The first 100 cases. Paper presented at the Annual Meeting of theCongress of Neurological Surgeons; Seattle, WA, 1998.

34. Perez-Cruet MJ, et al. Microendoscopic lumbar discectomy: technical note. Neurosurgery2002. 51(5 Suppl):129–136.

35. Ramirez LF, Thisted R. Complications and demographic characteristics of patients undergoinglumbar discectomy in community hospitals. Neurosurgery 1989;25(2):226–230; discussion,230, 231.

36. Deyo RA, et al. Morbidity and mortality in association with operations on the lumbarspine: the influence of age, diagnosis, and procedure. J Bone Joint Surg Am 1992;74(4):536–543.

37. Goodkin R, Laska LL. Unintended “incidental” durotomy during surgery of the lumbarspine: medicolegal implications. Surg Neurol 1995;43(1):4–12; discussion 12–14.

38. Davis RA. A long-term outcome analysis of 984 surgically treated herniated lumbar discs.J Neurosurg 1994;80(3):415–421.

376 Tredway and Fessler

Index 377

Index

377

Alajouanine, T. H., 6Ancient medicine, management of back and

leg pain, 1–3Anesthesia,

microendoscopic discectomy, 364, 365minimally invasive lumbar surgery,

75, 76percutaneous transpedicular discectomy,

188X-Stop interspinous process distraction

device, 320, 323Annular debridement, arthroscopic, 123Annular fenestration, historical perspective, 9Annular tear,

diagnosis, 122pain generation, 119–121

Annulus fibrosus, features, 42, 43, 45Antibiotics, prophylaxis, 76, 77Aristotle, 1Arthroscopes, posterolateral access, 50Arthroscopic microdiscectomy, see Lumbar

disc herniationAtavi®,

comparison with other internal fixationsystems, 233, 234

components, 230, 231fixation systems, 232, 233FlexPosure® retractor, 230, 231, 235MiCOR™ allograft block, 233

Avicenna, 1–3

B,C

Bozzini, P. D., 18C-arm, selection, 73Cannulas,

development , 17, 18lumbar disc herniation repair,

oval cannula positioning, 89, 90placement, 81–83

posterolateral access, 49Capsular ligamentum flavum complex,

anatomy, 45Chemonucleolysis,

Chymodiactin downfall factors, 354,355, 357

chymopapain patents and preparations,352, 353

Food and Drug Administration approval,352–354

negative press, 355–357rationale, 351, 364

Chymodiactin, see ChemonucleolysisChymopapain,

adjunctive therapy for disc herniations,213, 351

history of use, 351–357, 364Coccydynia, pain management, 275Computed tomography (CT),

annular tear, 119, 121intervertebral disc pain diagnosis, 122lateral recess stenosis, 145skin entry site selection for lumbar disc

herniation repair, 80stereotactic imaging, see Frameless

stereotactic imagingtranspedicular biopsy guidance, 175

Cotugno, Domenico, 3, 4Crouzon, O., 6–8CT, see Computed tomography

Degenerative disc disease, see Intervertebraldisc degeneration

Disc fragment retrieval,historical perspective, 15, 21lumbar disc herniation repair, 91, 92

Discectomy, see Lumbar disc herniation;Microendoscopic discectomy;Percutaneous transpediculardiscectomy; Selective endoscopicdiscectomy™

Discitis, postoperative prevention, 116Dorsal root ganglion (DRG),

anatomy, 31–33, 35vascularity, 160–163

DRG, see Dorsal root ganglionDural sac, anatomy, 39Dyesthesia, see Skin hypersensitivity

EElectromyography (EMG), neuromonitoring,

214, 251

378 Index

EMG, see ElectromyographyEndoscopy,

advantages, 252disadvantages, 252endoscope features, 251historical perspective, 18, 19, 21–23lumbar transperitoneal and retroperitoneal

spinal procedures,outcomes, 256, 257overview, 254, 255principles, 255, 256

thoracoscopic spinal endoscopy,outcomes, 253, 254overview, 252, 255principles, 252, 253

Epidural adipose tissue, features, 38Epidural bleeding,

complication prevention, 113control, 99, 100

Epidural injection,caudal epidural,

catheter delivery, 283, 284imaging, 282injectate, 283needle techniques, 283overview, 282patient positioning, 282

lumbar interlaminar epidural,complications, 282imaging, 280, 281injectate, 282needle techniques, 281overview, 279, 280patient positioning, 280

lumbar transforaminal epidural,imaging, 276–278injectate, 279needle techniques, 278nerve avoidance, 279overview, 276patient positioning, 276

steroid injection technique, 272–274Evocative chromodiscography™, see

Selective endoscopic discectomy™

Facet joint,anatomy, 271block,

duration of relief, 291

imaging, 289injectate, 291needle techniques, 289–291overview, 289patient positioning, 289

degeneration and pain generation, 121,271, 272

laser rhizotomy, 262, 263FlexPosure® retractor, see Atavi®Foraminoplasty, laser, 212, 213, 263, 264Frameless stereotactic imaging,

costs, 347image acquisition, 336, 337limitations, 347, 348overview, 335presurgical planning, 337, 338prospects, 345, 346spinal surgery applications, 341–344, 348stereotaxis principles, 335, 336surgical navigation systems, 338–340virtual fluoroscopy, 341

G,H

Growth factors, utilization in minimallyinvasive posterior fusion and fixation,229, 230

Herniation, see Lumbar disc herniationHijikata, S., 12, 17, 22, 205Hippocrates, 1

IDET, see Intradiscal electrothermalcoagulation

Infection, prevention, 76, 77, 116Interlaminar access, minimally invasive

surgery, 59Intervertebral disc degeneration,

diagnosis, 122epidemiology, 295history of study, 4, 359pain generation, 119, 295treatment,

expandable cage introduction, 123nuclear replacement, see Nucleus

replacementoverview, 295, 296prosthetic disc implantation, 296

Intracanalicular ligaments, anatomy, 39Intradiscal electrothermal coagulation

(IDET), indications and principles, 364

Index 379

K,LKyphoplasty,

complications, 246outcomes, 244–246overview, 240, 242technique, 242–244vertebroplasty comparison, 247

Lasers,facet rhizotomy, 262, 263foraminoplasty, 212, 213, 263, 264invention, 259medicine applications, 259percutaneous laser discectomy, 259–262selective endoscopic discectomy™, 212,

213sympathectomy, 264–267

Lateral recess stenosis, see also Spinalstenosis

clinical presentation, 145, 146diagnosis, 145history of study, 135surgical management,

outcomes, 147overview, 147skin hypersensitivity and management,

147–149Lazarevic test, lumbar disc herniation, 61–63Low back pain, see PainLumbar disc herniation,

arthroscopic and endoscopic disc surgeryadvantages,

complication minimization,epidural scarring, 67, 68nerve root damage, 67, 68

cost-effectiveness, 70documentation, 68epidural venous system protection, 68paraspinal muscle integrity

maintenance, 68postoperative imaging, 68reherniation risks, 66spinal stability maintenance, 68visualization, 66

chemonucleolysis, 364clinical features and diagnosis, 61, 63,

65, 66, 360–362complications and prevention,

cannulas,migration, 112, 113placement, 110, 112

discitis, 116infection, 116needle insertion, 109, 110power-driven suction nuclear resectors,

113skin hypersensitivity, 114–116

intradiscal electrothermal coagulation, 364magnetic resonance imaging, 359, 363microendoscopic discectomy,

anesthesia, 364, 365complications and prevention,

371.373, 374discectomy, 369outcomes, 374patient positioning, 365prospects, 374, 375soft tissue removal, 367–369surgical approach, 365–367wound closure, 369, 371

minimally invasive surgery technique,anesthesia, 75, 76annulotomy, 87, 88antibiotic prophylaxis, 76, 77C-arm, 73cannulas,

oval cannula positioning, 89, 90placement, 81–83

disc fragment retrieval, 91, 92fluid management, 92–95foraminal and extraforaminal

herniation retrieval, 95–97instrument lodging site selection, 81needle positioning, 77, 79operating room table, 70–73paramedial or central disc herniation

retrieval, 98patient positioning, 73, 75radiographic landmarks, 81radiolucent frame, 73recurrent disk herniation retrieval, 107,

108sequestered disc herniation retrieval,

98–101, 104–106skin entry site selection, 79–81surgical site identification, 77thoracic or thoracolumbar herniation

retrieval, 107triangular working zone examination,

83, 86, 87physical examination, 362, 363

380 Index

Magnetic resonance imaging (MRI),annular tear, 121, 221intervertebral disc pain diagnosis, 122,

221, 359lateral recess stenosis, 145lumbar disc herniation, 359, 363stereotactic imaging, see Frameless

stereotactic imagingMED, see Microendoscopic discectomyMedial branch,

block,diagnostics, 287–289imaging, 287injectate, 287needle techniques, 287overview, 286, 287patient positioning, 287

radiofrequency ablation, 272MiCOR™ allograft block, see Atavi®Microendoscopic discectomy (MED),

anesthesia, 364, 365complications and prevention, 371.373,

374discectomy, 369outcomes, 374patient positioning, 365prospects, 374, 375soft tissue removal, 367–369surgical approach, 365–367wound closure, 369, 371

Minimally invasive spinal surgery, see alsospecific techniques,

approaches, see Posterolateral access,historical perspective, 18, 19, 21–23

MRI, see Magnetic resonance imaging

Needles,insertion complications, 109, 110positioning, 77, 79posterolateral access, 49

Nerve root,anatomy, 31–33, 35damage minimization in arthroscopic and

endoscopic disc surgery, 67, 68dorsal root ganglion vascularity, 160–163endoscopic neurolysis, 123intradural supply of intradural roots,

155–157slow vs rapid compression effects, 160

vascular supply of intradural roots, 157,158

Nuclear mass reduction,historical perspective, 9, 11, 12, 14pain reduction outcomes, 123

Nucleus pulposus,anatomy, 41replacement, see Nucleus replacement

Nucleus replacement,historical perspective, 297hydrogel polymers,

cadaveric endoscopic implantation andbiomechanical studies, 304–308

ideal criteria, 298polyvinyl alcohol/

polyvinylpyrrolidone hydrogel,dissolution behavior, 299–302stability, 298–302stress relaxation and fatigue,

302–304prospects for study, 309, 310types, 297, 298

intervertebral disc mechanics, 298rationale in degenerative disc disease, 297

Operating table, selection, 70–73Osteoporosis, see Vertebral compression

fracture

Pain,annular tear, 119–121facet joint degeneration, 121inflammatory agents, 121intervertebral disc sources, 119low back pain economic impact, 359minimally invasive techniques in

management,caudal epidural,

catheter delivery, 283, 284imaging, 282injectate, 283needle techniques, 283overview, 282patient positioning, 282

coccydynia, 275epidural steroid injection, 272–274facet joint block,

duration of relief, 291imaging, 289injectate, 291

Index 381

needle techniques, 289–291overview, 289patient positioning, 289

lumbar interlaminar epidural,complications, 282imaging, 280, 281injectate, 282needle techniques, 281overview, 279, 280patient positioning, 280

lumbar transforaminal epidural,imaging, 276–278injectate, 279needle techniques, 278nerve avoidance, 279overview, 276patient positioning, 276

medial branch block,diagnostics, 287–289imaging, 287injectate, 287needle techniques, 287overview, 286, 287patient positioning, 287

radiofrequency medial branch ablation,272

sacrococcygeal joint injection,imaging, 291indications, 292injectate, 291needle techniques, 291overview, 291patient positioning, 291

sacroiliac joint block,diagnostics, 286imaging, 284–286injectate, 286needle techniques, 286overview, 284patient positioning, 284

sacroiliac joint steroid injection,274, 275

steroid therapy, 122venous hypertension role, 153–155

Paraspinal muscles,anatomy, 29integrity maintenance in arthroscopic and

endoscopic disc surgery, 68Percutaneous transpedicular biopsy, see

Transpedicular biopsyPercutaneous transpedicular discectomy,

pyogenic spondylodiscitis,advantages, 196, 197, 200contraindications, 199, 200indications, 188outcomes, 198, 199

technique,anesthesia, 188phases, 191, 193, 196Steinmann pin, 189, 191

Percutaneous vertebroplasty (PVP),complications, 241, 242kyphoplasty comparison, 247outcomes, 241overview, 240, 241polymethyl methacrylate injection, 240,

241Periannular bleeding, complication

prevention, 113Posterior longitudinal ligamentum, anatomy,

45Posterolateral access,

bilateral biportal approach, 53–55, 100,101

instruments, 49, 50interlaminar access, 59unilateral biportal approach, 55, 58uniportal approach, 53

Prosthetic discs, development, 123, 124PVP, see Percutaneous vertebroplastyPyogenic spondylodiscitis,

diagnosis, 180, 181management,

overview, 183surgery, see Percutaneous

transpedicular discectomynatural history, 180, 181

Q–S

Queckenstedt maneuver, 151Radiofrequency medial branch ablation,

technique, 272Sacrococcygeal joint injection,

imaging, 291indications, 292injectate, 291needle techniques, 291overview, 291patient positioning, 291

Sacroiliac joint,block,

diagnostics, 286imaging, 284–286

382 Index

injectate, 286needle techniques, 286overview, 284patient positioning, 284

pain generation, 274, 284steroid injection, 274, 275

Safe zone, see Triangular working zoneSciatica, history of diagnosis and

management, 3–8Segmental instability,

classification, 124definition, 124surgical stabilization,

arthroscopic anterior columnstabilization augmented withpercutaneously insertedpedicular bolts and sc plates,

applications, 141instruments, 126outcomes, 142screw fixation of first sacral

segment, 134–136, 138, 139,141

technique, 128–133arthroscopic interbody fusion

development, 125, 126historical perspective, 124, 125indications, 124minimally invasive posterior fusion

and fixation,Endius Atavi® system, see Atavi®growth factor utilization, 229, 230milestones, 227–229

thoracoscopic and laparoscopicdecompression and stabilization,142

Selective endoscopic discectomy™,chymopapain adjunctive therapy, 213electrothermal therapy, 209, 210evocative chromodiscography™,

207–209exclusion criteria, 220foraminal epidurography and therapeutic

injections, 216historical perspective, 205, 206imaging, 221inclusion criteria, 217–219indications, 217lasers in foraminoplasty, 212, 213neuromonitoring, 214–216prospects, 221technique, 221–223

Yeung Endoscopic Spine Surgery system,205, 207, 222

Skin hypersensitivity,lateral recess stenosis surgery

complication and management,147–149

postoperative prevention, 114–116Somatosensory evoked potential (SSEP),

neuromonitoring, 214Spinal fusion, see Segmental instabilitySpinal instability, see Segmental instabilitySpinal stenosis, see also Lateral recess

stenosis,anatomy, 315–317clinical features, 315history of study, 315treatment,

overview, 317, 318X-Stop interspinous process

distraction device,advantages, 327, 330anesthesia, 320, 323operative technique, 320, 322, 323outcomes, 318, 319, 323–327

Spine anatomy,annulus fibrosus, 42, 43, 45capsular ligamentum flavum complex, 45dural sac, 39epidural adipose tissue, 38intracanalicular ligaments, 39nerve roots and ganglia, 31–33, 35nucleus pulposus, 41paraspinal muscles, 29posterior longitudinal ligamentum, 45thoracolumbar fascia, 29triangular working zone, 29, 31vasculature, 35–38, 151–153, 157, 158

Spondylodiscitis, see Pyogenicspondylodiscitis

SSEP, see Somatosensory evoked potentialSteroid therapy,

coccydynia, 275epidural steroid injection, 272–274perineural therapy, 122sacroiliac joint steroid injection, 274, 275skin hypersensitivity management, 149transforaminal therapy, 122

Substance P, pain generation role, 121Sympathectomy, laser, 264–267

Thermal intradiscal therapy, principles, 124

Index 383

Thoracolumbar fascia, anatomy, 29Transpedicular biopsy,

anatomy, 172, 173complications, 174, 175computed tomography guidance, 175historical perspective, 167, 168rationale, 167, 168, 174technique, 168, 170, 171

Transpedicular discectomy, seePercutaneous transpediculardiscectomy

Triangular working zone,anatomy, 29, 31arthroscopic and endoscopic examination,

83, 86, 87identification for anchoring of

instruments, 15, 16Triphens, posterolateral access, 49, 50

Vasoactive intestinal polypeptide (VIP), paingeneration role, 121

VCF, see Vertebral compression fractureVenous hypertension,

animal experiment studies of venousradiculomedullary circulation, 159,160

neurogenic pain role, 153–155slow vs rapid compression of nerve roots,

160Vertebral compression fracture (VCF),

epidemiology, 239kyphoplasty,

complications, 246outcomes, 244–246overview, 240, 242technique, 242–244vertebroplasty comparison, 247

nonoperative management, 239percutaneous vertebroplasty,

complications, 241, 242kyphoplasty comparison, 247outcomes, 241overview, 240, 241polymethyl methacrylate injection,

240, 241prospects for management, 247, 248

Vertebroplasty, see Percutaneousvertebroplasty

VIP, see Vasoactive intestinal polypeptide

X-Stop interspinous process distractiondevice,

advantages, 327, 330anesthesia, 320, 323operative technique, 320, 322, 323outcomes in spinal stenosis, 318, 319,

323–327

YESS system, see Yeung Endoscopic SpineSurgery system

Yeung Endoscopic Spine Surgery (YESS)system, selective endoscopicdiscectomy™, 205, 207, 222

Arthroscopic and Endoscopic Spinal Surgery - Text and Atlas 2nd ed - P. Kambin (Humana, 2005) WW - [PDF Document] (2024)