Image-Guided Spine Interventions Second Edition
Image-Guided Spine Interventions Second Edition
Edited by
John M. Mathis, BS, MSc, MD
Medical Director, Center for Advanced Imaging, Roanoke, VA, USA
Stanley Golovac, MD
Co-Director and CEO, Space Coast Pain Institute, Merritt Island, FL, USA
Editors John M. Mathis Medical Director Center for Advanced Imaging Roanoke, VA USA
[email protected]
Stanley Golovac Co-Director and CEO Space Coast Pain Institute Merritt Island, FL USA
[email protected]
ISBN 978-1-4419-0351-8 e-ISBN 978-1-4419-0352-5 DOI 10.1007/978-1-4419-0352-5 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2009941057 © Springer Science+Business Media, LLC 2010 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
To the women in my life who make all things possible: Krista, Jamie, Juanita, Jean, Ida, Mildred, and Vernice. John M. Mathis To my loving wife, Elena, and my two children, Alexandra and Conrad. Stanley Golovac
Preface to the Second Edition
The field of minimally invasive spine intervention continues to evolve rapidly. Numerous changes and developments have been added to our tools and capabilities over the past decade. Image-Guided Spine Interventions, Second Edition, describes the varied and numerous tools that are available to the physicians who are involved in providing these procedures to appropriate patients. This book embraces clinical evaluation, pharmacological requirements, procedural recommendations, and a spectrum of procedures that will be of interest to the minimally invasive spine interventionalist. It covers a broad range of material that is presented by experts in each field, such as discography, percutaneous discectomy, vertebroplasty and balloon kyphoplasty, epidural steroid injections, selective nerve root blocks, autonomic nerve blocks, spinal stimulator implantation, and spine vascular intervention. This book will be useful to all physicians who deal with back pain, including pain anesthesiologists, spine neurosurgeons, orthopedists, and radiologists. Both editors have clinical practices devoted solely to spine procedures and pain management. This rewarding area has grown rapidly, and it provides great opportunities for clinical development as well as potential to aid a very needy patient population. We sincerely hope this work will be useful in helping our colleagues – new and old – establish, grow, and refine a minimally invasive spine interventional practice. Roanoke, VA Merritt Island, FL
John M. Mathis, MD, MSc (Rad Physics) Stanley Golovac, MD
vii
Contents
1 Spine Anatomy 1 John M. Mathis, Ali Shaibani, and Ajay K. Wakhloo 2 Materials Used in Image-Guided Spine Interventions 29 John M. Mathis 3 Patient Evaluation and Criteria for Procedure Selection 39 Harold J. Cordner 4 The Surgeon’s Perspective: Image-Guided Therapy and Its Relationship to Conventional Surgical Management 57 F. Todd Wetzel 5 Image-Guided Percutaneous Spine Biopsy 75 A. Orlando Ortiz, Gregg H. Zoarski, and Allan L. Brook 6 Discography 107 Aaron Calodney and Duane Griffith 7 Percutaneous Lumbar Discectomy 147 Stanley Golovac 8 Epidural Injections for the Treatment of Spine-Related Pain Syndromes 157 Matthew T. Ranson and Timothy R. Deer 9 Pulsed Radiofrequency Procedures in Clinical Practice 175 Richard M. Rosenthal 10 Facet Joint Injections and Sacroiliac Joint Injections 207 Louis J. Raso 11 Autonomic Nerve Blockade 229 Stanley Golovac and John M. Mathis
ix
x
Contents
12 Percutaneous Vertebroplasty 249 John M. Mathis and Charles Cho 13 Implanted Drug Delivery Systems 279 Lisa Jo Stearns 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine 299 Shahram Rahimi, Ali Shaibani, and Ajay K. Wakhloo 15 Kyphoplasty: Balloon Assisted Vertebroplasty 337 John M. Mathis, Charles H. Cho, and Wayne J. Olan 16 Sacroplasty 355 Charles H. Cho, John M. Mathis, and Keith E. Kortman 17 Spinal Cord Stimulation: Uses and Applications 375 Stanley Golovac Index 397
Contributors
Allan L. Brook, MD Associate Professor of Clinical Radiology and Neurosurgery, Albert Einstein College of Medicine, Director, Interventional Neuroradiology, Montefiore Medical Center, Department of Radiology, Bronx, NY 10467, USA Aaron Calodney, MD Texas Spine and Joint Hospital, Tyler, TX 75701, USA Charles H. Cho, MD, MBA Neuroradiologist, Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA 02115, USA Harold J. Cordner, MD Department of Pain Management, Sebastian River Medical Center, Sebastian, FL 32958, USA Timothy R. Deer, MD Clinical Professor, Department Anesthesiology, West Virginia University, Charleston, WV 25301, USA Stanley Golovac, MD Co-Director and CEO, Space Coast Pain Institute, Merritt Island, FL 32953, USA Duane Griffith, MD Texas Spine and Joint Hospital, Tyler, TX 75701, USA Keith E. Kortman, MD Department of Radiology, Sharp Memorial Hospital, Del Mar, CA 92014, USA John M. Mathis, BS, MSc (Rad. Physics), MD Medical Director, Center for Advanced Imaging, Roanoke, VA 24018, USA Wayne J. Olan, MD Director, Neuroradiology, Department of Radiology, Suburban Hospital, Bethesda, MD 20854, USA
xi
xii
Contributors
A. Orlando Ortiz, MD, MBA Chairman, Department of Radiology, Winthrop-University Hospital, Mineola, NY 11501, USA Shahram Rahimi, MD Research Associate, Department of Radiology, Northwestern University, Feinberg School of Medicine, Chicago, IL 60611, USA Matthew T. Ranson, MD Attending Physician, The Center for Pain Relief, Charleston, WV 25301, USA Louis J. Raso, MD Physician, Jupiter, FL 33477, USA Richard M. Rosenthal MD, DABPM, FIPP Medical Director,Utah Center for Pain Management and Research, Provo, UT 84604, USA Ali Shaibani, MD Associate Professor, Radiology and Neurosurgery, Department of Radiology and Neurosurgery, Northwestern University, Feinberg School of Medicine, Chicago, IL 60611, USA Lisa Jo Stearns, MD Medical Director, Center for Pain and Supportive Care, Scottsdale, AZ 85258, USA Ajay K. Wakhloo, MD, PhD Professor of Radiology, Neurology, and Neurosurgery, Chief, Division of Neurointerventional Radiology, Department of Radiology, University of Massachusetts, Worcester, MA 01655, USA F. Todd Wetzel, AB, MD Professor, Orthopedic Surgery and Neurosurgery, Department of Orthopedic Surgery, Temple University, Philadelphia, PA 19140, USA Gregg H. Zoarski, MD Director, Diagnostic and Interventional Neuroradiology, Department of Diagnostic Imaging, University of Maryland, Baltimore, MD 21201, USA
1 Spine Anatomy John M. Mathis, Ali Shaibani, and Ajay K. Wakhloo
Introduction The spine and its anatomical components are complex. Authors have approached it from a variety of perspectives including surgical, anatomical, and diagnostic (imaging). The interest of the authors in spinal anatomy concerns the minimally invasive treatment of pathological processes affecting the spine. This chapter describes spine anatomy that is of interest to the image-guided interventionist.
Physical Components Bones The spine is composed of 33 bones: there are 7 cervical vertebra, 12 thoracic vertebra, 5 lumbar vertebra, 5 sacral segments (fused), and 4 coccygeal segments (variably fused).1 Natural curvature is found throughout the spine (Figure 1.1). Viewed from the side, the cervical spine is convex forward, the thoracic spine is convex backward (centered at T7), the lumbar spine is convex forward, and the sacral bone is convex backward. The vertebrae progressively enlarge from the cervical through the lumbar regions. There is also variability in vertebra size at any particular level based on the individual’s body size (Figure 1.2). The size of a vertebra is of extreme importance when one is performing percutaneous vertebroplasty (PV) or kyphoplasty (KP). In both PV and KP, the most common side effects are created by cement leak. These leaks result from natural or pathological holes in vertebra as well as overfilling. To avoid overfilling, it is important to appreciate the volume range of vertebral bodies between the cervical and lumbar regions (Table 1.1). Theoretical volume calculations show vertebral body volumes ranging from 7.2 mL in the cervical spine to 19.6 mL in the lumbar region. These volumes are computed for a hollow cylinder with dimensions taken from each spine region. Because of the From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_1, © Springer Science + Business Media, LLC 2010
1
2 Chapter 1 Spine Anatomy Figure 1.1. The spine viewed from the lateral projection. Natural curvature varies from convex forward in the cervical and lumbar region to concave forward in the thoracic and sacral regions. The thoracic curvature is particularly prone to kyphosis with vertebral compression fractures in this territory (Reprinted with the kind permission of Springer Science+Business Media from Mathis JM ed. ImageGuided Spine Interventions. New York: Springer Science+Business Media, 2004).
thickness of cortical and trabecular bone, the fillable volume is on the order of 50% of the theoretical volume. The fillable volume will again be diminished by the amount of the vertebral collapse following a compression fracture. As seen in Table 1.1, the 50% compressed volume for a C5 vertebra is between 1.8 and 2.2 mL. In the thoracic spine (T9), the 50% compressed volume is 3.8 mL. At L3, the 50% compressed volume is 4.9 mL. It is easy to see why very small volumes of cement sometimes
Physical Components 3
Figure 1.2. Representative vertebrae from the cervical, thoracic, and lumbar regions. Relative vertebra body sizes and configuration changes are shown (Reprinted with the kind permission of Springer Science+Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
can achieve adequate biomechanical augmentation for pain relief. These volumes differ considerably from region to region in the spine. The spinal canal is formed by the posterior wall and the posterior elements of the vertebral body (pedicles and lamina). The pedicles join the vertebral body to the posterior lamina. The vertebral pedicle is a complex three-dimensional cylindroid structure that consists of a thin shell of compact bone (which is thickest on the medial surface). Table 1.1. Vertebral volume estimates from the cervical to lumbar regions Vertebral level C5
Theoretical volume (mL) 7.2
Fillable volume (mL) 3.60
50% Compressed volume (mL) 1.8
T9
15.3
7.65
3.8
L3
19.6
9.80
4.9
Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004.
4 Chapter 1 Spine Anatomy
Figure 1.3. Axial CT scans of three vertebrae. (A) Scan of a T11 vertebra demonstrates the sagittal configuration (straight posterior to anterior) of the pedicle with respect to the vertebral body. The line demonstrates the general tract that a needle would take during vertebroplasty by means of a transpedicular approach. In the scan of an L5 vertebra (B), the transpedicular approach (black line) is nearly 45° away from the sagittal plane. In the scan at T1 (C), the transpedicular angle with the sagittal plane (black line) approaches 45°, similar to the angle found in the lowest lumbar vertebra (Reprinted with the kind permission of Springer Science+Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Physical Components 5
Figure 1.3. Continued.
The compact shell surrounds a much larger center that is filled with cancellous bone.2–8 The pedicles are extremely important because they provide a safe tunnel through which the interventionist can gain access to the vertebral body for biopsy, vertebroplasty, and kyphoplasty. The pedicles in the cervical region are small and present a poor access to the vertebral body in this region. However, the thoracic and lumbar pedicles provide good potential access. Pedicles progressively increase in size from the upper thoracic (T4) to the lower lumbar (L5) spine. The angle of the pedicles relative to the vertebral body changes as does their size. From T4 to T12, the pedicles have a relatively straight sagittal (anterior-to-posterior) orientation (Figure 1.3a). In the lumbar spine from L1 to L4, there is a slow but progressive angle away from the sagittal orientation. At L5, the angle is extreme and can approach 45° away from the sagittal plane (Figure 1.3b). Progressive angulation also occurs from T4 toward the cervical region (Figure 1.3c). Therefore, both pedicle size and angulation are important when one is planning a transpedicular approach during intervention. Though the size of the pedicles varies from region to region and from individual to individual, one can be comfortable that a 13-gauge cannula (0.095 in., outside diameter) will fit through essentially all adult pedicles from T4 to L5. In most individuals, a 10- to 11-gauge cannula (0.134–0.120 in., outside diameter) will safely pass through pedicles from T12 to L5. When the size of the pedicle (or its absence in neoplastic disease) precludes a transpedicular approach, a parapedicular route may be necessary. This route takes the entry device along the lateral margin of the pedicle and above the tranverse process. In the thoracic spine, this trajectory is
6 Chapter 1 Spine Anatomy
Figure 1.4. The parapedicular approach. (A) In this lateral view, notice that the needle enters above the transverse process. (B) Needle placement position for a parapedicular approach in vertebroplasty (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
generally along the junction of the rib with the adjacent transverse process and vertebral body (Figure 1.4a, b). The articulation of the rib and vertebral body forms the costovertebral joint. The costotransverse joint is the junction of the rib and transverse process, with the intervening space filled with the costotransverse ligament. The parapedicular needle entry point will be along the lateral and posterior vertebral border in the paraspinal soft tissues. The paraspinus space is filled with fatty tissue and venous structures. Venous bleeding is common here, but this is usually self-limiting as long as no coagulopathy exists. Occasionally, the posterior costophrenic sulcus contains lung that bulges beyond the border of the rib, making pneumothorax also possible. The bones of the vertebra make up part of the central skeleton, inside of which the elements of the blood are made. This occurs in the intertrabecular (marrow) space. The venous system connects and becomes continuous with this marrow space (Figure 1.5a, b). This venous connection provides one of the main avenues for cement leakage during vertebroplasty or kyphoplasty. The venous route most important for
Physical Components 7
Figure 1.5. (A) The venous communications typical in a vertebra: AEVP anterior external venous plexus; IVV intervertebral vein; ARV anterior radicular vein; PRV posterior radicular vein; PIVP posterior internal venous plexus; PEVP posterior external venous plexus; BVV basivertebral vein. (B) Axial CT scan demonstrating the posterior wall opening (black arrows) that allows the major veins of the interior of the vertebra (BVV basivertebral vein) to communicate with epidural veins (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
8 Chapter 1 Spine Anatomy
potential leakage is through the posterior vertebral wall, communicating with the veins in the epidural space. Leakage into this location can create compression of the cord or nerve roots. Venous leak anterior or laterally can result in cement migration into central veins carrying blood to the lungs (resulting in pulmonary emboli). Intervertebral Discs and Joints The intervertebral discs and joints interface with the various vertebrae in the spine. Together with the ligamentous attachments, these elements allow the vertebrae to move through bending and rotation. However, these discs and joints wear and may be the source of pain caused by degeneration. The image-guided interventionist must deal with these structures during discography, percutaneous discectomy, intradiscal electrothermal therapy, facet blocks, and dorsal ramus neurolysis. The intervertebral discs are composed of an outer ring of fibrocartilage called the annulus fibrosis (Figure 1.6a, b). The annulus is attached to the cartilaginous endplates of the vertebrae and constrains the inner disc core called the nucleus pulposus. The annulus is the thickest anteriorly. It is thin posteriorly, which coincides with the area most commonly associated with annular tears and disc herniations. The outer annular fibers, which are more densely packed, are referred to as Sharpey’s fibers. The nucleus pulposus is made of cells that are notochordal remnants. It is composed of collagen fibrils that are embedded in a proteoglycan matrix that contains water. With aging and degeneration, water is lost, and the nucleus becomes progressively fibrotic and smaller. Because of the spine curvature (Figure 1.1), the angle of the plane of the disc between the vertebral endplates is variable through the spine. This variation requires different imaging angulation to enter the disc without obstruction by the adjacent vertebral margins. Appropriate imaging angulation is necessary for accurate needle placement in discography and percutaneous disc therapy. The apophyseal or facet joints are paired joints between the posterior elements of two adjacent vertebrae. They are curved joints that are oriented obliquely to the sagittal plane (Figure 1.6c). The joints are asymmetric in about 30% of the population.9 Every joint consists of an articular process from each of the adjacent vertebra. The joint has a synovial lining with a fibrous capsule (Figure 1.6a, b). The nerve supply is from the medial division of the dorsal ramus of the spinal nerve that reaches the joint from both the nerve above and below the joint on the ipsilateral side (Figure 1.7a). The joint is believed to be a source of Figure 1.6. The intervertebral disc. (A) Lateral drawing depicting the disc components and their association with the adjacent hyaline cartilaginous endplates. (B) Axial drawing demonstrating that the annulus fibrosus is thickest anteriorly. The capsule and lining of the facet joint also are shown. (C) Axial CT scan showing the complex configuration of the facet joints (black arrows) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Physical Components 9
10 Chapter 1 Spine Anatomy
Figure 1.7. (A) Axial drawing of the nerve exiting the neural foramina of a lumbar vertebra and giving off the posterior ramus. A medial branch of this nerve will supply the capsule of the facet joint. These innervations arise from medial branches from both above and below each joint. The gray and white rami communicantes connect the autonomic ganglia with the anterior division of the spinal nerves. (B) Axial MR scan showing the neural foramina (white arrows) of a lumbar vertebra containing the dorsal root ganglia (white arrowhead).
nonradiating axial pain that is typically aggravated by hyperextension and rotation. Because the joint is curved, image guidance can be confusing, and entry into the joint may be difficult, particularly when there is degenerative disease. A small synovial recess along the superior and
Physical Components 11
(C) Axial drawing of a cervical vertebra. This highlights the close proximity of the vertebral artery with the exiting spinal nerve. (D) Axial MR scan of a cervical vertebra demonstrating the vertebral artery (black arrow) along the anterior neural foramina. IJ internal jugular vein; CA carotid artery; VA vertebral artery (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
inferior margins of the joint will allow access without passing through the curved bone margins. Facet blocks are used for therapeutic treatment and diagnostic confirmation of the pain source. As they rarely have prolonged therapeutic benefit, neurolysis of the joint nerve supply with chemical or radiofrequency (RF) ablation is most often used for long-term pain control. Spinal Nerves Entire books have been written about the anatomy of the spinal nerves. For the purpose of this text, the authors emphasize the elements that are of prime importance to the interventionist.
12 Chapter 1 Spine Anatomy
In the spine, as in the brain, there are central (spinal cord) and peripheral components (peripheral nerves) of the nervous system. The peripheral nerves are the components that are of major importance from the standpoint of image-guided therapy. The peripheral nerves are responsible for somatosensory, somatomotor, and autonomic nerve function. The spinal nerves exiting the neural foramina are composed of an anterior and a posterior division that coalesce into a single nerve in the neural foramina (Figure 1.7a). The anterior division of the spinal nerve contains the motor fibers that originate in the cell bodies in the anterior horn of the spinal cord. Preganglionic autonomic fibers course in this anterior division as well and originate in the anterolateral horn of the spinal cord. These fibers branch to become the white rami communicantes, and synapse with postganglionic autonomic fibers in the autonomic ganglia along the spine to form the sympathetic trunk, or extend to ganglia adjacent to end organs (celiac, mesenteric, etc.) via the splanchnic nerves. The sensory neurons (primary afferent) are found in the dorsal root of the spinal nerve. The dorsal root ganglia contain sensory cell bodies; the axons of these sensory nerves originate in specialized sensory structures (Golgi tendon organs, Ruffini endings from the joints, muscle spindles, pacinian corpuscles in fascial planes, etc.) and carry somatic sensory information about touch, proprioception, stereognosis, pain, and temperature. Visceral afferent information is also returned through the dorsal horn. The sensory nerves separate within the cord and take characteristic routes to the brain, where they reach varying levels of consciousness based on their type. The various types of peripheral nerve are different not only because of their relative function but also because of physical size and conduction velocity. The motor fibers are the largest and have the fastest conduction velocity. General sensory fibers mediating touch and proprioception are intermediate in size, while pain and nocioceptive fibers are the smallest and have the slowest conduction velocity. To block these fibers, an anesthetic must bind to (and block) three consecutive sodium channels (nodes of Ranvier). This means that in clinical practice, a smaller amount of anesthetic is needed to block smaller fibers (pain) and that regular sensory and motor fibers are more resistant to anesthetic block. This provides for the ability to obtain differential neurologic blockade and allows pain to be blocked without the loss of motor function (if appropriate volume and concentration of anesthetic are chosen). Selective nerve root blocks are used for diagnostic and therapeutic purposes. The injectate (chosen for a specific effect) is introduced into or just lateral to the neural foramina. This places the injected agent around or peripheral to the dorsal root ganglion. In the lumbar region, the foramina are larger than in the thoracic and cervical spine. Venous vascular structures are common in the lumbar foramina, but a much lower chance of an arterial injection exists here (Figure 1.7a, b). In the cervical region, the vertebral artery lies along the anterior border of the foramina and there are numerous feeding branches to the cord that pass through cervical foramina from the vertebral, ascending cervical and deep cervical vessels (Figure 1.7c, d). Great care must be exercised when one is doing nerve blocks in this region because direct injury
Physical Components 13
(dissection) to any of these vessels can occur and result in infarction of the cervical cord. Injection of anesthetics or steroids into the artery can create seizure or stroke, respectively. Nerve blocks of the autonomic nerves are also of great use in the mediation of visceral pain in processes such as cancer and pelvic inflammatory disease, or to provide relief from reflex sympathetic dystrophy. To specifically block the autonomic nerves, leaving the somatosensory and somatomotor nerves in tact, injections are placed around the autonomic ganglia in the location where the problem exists. The autonomic nervous system has two components, which are called the parasympathic nerves and the sympathetic nerves (Figure 1.8). There are no sympathetic cells in the brain. Parasympathetics originate from the brain (and run in the cranial nerves) and the sacral cord (S2–S4). The sympathetics originate in the spinal cord between T1 and L2.
Figure 1.8. Artist’s conception of the autonomic nervous system, with its sympathetic and parasympathetic components. Note that these elements have very different origins within the central nervous system (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
14 Chapter 1 Spine Anatomy
Both systems synapse in peripheral ganglia, and each carries both motor and sensory nerves to visceral organs (blood vessels, glands, heart, bowel, etc.). The parasympathetic and sympathetic systems function antagonistically. The parasympathetic system constricts the pupil, decelerates the heart, lowers blood pressure, relaxes the sphincters, and contracts hollow visceral organs. The sympathetic system dominates in periods of excitement and causes dilation of the pupil, accelerates the heart, increases blood pressure, contracts sphincters, and relaxes smooth muscle of the hollow viscera.10–12 Most organs receive innervations from both parts of the autonomic system. Though sympathetics do not exist in the brain, they reach the organs of the head and neck through ganglia located in the cervical region with preganglionic fibers arriving via the sympathetic track and coursing into and through the inferior (stellate), middle, and superior cervical ganglia. Postganglionic fibers are distributed along blood vessels to the various end organs. Nerves to thoracic, abdominal, and pelvic viscera arrive from the sympathetic chain (traveling along the lateral vertebral bodies) or splanchnic nerves to the ganglia adjacent to end organs such as the heart or pancreas. Once again, blockade of selected ganglia can reduce visceral pain or hyperactivity of the sympathetic system. Anatomical Spaces The anatomical spaces around the spine that are of primary interest to the image-guided interventionist are those found around the thecal sac (epidural space) and the outlet space for the exiting nerve roots (foraminal space). The epidural space begins immediately inside the bony spinal canal and extends from the foramen magnum (it does not exist inside the skull) to the caudal hiatus of the sacrum (Figure 1.7a, c). It surrounds the thecal sac, and the exiting nerve roots pass through it into the neural foramina. It is filled with fibro–fatty areolar tissue and vascular elements, mostly venous. The arterial supply that traverses the epidural space is basically limited to spinal arteries that enter along spinal nerves to supply the cord, nerve roots, and the parts of the vertebrae adjacent to and circumscribing the epidural space. The epidural space varies in size. It is the smallest in the cervical region (1–2 mm) and enlarges progressively toward the lower lumbar and sacral area. At L2–3, the space is 5–6 mm wide. If the neck is flexed, the posterior cervical epidural space can increase to 3–4 mm.13,14 The epidural space is easily accessed in the sacral and lumbar regions via needle placement through the intralaminar, transforaminal, and caudal (caudal hiatus) routes. Access is via the intralaminar and transforaminal routes for the thoracic and cervical epidural space. The epidural space can be septated naturally. Postoperative scarring, which can locally obliterate the space, commonly occurs along the posterior and lateral borders of the thecal sac in the site of the operative field. When this occurs, the intralaminar approach is of reduced utility, and puts puncture of the thecal sac at higher risk. Transforaminal epidural access then becomes the most dependable method.
Vascular Anatomy of the Spinal Cord 15
The neural foramina of the spine exist bilaterally from the cervical through the sacral regions. In the sacral region, the foramina are bounded by sacral bone on all sides with exit points both dorsally and ventrally. From the cervical through the lumbar spine, each foramen is bounded by a vertebral body and disc anteriorly, the pedicle superiorly and inferiorly, and facet articular processes posteriorly. The nerve root passes through foramina accompanied by small branches of the spinal artery and veins, and surrounded by fatty tissue. The veins communicate with the epidural venous plexus. In the cervical spine, the vertebral artery runs immediately anterior to the exiting nerve root in the foramen transversarium (Figure 1.7c, d). This close association of the nerve and artery can put the artery (or one of its many central communicating branches) at risk during transforaminal approaches for epidural injections and nerve blocks. Direct injury can result in dissection or occlusion. Anesthetic or steroid injection into the artery may cause seizure or stroke, respectively.
Vascular Anatomy of the Spinal Cord Neuroradiologists, and interventional neuroradiologists in particular, need an accurate understanding of the normal vascular anatomy of the spine and spinal cord. This is especially true because spinal angiography is less commonly performed than cerebral angiography, and safe and adequate performance of spinal angiography and intervention is predicated on accurate and complete knowledge of the normal vascular anatomy. This section provides a concise and accurate vascular anatomy of the spine and spinal cord, with guidelines for performing spinal angiography in a safe and complete manner.15 Arterial supply can be divided conceptually into a macrocirculation (the supply up to the cord surface) and a microcirculation (the supply beyond the anterior and posterior spinal arteries).16 Macrocirculation Conceptually, the arterial supply to the cord can be described from the “outside in” as a segmental supply based on the embryological development of the body. In the first few weeks of development in the embryo, the embryo is divided into 31 somites in a rostral–caudal direction. These 31 somites correspond to the 31 pairs of spinal nerves (8 cervical, 12 thoracic, 5 lumbar, 5 sacral, 1 coccygeal). The segmental artery, through its branches, supplies blood to all the ipsilateral derivatives of its corresponding metamer (neural crest, neural tube, and somites), that is, muscle, skin, bone, spinal nerve, and spinal cord. Each segmental/metameric artery is named for the nerve it accompanies in the neural foramen. In the beginning of embryological development, each segmental artery has a branch supplying the cord, but most regress over time and only a few are left to provide flow to the spinal cord. The remainder will remain unimetameric, supplying the related nerve, dura, vertebral body, and paraspinous muscles. At the end of embryological development, of the 62 metameric arteries
16 Chapter 1 Spine Anatomy
(31 pairs), four to eight will supply the ventral spinal axis (anterior spinal artery), and 10–20 will supply the dorsolateral/pial network (posterior spinal arteries). The process of regression of cord supply is more pronounced caudally, which results in fewer sources of medullary supply, such as the dominant artery of Adamkiewicz. The simplified algorithm for the vascular supply at each segmental level is: Major arterial trunk → spinal/segmental artery (31 pairs) → radicular artery Or Radiculopial or radiculomedullary artery → paired posterior or single anterior spinal artery (Figure 1.9).
Figure 1.9. Illustration depicting (1) a segmental artery, (2) the somatic branches (vertebral body supply), (3) an intercostal artery or muscular branch, (4) the dorsospinal trunk, (5) paravertebral longitudinal anastomosis, (6) a radiculomedullary artery, (7) the dorsal somatic branch, (8) nerve, (9) the dura, (10) the radicular branches to the dorsal nerve root, (11) the radicular branches to the ventral nerve root, (12) the ventral spinal axis (ASA), (13) a radiculopial artery (dorsal radiculomedullary), (14) the dorsolateral spinal network (or posterior spinal arteries) (PSA) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Vascular Anatomy of the Spinal Cord 17
Figure 1.10. Selective intercostal artery injection showing longitudinal pretransverse anastomosis between segmental arteries (open arrow) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
The segmental arteries form extraspinal and extradural longitudinal anastomoses, which can be divided as follows: 1. Ventrolateral: An example is the ascending cervical artery (from the thyrocervical trunk) in the neck. 2. Pretransverse (anterior to transverse processes: An example is the vertebral artery in the cervical region, or lateral–sacral arteries. These supply the sympathetic system in the thoraco-lumbar area (Figure 1.10). 3. Dorsal–longitudinal: These anastomoses branch to the midline insertion of the spinous process muscles. The deep cervical artery from the costocervical trunk is an example. The major arterial trunks supplying the radicular arteries at each level are the following: 1. Cervical: Vertebral arteries, ascending cervical branch of the thyrocervical trunk, deep cervical branch of the costocervical trunk, occipital branch of the External Carotid Artery (ECA), and ascending pharyngeal branch of ECA. 2. Thoracic: Branches of the costocervical trunk, internal thoracic branch of the subclavian artery, supreme intercostal branch of the aorta, and intercostal branches of the aorta. 3. Lumbosacral: Lumbar branches of the aorta, middle sacral branch of the aorta, lateral sacral branches of the internal iliac arteries, and iliolumbar branch of the common iliac arteries.
18 Chapter 1 Spine Anatomy
At each of the 31 levels, each segmental artery supplies blood to the dorsal and ventral nerve roots, thereby being given the designation “radicular” artery. At some levels, the segmental artery supplies blood not just to the nerve root but also beyond, to the spinal cord, via branches connecting either to the pial/coronal arterial network, or directly to the anterior spinal artery. In the former condition, these segmental arteries are named “radiculopial,” and in the latter condition, they are called “radiculomedullary.” At the level of the surface of the spinal cord, there is a single anterior spinal artery (ventral spinal axis) and there are paired posterior spinal arteries. Connecting these two networks is the pial/coronal (centripetal) network of small arteries. Some experts consider the paired posterior spinal arteries to be part of the pial/coronal network, representing more dominant craniocaudally oriented channels. According to this definition, those segmental arteries providing supply to the posterior spinal arteries are more accurately designated as radiculopial rather than radiculomedullary. The authors use this definition for this chapter. The flow in the spinal arteries, anterior and posterior, is bidirectional, depending on the dominant medullary artery at each level, as well as the time needed for the aortic systolic pulse wave to reach each radiculomedullary or radiculopial artery (more distal arteries will experience the aortic systolic pulse wave later, which also contributes to bidirectional flow). Radicular Arteries At each of the 31 levels, the spinal/segmental artery provides branches to the dorsal and ventral nerve roots, after giving off branches to the paraspinous musculature, vertebral body, and dura. The only exception is the C1 level, where there may be congenital absence of the radicular branches. Under normal physiological circumstances, the radicular branches are usually too small to be seen angiographically. Radiculopial Arteries The radiculopial arteries supply the nerve roots (via radicular branches), then run ventral to either the dorsal or the ventral nerve root to supply blood to the pial/centripetal (vasa corona) network. These arteries do not supply the anterior spinal artery (ventral axis) directly. They do have anastomoses with pial branches of the anterior spinal artery, however. There are more dorsal than ventral radiculopial arteries. The dorsal radiculopial arteries (called dorsal radiculomedullary arteries by some authors) are more important, and these are the ones referred to as the radiculopial arteries henceforth in this chapter. Their numbers vary from individual to individual. On average, there are 3–4 dorsal radiculopial arteries in the cervical region, 6–9 in the thoracic region, and 0–3 in the lumbosacral region. Radiculomedullary Arteries The radiculomedullary arteries provide the only segmental supply to the ventral spinal axis (anterior spinal artery) and are the dominant source of supply to the cord over several functional segments. After giving off their radicular branches to the nerve roots, they run along the ventral surface of the nerve root, occasionally giving off a pial
Vascular Anatomy of the Spinal Cord 19
Table 1.2. Diameter of spinal arteries Artery Diameter (Mm) Artery
Cervical
Artery of cervical enlargement
0.4–0.6
Thoracic
Lumbosacral
Artery of Adamkiewicz
0.55–1.2
Ventral spinal axis (anterior spinal artery)
0.2–0.5
0.2–0.4
0.5–0.8 (artery of the filum)
Dorsolateral spinal arteries (posterior spinal arteries)
0.1–0.2
0.2–0.25
0.1–0.4
Reprinted with the kind permission of Springer Science+Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004; Modified with permission from Lasjaunias P, Berenstein A, TerBrugge KG. Surgical Neuroangiography, 2nd ed. New York: Springer-Verlag, 2001.
collateral, then supply the anterior spinal artery. Their numbers vary from individual to individual. On an average, there are 2–4 (ventral) radiculomedullary arteries in the cervical region, 2–3 in the thoracic region, and 0–4 in the lumbosacral region. Classically, two radiculomedullary arteries have received special attention: the arteries of the cervical and lumbar enlargements. The artery of the lumbar enlargement is also known as the artery of Adamkiewicz (Table 1.2). In 75% of patients, the artery of Adamkiewicz arises between T9 and T12, more commonly on the left. When its origin is above T8 or below L2, there is another major contributor to the anterior spinal artery either caudally or cranially. In 30–50% of cases, it also gives a major contribution to dorsolateral pial system (paired posterior spinal arteries) (Figure 1.11). The connection of the radiculomedullary artery to the ventral spinal axis is Y shaped in the cervical area because the artery does not have to ascend very high before it meets the ventral spinal axis. The classic hairpin anastomosis is seen at the thoracic and lumbar levels. The single ventral spinal axis (anterior spinal artery) is continuous from the basilar artery to the artery of the filum terminale. The artery of the filum terminale is the caudal extension of the anterior spinal artery. The anterior spinal artery may be focally discontinuous, especially at the thoracic level. The ventral spinal axis runs in the subpial space in the ventral sulcus of the spinal cord, dorsal to the veins. In the cervical region, there may be congenital lack of fusion of the embryological dual ventral spinal axes, resulting in a short unfused segment. Microcirculation The circulation beyond the level of the ventral spinal axis and the dorsolateral pial network (posterior spinal arteries) is conceptually divided into a centrifugal (from the center of the cord out) and a centripetal (from the pial surface toward the center of the cord) system. Centrifugal System The centrifugal system is also known as the sulcocommisural system. The ventral spinal axis (anterior spinal artery) gives rise to 200–400 sulcocommissural arteries within the ventral sulcus of the spinal cord.
20 Chapter 1 Spine Anatomy
Figure 1.11. Selective injection of an intercostal branch supplying the ventral spinal axis showing the artery of Adamkiewicz (artery of the thoracolumbar enlargement (small arrow)) and the ventral spinal axis (anterior spinal artery (arrowhead)), and classic hairpin loop of the radiculomedullary artery (open arrow) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
These arteries penetrate the sulcus and enter the central gray matter, where they give off branches radiating outward toward the peripheral white matter. Each sulcocommissural artery usually supplies one half (right or left) of the cord. The sulcocommissural system will supply the majority of the gray matter and the ventral half of the cord. Before entering the cord substance, each sulcocommissural artery gives off cranial and caudal anastomotic branches to other sulcocommissural arteries. Craniocaudal anastomoses also are seen within the substance of the cord. Early in development, before the disproportional elongation of the spinal column in relation to the cord, the sulcal arteries have a completely horizontal course. With growth and the disproportionate elongation of the spinal column, they assume an ascending course. Yoss found that occlusion of the artery of the lumbar enlargement in primates caused severe damage to the ventrolateral two thirds of the cord, where the artery entered, and for a distance above and below.17 The territory of the cord supplied by the centrifugal system (from the ventral spinal axis)
Vascular Anatomy of the Spinal Cord 21 Figure 1.12. The “centrifugal” arterial system: (1) the radiculomedullary artery, (2) the ventral spinal axis, and (3) the sulcocommissural arteries (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
is comparatively as large as that supplied by the internal carotid artery relative to one cerebral hemisphere (Figure 1.12). Centripetal System The centripetal system is also known as the dorsolateral pial supply (from posterior spinal arteries). This network covers the dorsal and dorsolateral surface of the cord and has two dominant craniocaudal channels known as the posterior spinal arteries. At the craniocervical junction, supply to this system is directly from the transdural vertebral arteries, or from posterior inferior cerebellar arteries when their origin is below the dura. Below this level, arterial supply is from radiculopial arteries (Figure 1.13). This system has a dorsal component and a lateral component (located between the dorsal and ventral nerve roots), which are interconnected. This network gives rise to radial/coronal arteries (vasa corona), which extend around the circumference of the cord and have anastomoses to the ventral spinal axis. The radial/coronal arteries give off perforating branches to the cord all along their course. These short perforating branches extend axially into the white matter and a portion of the gray matter of the dorsal horns. The perforating branches of the radial/coronal arteries have intramedullary anastomoses with branches of the sulcocommissural arteries dorsolaterally, ventrolaterally, and ventrally. There are also short, extramedullary longitudinal (craniocaudal) anastomoses between the radial/coronal arteries. These anastomoses are relatively small, however, and cannot provide adequate craniocaudal supply in the case of arterial occlusion. The dorsolateral pial network must therefore be regarded primarily as an axial system of arterial supply.
22 Chapter 1 Spine Anatomy Figure 1.13. The “centripetal” arterial system: (1) The radiculomedullary artery, (2) the ventral spinal axis, (3) the sulcocommissural artery, and (4) the coronary arteries from the dorsolateral spinal network (posterior spinal arteries) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Somatic Arterial Supply The metameric/segmental artery is centered at the level of the intervertebral disc, the corresponding nerve, and the myelomere (cord). Therefore, the vertebral body is fed by two consecutive segmental arteries on each side (for a total of four). Each of the four will supply approximately 25% of the vertebral body. However, extensive anastomoses within the substance of the vertebrae often permit all or most of the vertebral body to be seen from one arterial injection. The somatic arteries anastomose on the posterior surface of the vertebral body, making a characteristic hexagon or diamond-shaped network on anterior–posterior angiography (Figures 1.14 and 1.15). Angiography 1. Lumbar and lower thoracic: Usually a hemivertebral blush is seen from one segmental arterial injection; this effect is evident only 25% of the time. 2. Upper thoracic: The right intercostal artery will opacify the right hemivertebra and the ventral half of the left hemivertebra. 3. Cervical and sacral: Symmetry is the rule, with opacification of the ipsilateral hemivertebra.
Spinal Venous Anatomy The authors will approach the description of the venous anatomy of the spinal cord from the inside out. Venous drainage of the cord is divided into an intrinsic system (in proximity to the centrifugal arterial
Spinal Venous Anatomy 23
Figure 1.14. Retrocorporeal hexagonal anastomosis of dorsal somatic branches to the vertebral body: (1) Vertebral body, (2) nerve root sleeve, (3) pretransverse longitudinal anastomosis, (4) segmental artery, (5) radicular artery, (6) disc and dorsal somatic branch (open arrow), (7) intravertebral disc (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
system but, naturally, with an opposite direction of flow) and the extrinsic system (in proximity to the centripetal arterial system). In general, the ventral dominance of the arterial system is not seen in the venous system. The venous drainage of the cord is relatively equally divided dorsally and ventrally. The intrinsic venous system comprises dorsal and ventral sulcal (sulcocommissural) veins that collect the venous outflow from the central gray matter. The extrinsic venous system can be thought of as containing the venous perforators draining into the radial/coronal veins, which in turn drain into the primary dorsal and ventral longitudinal collecting veins. These longitudinal collecting veins in turn drain into the radicular veins (analogous to the radiculomedullary and radiculopial veins), which eventually empty into the ventral epidural venous plexus. In addition to the main dorsal and ventral draining veins, there are short, intersegmental, lateral longitudinal veins linking adjacent radial veins. These lateral longitudinal channels are not large enough, however, to form a functional dominant craniocaudal channel like the dorsal and ventral systems.
24 Chapter 1 Spine Anatomy
Figure 1.15. (A) Subtracted and (B) unsubtracted selective injection of the lumbar artery (solid arrow) showing hexagonal dorsal anastomosis of dorsal somatic arteries (open arrows) (supply to the vertebral body) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Spinal Venous Anatomy 25
Table 1.3. Number of spinal veins Region Number of longitudinal veins
Cervical
T1–T8
T9–T12
Lumbosacral
Ventral surface:
3 > 1
3 > 1
1 > 3
1
Dorsal surface:
1 > 3
3 > 1
1
1
3 > 1: in most patients, three veins would be present; some would have only one; 1 > 3: in most patients, one vein would be present; some would have three. Reprinted with the kind permission of Springer Science+Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004.
Flow in the thoracic longitudinal channels is bidirectional, with cervical drainage of its most cranial portion and lumbar drainage of its most caudal part. There can be multiple longitudinal venous channels, especially in the thoracic region and ventrally (Table 1.3). The main ventral longitudinal venous channel is known as the anterior median vein (Figure 1.16).
Figure 1.16. The venous drainage of the spinal cord: (1) The dorsal root, (2) the ventral nerve root, (3 and 10) the coronal venous plexus (radial veins), (4) the anterior median vein of the ventral longitudinal venous system, (5) a dorsal longitudinal vein, (6) a transmedullary anastomotic vein, (7) a dorsal sulcal vein, (8) a radiculomedullary vein, (9) a ventral longitudinal vein, (10) circumferential spinal vein (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
26 Chapter 1 Spine Anatomy
The radicular (radiculomedullary) veins drain into either spinal nerve venous channels in the neural foramina or to a dural venous pool, both of which eventually empty into the ventral epidural venous plexus. The epidural (extradural) venous system has a prominent ventral component and a small, much less important, dorsal component. The ventral epidural veins receive venous drainage from the vertebral bodies (through anterior and posterior venules), the spinal cord (via radiculomedullary veins), and the dura, and they are also involved in some resorption (via arachnoid granulations along the nerve root sleeves) of the cerebrospinal fluid. The ventral epidural venous plexus forms a valveless, retrocorporeal, hexagonal anastomotic plexus, which is essentially continuous craniocaudally. The direction of flow within this plexus is not unidirectional; rather, it depends upon the location of the outflow vein at each anatomic level. The ventral epidural venous plexus drains into multiple different outflow veins, depending upon the anatomical level. These are as follows: 1. Cervical: Drainage is into the vertebral veins, which in turn empty into the innominate veins. 2. Thoracic: Drainage is into the intercostal veins, which then empty into the azygous and hemiazygous systems and subsequently the inferior vena cava. 3. Lumbar: Drainage is multiple, involving the ascending lumbar vein (on the left), the azygous and hemiazygous systems, and the left renal vein. The final common pathway is generally the inferior vena cava. 4. Sacral: Drainage is into sacral veins, emptying into the lateral sacral veins, and subsequently the internal iliac veins. References 1. Netter FH. CIBA Collection of Medical Illustrations: Nervous System. Vol 1. CIBA, West Caldwell, NJ 1962:21–30. 2. Kothe R, O’Holleran JD, Liu W, Panjabi MM. Internal architecture of the thoracic pedicle. Spine 1996;21:264–270. 3. Misenhimer GR, Peek RD, Wiltse LL, Rothman SLG, Widell EH. Internal architecture of the thoracic pedicle cortical and cancellous diameter as related to screw size. Spine 1989;14:367–372. 4. Phillips JH, Kling TF, Cohen MD. The radiographic anatomy of the thoracic pedicle. Spine 1994;19:446–449. 5. Panjabi MM, Goel V, Oxland T, Takata K, Duranceau J, Krag M, Price M. Human lumbar vertebrae. Quantitative three-dimensional anatomy. Spine 1992;17:299–306. 6. Ebraheim NA, Rollins JR, Xu R, Yeastling RA. Projection of the lumbar pedicle and its morphometric analysis. Spine 1996;21:1296–1300. 7. Robertson PA, Stewart NR. The radiologic anatomy of the lumbar and lumbosacral pedicles. Spine 2000;25:709–715. 8. Ebraheim NA, Xu R, Ahmad M, Yeastling RA. Projection of the thoracic pedicle and its morphometric analysis. Spine 1997;22:233–238. 9. Griffiths HJ, Parantainen H, Olsen PN. Disease of the lumbosacral facet joints. Neuroimaging Clin North Am 1993;3:567–575. 10. Group M, Stanton-Hicks M. Neuroanatomy and pathophysiology of pain related to spinal disorders. Radiol Clin North Am 1991;29:665–673. 11. Bonica JJ. The Management of Pain. Philadelphia: Lea & Febiger, 1990:95–121. 12. DeMyer W. Neuroanatomy. Philadelphia: Williams & Wilkins, 1988:87–126.
Spinal Venous Anatomy 27 13. Bromage PR. Anatomy of the epidural space. In Bromage PR (ed): Epidural Analgesia. Philadelphia: WB Saunders, 1978:8–20. 14. Reynolds AF, Roberts PA, Pollay M, Stratemeier PH. Quantitative anatomy of the thoracolumbar epidural space. Neurosurgery 1985;17:905. 15. Lasjaunias P, Berenstein A, TerBrugge KG. Surgical Neuroangiography, 2nd ed. New York: Springer, 2001. 16. Krauss WE. Vascular anatomy of the spinal cord. Neurosurg Clin North Am 1999;10(1):9–15. 17. Yoss RE. Vascular supply of the spinal cord: the production of vascular syndromes. Univ Mich Med Bull 1950;16:333–345.
2 Materials Used in Image-Guided Spine Interventions John M. Mathis
Imaging Equipment Most image-guided spine interventions are accomplished well with fluoroscopic guidance. It goes without saying that good visualization of the anatomical area being treated is necessary. Most modern fluoroscopic equipment will provide this capability. It is important to view the target anatomy from multiple projections, and therefore a C-arm configuration is used. Fixed-plane fluoroscopic equipment (commonly used for gastrointestinal work) is not sufficient. The most sophisticated equipment in the multidirectional category is the fixed-base, biplane fluoroscopic room (Figure 2.1a). These rooms are common for interventional neuroradiologists but are not routinely available otherwise. The ability to view the target anatomy in two projections at once is a definite luxury and offers the fastest possible needle insertion capability. However, single-plane C-arm systems are fine for all these procedures. The greatest disadvantage is the reduced speed experienced with vertebroplasty and kyphoplasty, but these procedures can also be performed adequately without biplane capability. Fixed-base C-arm (dedicated angiographic) rooms (Figure 2.1b) are more desirable than portable C-arms (Figure 2.1c). This is primarily because of image quality, but also because of the ease of use by the operating physician. Fixedbase angiographic equipment is motorized and can be controlled by the physician. By contrast, in most portable units, projection changes must be made manually by a technologist. This requirement has the disadvantage of requiring the physician to describe the desired projection rather than being able to select it personally, and this generally slows the process. Also, projections that are repeatedly used can be programmed into memory on a fixed-base machine and automatically retrieved with the press of a button. These features make use of the fixed-base rooms simpler and faster. New fluoroscopic equipment generally has good image quality. This may not be true of older equipment; therefore, old equipment From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_2, © Springer Science + Business Media, LLC 2010
29
Figure 2.1. (A) Biplane fluoroscopic equipment is the most sophisticated and useful of the possible room setups. However, it is extremely expensive and not generally available for the average spine imaging operator. (B) A fixed, single-plane fluoroscopic arrangement that affords a fluoroscopic image of excellent quality. The C-arm is motor driven and computer controlled for ease of operation. It will be slower than the maximally efficient biplane room seen in (A).
Imaging Equipment
31
Figure 2.1. Continued. (C) A modern, mobile C-arm fluoroscopic arrangement commonly used in operating room situations. This apparatus offers good imaging capability but is much more cumbersome to use than the fixed-base systems. Though acceptable, it is the least desirable setup (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
should be checked by a certified radiological physicist for image quality and radiographic exposure or output. Additionally, portable equipment may not have enough power to penetrate thick body areas. This can limit visualization in some situations and reduce the safety of procedures. Ultimately, all fluoroscopic images face this limit. Large patients or difficult locations, such as the high thoracic region (lateral T1–T4, which are blocked by the shoulders), will have limited visualization. In these situations, alternate imaging should be considered. The use of Computed Tomography (CT) has grown both because of availability and the limitations of fluoroscopy. Some operators use CT simply because it is what they have available. Certain regions of the body that may be hard to image with fluoroscopy are better suited to CT imaging. Additionally, complex clinical situations, such as percutaneous vertebroplasty when used to treat neoplastic destruction of the posterior wall of vertebra, may be aided by CT imaging. While CT offers some potential advantages for a limited number of situations, it is
32
Chapter 2 Materials Used in Image-Guided Spine Interventions
generally less available, more expensive, and slower than fluoroscopic imaging. Finally, the user of CT relinquishes the capability of real-time visualization of contrast or cement injection. These restrictions keep the use of CT limited to a small number of cases and complex anatomic situations.
Pharmacological Agents for Spine Intervention Corticosteroids Corticosteroids have a long history in the treatment of pain related to spine disease and have been used since the 1960s. At that time, they were injected both epidurally and intrathecally for pain management. By the 1980s, there were reports of complications that included arachnoiditis, meningitis, and paraparesis/paraplegia.1,2 Controversy was sufficient in Australia to prompt explicit government warnings about the use of corticosteroids for epidural pain management.3 Review of the scientific literature regarding these findings suggests that many of the complications resulted from or were associated with the intrathecal use of corticosteroids.1–5 It is known that some definite side effects can result from these drugs; physicians should be aware of these and should discuss potential complications with their patients. When used in spine injections, corticosteroids are believed to help produce chemical stabilization of the local environment. This is accomplished by reducing the local amount of phospholipase A2 and arachidonic acid as well as by decreasing the cell-mediated inflammatory and immunological responses. The most common corticosteroid used for spine injections has been a long-acting form of methylprednisolone acetate (Depo-Medrol®; New England Compound, Framingham, MA). This material is available in doses of both 40 and 80 mg/mL. The acetate formulation is quite insoluable in water and has a long half-life in tissues. Its relative strength is approximately five times that of hydrocortisone. It often contains the preservative polyethylene glycol, which is thought to be potentially neurotoxic. Indeed, this material may be the source of arachnoiditis created with intrathecal injection of Depo-Medrol®. Depo-Medrol® is particulate and therefore can cause stroke if injected intra-arterially (i.e., into the vertebral or radicular artery during the attempted cervical foraminal injection). Adding anesthetic solutions exacerbates the problem because the combination increases precipitation within the syringe. A more recent option for an injectable corticosteroid is the combination of betamethasone sodium phosphate and betamethasone acetate (Celestone Soluspan®, Schering Corp., Kenilworth, NJ). This mixes a short- and a long-acting form of betamethasone in the same injectable solution. It contains no preservative and comes in doses of 6 mg/mL. Betamethasone is approximately 30 times as strong as hydrocortisone. It seems to have a less particulate nature and a decreased tendency to precipitate when mixed with anesthetics (compared to Depo-Medrol®). All these properties make it less apt
Pharmacological Agents for Spine Intervention
to create arachnoiditis when injected intrathecally and less prone to create stroke if given intra-arterially. In the past, both Depo-Medrol® and Celestone® have gone through periods of decreased availability. This has caused some labs to use a long-acting form of triamcinolone (Aristocort®, Astellas Pharma US, Deerfield, IL).6 This material is particulate (similar to Depo-Medrol®) and also contains the preservative polyethylene glycol. It is available in doses of 25 and 40 mg/ mL (depending on the supplier) and is approximately five times as strong as hydrocortisone. It is long acting with similar properties to Depo-Medrol®. It seems to offer no advantage over Depo-Medrol®. Recently, some physicians have begun to substitute dexamethosone as the injectable steroid when there is concern of possible intra-arterial injection (as in the cervical region). This material can be given intravenously and may offer less risk of particulate risk of embolization intra-arterially. However, this protection is empiric and increased safety is not proven for this application. Anesthetic Agents Local anesthetic agents are commonly added as part of the injectate in numerous spinal and pain management injection procedures. Local anesthetics block the sodium channel, completely halting electrical impulse conduction in peripheral nerves, spinal roots, and autonomic ganglia.7 To block nerve conduction, the local anesthetic must cover at least three consecutive sodium channels (nodes of Ranvier). Differential blocking occurs because fibers carrying different types of information (pain, sensory, motor) are of different size. The smallest of these are the nociceptive (pain) fibers. These fibers attain calcium channel blockade with the smallest amount of anesthetic. Progressively, larger fibers require a larger volume of anesthetic to block enough adjacent channels to stop conduction. Pain fibers are the most sensitive, followed by sensory fibers and finally motor fibers. This differential blocking allows pain relief without obligatory motor blockade. Local anesthetics are organic amines with an intermediary ester or amide linkage separating the lipophilic ringed head from the hydrophilic hydrocarbon tail. The amino ester group of anesthetics includes procaine, tetracaine, and benzocaine. These anesthetics have been used for a long time and are known to have a higher allergic potential than the amide-linked group of anesthetics (lidocaine, bupivacaine, and ropivacaine) now in common usage. The amino ester group is thought to have its allergic potential because of its metabolite p-aminobenzoic acid (PABA). The members of the amide group, which do not have this metabolite, are known to have a very low allergic potential and little cross-reactivity. However, the amide group may contain the preservative methylparaben, which is metabolized to PABA and can produce cross-reactivity for potential allergic reactions with the ester group. Preservative-free amide anesthetics, therefore, are recommended for all injection procedures. Lidocaine is a common first-generation member of the amide anesthetic group. It was found safe except in large quantities that generally
33
34
Chapter 2 Materials Used in Image-Guided Spine Interventions
exceeded 500 mg. It has a relatively short duration of action, usually lasting only several hours. Bupivacaine is a second-generation amide anesthetic that has a prolonged duration of action. It is, however, associated with more cardiac and neurotoxic reactions, and has a maximum recommended safe dose of 150 mg. Because of the poorer cardiac profile of bupivacaine, third-generation amide anesthetics were developed. Ropivacaine is a member of this group that produces long-term local anesthesia like bupivacaine but with a better cardiac profile, allowing larger ultimate dosage without cardiac or neurotoxity. Injections of local anesthetic are small enough that one should generally never approach the maximum allowable dosages. Bupivacaine and ropivacaine come in different concentrations (0.25, 0.5, and 0.75% or 0.2 and 0.5%, respectively). The lower dosages are useful for pain relief in epidural and nerve blockage injections. The more concentrated dosages will produce motor blockade in small quantities, and this is not desired with these procedures. Antibiotics Antibiotics are needed for only selected procedures in spine intervention. These include discography, intradiscal electrothermal therapy, percutaneous discectomy, vertebroplasty, kyphoplasty, and the implantation of pumps and stimulators. Most injection procedures do not require antibiotics. The purpose of antibiotic coverage in most of these procedures is to decrease the chance of seeding bacteria into poorly vascularized sites such as the disc or around foreign bodies (implantables). Since penicillin allergy is not uncommon, a broad-spectrum antibiotic with minimal or no penicillin cross-reactivity is generally chosen. Though some penicillin cross-reactivity with the cephalosporins exists, it is minimal, and therefore a reasonable choice is cefazolin (Ancef®, GlaxoSmithKline, Brentford, London). This is the most common antibiotic used for this purpose and is given in a 1 g dosage intravenously or intramuscularly (IV or IM) 30 min prior to the procedure. Additionally, it can be put into the contrast for discographic procedures (usually 20–100 mg, with the upper range used when no IV antibiotics are given). It must be borne in mind that this antibiotic will cause grand mal seizure activity if given intrathecally. No Ancef should be injected (in the contrast solution) if a transdural approach is employed. An alternative to Ancef is Centamicin which is not neurotoxic. It can be added to the contrast solution at a dosage of 1 mg/ml. In some patients, allergy or lack of access to an IV hookup may make alternate choices more desirable. Another commonly utilized antibiotic in the interventional lab is ciprofloxacin (Cipro®, Bayer Pharmaceuticals, West Haven, CT). This is a fluoroquinolone with a broad spectrum of coverage and without cross-reactivity to penicillin. It is usually given orally in dosages of 500 mg twice a day. It can be given intravenously (400 mg), but it must be given slowly over a 60-min period to avoid pain and IV site reaction. This generally limits its use to oral administration. Another good alternative is levofloxacin (Levaquin®, Ortho-McNeilJanssen Pharmaceuticals, Inc., Titusville, NJ), a fluorinated carboxyquinolone. It may be given orally or intravenously, and has similar plasma and time profiles for both, making it a good choice for either route.
Pharmacological Agents for Spine Intervention
Again, slow administration is required for IV use. The general dosage is 500 mg every 24 h. Analgesics Conscious sedation sometimes is needed for procedures in the realm of image-guided spine pain management (e.g., percutaneous vertebroplasty) and works fine while the patient is on the table. However, some procedures, are frankly painful (e.g., discography) and may be associated with a postprocedural pain flare-up. If persistent pain occurs, one may need to prescribe analgesics appropriate for the patient’s pain level and suspected duration. This will not usually take the form of a long-term or chronic analgesic administration. The two mainstays for postprocedural pain management are opioids or nonsteroidal antiinflammatory (NSAID) drugs (and combination agents that contain drugs of both types). Mild to intermediate pain may be handled by the use of NSAIDs alone or in combination with a weak opioid (codeine, hydrocodone, dihydrocodeine, oxycondone). Controlled trials show little difference in efficacy of the NSAID category, and, therefore, finding one that works will usually be sufficient. There is potential toxicity from the NSAIDs to the gastrointestinal, genitourinary, central nervous, and hematological systems. Consider avoiding NSAIDs in patients predisposed to developing gastropathy or bleeding diathesis. Ketoralac (Toradol®, Hoffman-La Roche, Nutley, NJ) is very effective for shortterm use in intermediate pain relief.8 It is recommended only for shortterm use and should be administered with an initial IV or IM loading dose given prior to oral dosing. Multidose (IV or IM) administration recommended for patients less than 65 years old is 30 mg every 6 h, not to exceed 120 mg/day. For patients over 65, renally impaired patients, and those weighing less than 50 kg, the dosage is 15 mg every 6 h, not to exceed 60 mg/day. If there is breakthrough pain, one should not increase the NSAID dosage but add additional analgesic coverage. Regular, rather than intermittent, therapy promotes both anti-inflammatory and analgesic effects. Intermediate pain is often managed with the weaker opioids such as codeine, hydrocodone, dihydrocodeine, or oxycodone. These drugs are usually formulated as combination products and are weak only insofar as the inclusion of aspirin, acetaminophen, or ibuprofen results in a ceiling dose above which the incidence of toxicity increases. Prescribed alone, some of these drugs can manage even severe pain. Codeine is emetic and is prescribed much less than in the past. Hydrocodone preparations (Vicodin®, Abbott Labs, Abbott Park, IL; LorTab™, UCB Inc., Smyrna, GA) are now more commonly used. The potency is between that of codeine and oxycodone. Hydrocodone is not available as a single entity preparation. Oxycodone, now available as a combination product (e.g., Percocet® , Percodan®, Endo Pharmaceuticals, Chadds Ford, PA) as well as a single-entity preparation (e.g., Roxicodone®, Xanodyne Pharmaceuticals Inc., Newport KY), is very effective. It is also now available in a slow-release formulation (OxyContin®, Purdue Pharma LP, Stamford, CT) that is very potent.
35
36
Chapter 2 Materials Used in Image-Guided Spine Interventions
The most potent opioids are reserved for severe pain (e.g., the intractable pain associated with cancer). The members of this group include morphine, controlled-release morphine (MS Contin®, Purdue Pharma LP, Stamford, CT), hydromorphone (Dilaudid®, Purdue Pharma LP, Stamford, CT), meperidine (Demerol®, Sanofi Pharmaceuticals, Inc., NY, NY), and methadone (Dolophine®, Roxane Laboratories, Columbus, OH). Oxycodone also falls somewhat within this category when used as a single-entity preparation. Adjuvant Analgesics Classic pain is usually well handled by one of the NSAIDs, an opioid, or a combination product. These analgesics effectively deal with pain resulting from classic nociceptor response to intense, potentially tissuedamaging stimuli. However, neuropathic pain results from spontaneous discharge of injured nerves. It may be enhanced by sympathetic afferent activity as well. This type of pain is not as easy to control with standard analgesics; successful treatment has been achieved by means of adjuvant drugs such as antidepressants and anticonvulsants. When neuropathic pain is described as burning and constant, the tricyclic antidepressants become the first line of therapy. Syndromes such as postherpetic neuralgia and phantom limb pain are examples. Amitriptilyne (Elavil®, AstraZeneca, Wilmington, DE) is the most widely studied drug used for this type of dyesthetic pain. The operative mechanism for antidepressant mediated analgesia is believed to be the increase in circulating pools of norepinephrine and serotonin created by reductions in the postsynaptic uptake of these neurotransmitters. The quantities of drug administered are well below what is needed to relieve depression and suggest a separate mechanism of action. When neuropathic pain is described as intermittent but sharp and lancinating, anticonvulsant drugs have been used with success and should be tried before the antidepressants. It is believed that they relieve pain by damping ectopic foci of electrical activity and spontaneous discharge from injured nerves. Though carbamazepine and phenytoin have been useful as adjuvant analgesics, gabapentin (Neurontin®, Pfizer, NY, NY) is a new anticonvulsant that has been found to be effective for neuropathic pain relief while avoiding most of the side effects found with the other anticonvulsants. These and other adjuvant analgesics should be used when neuropathic pain contributes to a patient’s discomfort. Radiographic Contrast Agents Always an area of potential controversy for the image-guided physician, the choice of an appropriate contrast agent is challenging. The main concern is related to allergic potential and use within the thecal sac. There is no method that completely avoids the potential for allergy. Premedication is indicated in all patients with known allergy or prior reaction. If that reaction was severe, then all methods should be used to avoid the use of iodinated contrast. Substitution of another type of material may be useful (e.g., gadolinium). Pretreatment should include
Pharmacological Agents for Spine Intervention
oral corticosteroids (prednisone, 50 mg, 13, 7, and 1 h before the procedure) and oral H1 and H2 blockers 1 h before the procedure (diphenhydramine, 50 mg; tagamet, 300 mg).9 Although allergic reaction to nonionic contrasts exists, it may be lower than the incidence found with the ionic media. Routine use of nonionic contrast (Isovue®, Bracco Diagnostics, Inc., Princeton, NJ; Omnipaque®, GE Healthcare, Waukesha, WI; Optiray®, Mallinckrodt Medical Inc., Hazelwood, MO) is effective and safe for facet and sacroiliac joint injections. However, when there is a chance of injection into the thecal sac (e.g., epidural steroid injections), an agent that is approved for intrathecal use is recommended (Isovue M® 200, Isovue M® 300, Bracco Diagnostics, Princeton, NJ). Neurolytic (Cytotoxic) Agents Chemical and thermal agents intended for neurolysis have been used for decades.10 Commonly used agents or procedures include absolute alcohol, phenol, cryoanalgesia, and radiofrequency lesions. These materials or methods are intended to create long-term or permanent damage. This must be taken into account when one is planning therapy and discussing the procedure with the patient. Absolute alcohol is commercially available as a 95% concentration. Its use at this concentration is very painful, and, therefore, substantial sedation or anesthesia is necessary during injection. Being hypobaric to Cerebrospinal Fluid (CSF), alcohol rises if injected into the thecal sac. When injected near the sympathetic chain, alcohol destroys the ganglion cells and blocks postganglionic fibers.11 Postinjection neuralgia, hypesthesia, or anesthesia can be the side effects of alcohol use. Phenol (carbolic acid), like alcohol, has been used extensively and for a long time.12 It is not available commercially as an injectable preparation, but it can be made by the hospital pharmacy. It has the advantage of causing much less local pain during injection than absolute alcohol. Phenol is usually prepared in concentrations of between 4 and 10% and is hyperbaric to CSF. It is not stable at room temperature. Phenol produces a shorter and less intense blockade than alcohol. Moller and colleagues estimated that 5% phenol was equivalent to 40% alcohol.13 In intractable pain, the analgesic effects of phenol and alcohol have been found to be equal.14 References 1. Bodduk B, Cherry D. Epidural corticosteroid agents for sciatica. Med J Aust 1985;143:402–406. 2. Dilke TFW, Burry HC, Grahame R. Extradural corticosteroid injection in management of lumbar nerve root compression. Br Med J 1973;2:635–637. 3. National Health and Medical Research Council. Epidural Use of Steroids in the Management of Back Pain. Canberra: Commonwealth of Australia, 1994. 4. Cicala RS, Turner R, Moran E, Henley R, Wong R. Methylprednisolone acetate does not cause inflammatory changes in the epidural space. Anesthesiology 1990;72:556–558.
37
38
Chapter 2 Materials Used in Image-Guided Spine Interventions 5. Delaney TJ, Rowlingson JC, Carron H, Butler A. Epidural steroids: effects on nerves and meninges. Anesth Analg 1980;58:610–614. 6. Abram SE. Epidural steroid injections for the treatment of lumbosacral radiculopathy. J Back Musculoskel Rehabil 1997;8:135–149. 7. De Jong RH. Local Anesthetics, 2nd ed. St. Louis, MO: CV Mosby, 1994. 8. Patt RB. Pain management. In Abram SE, Haddox DJ (eds): The Pain Clinic Manual, 2nd ed. Philadelphia: Lippincott Williams & Wilkins, 2000:293–351. 9. Pittman A, Castro M. Allergy and immunology. In Ahya SN, Flood K, Paranjothi S (eds): Washington Manual of Medical Therapeutics, 30th ed. Philadelphia: Lippincott Williams & Wilkins, 2001:241–255. 10. Swetlow GI. Paravertebral alcohol block in cardiac pain. Am Heart J 1926;1:393. 11. Merrick RL. Degeneration and recovery of autonomic neurons following alcohol block. Ann Surg 1941;113:298. 12. Putman TJ, Hampton OJ. A technique of injection into the Gasserian ganglion under roentgenographic control. Arch Neurol Psychiatr 1936;35:92–98. 13. Moller JE, Helweg J, Jacobson E. Histopathological lesions in the sciatic nerve of the rat following perineural application of phenol and alcohol solutions. Dan Med Bull 1969;16:116–119. 14. Wood KA. The use of phenol as a neurolytic agent: a review. Pain 1978;5:205–229.
3 Patient Evaluation and Criteria for Procedure Selection Harold J. Cordner
Introduction It may seem obvious, but the main objective in patient evaluation is to make a diagnosis of the pain disease state as well as to identify the source of the pain. Determining the source of the pain can be extremely challenging because of the vast number of structures that can generate pain. Pain can arise from bones, joints, muscles, ligaments, nerve structures, and/or alterations in vascular supply. In addition, pain has numerous etiologies, ranging from structural malalignment to somatoform disorders. Pain also may arise from several different pain generators, and make the picture more complex and difficult to diagnose and treat. The first step in determining the source of pain is to perform a thorough history and a physical examination, to be supplemented with appropriate diagnostic tests to make an accurate diagnosis. Only then can one take the second step: determining which tool to use to help the patient with pain. It is critical to understand that an accurate and correct diagnosis must be made prior to any treatment or intervention, otherwise these treatments will be doomed to failure. A multidisciplinary diagnostic effort by a trained team best serves patients suffering from chronic pain. After reaching a diagnosis, the team can then determine the best strategy to treat the underlying disease and the pain. When did the pain start? Where does it hurt? What sort of pain is it? How much does it hurt? The answers to these questions provide important clues about why a person is in pain. When taking the patient’s history, one must rely on the patient’s answers about the when, where, what, and how of the pain to shed light on the biological basis of most pain conditions. While performing the physical examination, one must rely on the patient to communicate where it hurts and how much it hurts, without any objective tests to measure pain. One must understand the interaction of various aspects of pain sufficiently to know when a patient may be malingering for financial or emotional
From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_3, © Springer Science + Business Media, LLC 2010
39
40
Chapter 3 Patient Evaluation and Criteria for Procedure Selection
Figure 3.1. Targets for pain treatment: TCAs tricyclicantidepressants; NMDA N-methyl-d-aspartate (Adapted by Peter S. Staats; reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
gain or to decide which tests may allow diagnosis of an underlying pain-generating condition or disease. Pain physicians must have an armamentarium of knowledge about painful conditions and about interventional skills and procedures they can perform. These are the tools a physician uses when treating a patient. Pain physicians also must understand all the tools in their toolbox and know when to apply them. These tools include medical management, physical medicine techniques, radiation and chemotherapeutic options, neuromodulation techniques (electrical stimulation and intraspinal infusion therapy), therapeutic neural blockade, anatomical procedures to fix structural abnormalities, and, of course, ablative techniques (Figure 3.1). If physicians offer only interventional techniques, patients will not receive the most comprehensive care. On the other hand, if physicians fail to offer interventional options, they are neglecting the most highly effective options. To minimize risk and discover the least invasive/ most successful treatment for a patient, one generally begins with the most conservative approaches (medical management, rehabilitation strategies, lifestyle changes, psychological approaches, and alternative strategies) and works up the continuum of complexity and risk to interventions such as spinal cord stimulation and intrathecal drug delivery with an implanted pump. Conservative therapies can offer pain control without the risks associated with invasive techniques. Conservative therapies, however, do not always work and are not permanent. When conservative therapies fail or the side effects of these therapies become intolerable, a physician should consider the use of an interventional technique (Figures 3.2–3.4).
Introduction
Figure 3.2. Chronic pain treatment continuum.
This text concentrates on the importance of interventional techniques in the management of pain. Although each chapter highlights indications, techniques, outcomes, and complications, it is important to recognize that interventional therapies are not the only options for patients with pain. Before considering interventional techniques, an accurate diagnosis must be made, and conservative therapies should be considered, if not exhausted. This chapter begins by reviewing the diagnostic tools that are invaluable in evaluating patients and identifying appropriate candidates for various therapeutic and palliative procedures: review of the patient’s medical history, a thorough physical examination, imaging studies, electrodiagnostic tests, laboratory tests, diagnostic blocks, and psychological evaluation. Finally, the criteria for procedure selection is discussed.
Figure 3.3. Neuropathic pain treatment algorithm (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. ImageGuided Spine Interventions. New York: Springer Science+Business Media, 2004).
41
42
Chapter 3 Patient Evaluation and Criteria for Procedure Selection
Figure 3.4. Neuropathic pain treatment algorithm. TENS transcutaneous electrical nerve stimulation (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
History and Physical Examination Reviewing a patient’s medical history and conducting a thorough physical examination provide healthcare practitioners with vital information for making diagnostic and treatment decisions. Most of the information about a patient’s medical history is gleaned simply by asking the patient and/or the patient’s family members pertinent questions. The asking part is easy. Knowing what to ask is harder and, of course, crucial. An astute pain physician will read between the lines and extract the information he or she needs from the patient. One can augment or confirm some aspects of the patient’s medical history by obtaining medical records and test results, or even by speaking directly with prior treating physicians or care givers. Recording and reviewing the patient’s medical history highlights what one should expect and check for during the physical examination. This activity also helps establish a productive patient–physician relationship by assuring the patient of the physician’s interest, which helps the physician gain the patient’s trust and confidence. By providing a clear picture of the patient’s functional status prior to the onset of pain, the history also will help define the treatment goal. History Gathering To obtain a patient’s medical history, the physician must be a good listener and must direct the questioning appropriately to reveal and/or confirm vital information. Asking patients in pain the right questions will provide a clear picture of the onset and progression of the pain as
History and Physical Examination
well as of the effect of the pain on each patient’s daily life. These questions must elicit the chronology of events leading up to the consultation and must cover psychosocial and behavioral factors that affect the pain and interfere with the achievement of treatment goals. Thus, it is important to find out whether the patient likes his or her job, or if the patient has had previous worker’s compensation injuries, or litigated motor vehicle injuries, especially if the person is on disability leave and/or is receiving worker’s compensation. It is also important to note the existence of pending litigation or other sources of secondary gain related to a patient’s condition. Questions about the biological aspects of the pain should reveal: Location Quality Intensity (measured on a scale) Time course and whether it is constant What exacerbates it What alleviates it Effect, if any, on functional status The clinician should review the results of any diagnostic tests or treatments for the pain (especially the efficacy, dose, frequency, and any side effect of pain medications and any psychological interventions) and gather information about the patient’s: General state of health Current medications Allergies (distinguishing between true drug allergies and transient adverse effects) Sleep patterns (as a sign of possible emotional depression) Consumption of tobacco, alcoholic beverages, illegal drugs, drugs prescribed to another person, and over-the-counter medications. The history also should include information on any of the patient’s childhood illnesses, physical and psychiatric adult illnesses, surgical procedures, major injuries, and hospitalizations that could affect the current pain problem. Perhaps, most importantly, the history of prior treatments or interventions and their effectiveness should be documented. Physical Examination The patient’s history and presenting complaint will help indicate what needs to be assessed during the physical examination. The physical examination also will help to formulate the differential diagnosis as well as to strengthen the doctor–patient relationship. There is no substitute for actually placing your hands on a patient and examining to help show interest, involvement, and competence. The examination should encompass both a thorough, focused exam on the involved area of pain as well as a complete physical examination. If a patient’s pain complaints change over the course of time, a repeat examination should be performed. Having the patient show where it hurts is important. Many times, patients will describe pain in their hip or back; but, when asked to show where it hurts, they point to their buttock. Often, patients will
43
44
Chapter 3 Patient Evaluation and Criteria for Procedure Selection
be referred for treatment of a thoracic vertebral fracture for back pain, and, when asked to show where it hurts, they will point to their lower lumbar spine. Trying to correlate the patient’s pain complaint with the physical examination is essential. In patients with unilateral limb pain, physicians should first examine the unaffected contralateral limb for comparison. The physician should inspect all areas where the patient feels pain for the presence of erythema, discoloration, abnormal nail growth, masses, induration, or scars. Light palpation of the painful area will reveal the presence of hyperalgesia. If the patient has symptoms of neuropathic pain, a thermal stimulus applied to the painful area will uncover thermal hyperalgesia. If the patient has a lesion, palpation will indicate the presence of a mass and palpation-induced pain. Testing for sensory and motor function and deep tendon reflexes will uncover any involvement of peripheral nerves or nerve roots. In patients with neck or low back pain, it is important to examine the spine and determine range of motion. One may be able to determine where the pain originates (e.g., in such areas as the hip, sacroiliac joint, or lumbar spine) by performing appropriate procedures and maneuvers. Physicians can test for the presence of three or more of Waddell’s signs (tenderness, simulation, distraction, regional disturbances, and overreaction) to determine whether or not low back pain is psychological in origin. The physical examination also provides an important opportunity to gauge the patient’s mood, affect, and degree of pain behavior. Unfortunately, even with an excellent history and an excellent physical examination, the source of the patient’s pain is not obvious. One must then choose diagnostic tests – which may encompass imaging studies, electrodiagnostic tests, laboratory studies, psychological evaluation, or diagnostic injections – to aid in determining the proper diagnosis or pain generator.
Imaging Studies Imaging studies are crucial for identifying anatomical abnormalities that corroborate physical findings. However, one must realize that imaging studies do not tell how much a structure or abnormality hurts. Findings that may look painful may not hurt at all. And vice versa: Findings that may not look terribly painful may be quite painful. Understanding the limitations and usefulness of various imaging studies is vital so that a physician can order the appropriate test and truly evaluate the results. Conventional Radiographs Conventional radiographs are particularly helpful in diagnosing the cause of skeletal pain in the back, neck pain, and pain in the limbs and/ or joints for the following reasons: They can indicate whether a bone is healing and aligning properly They can indicate whether a patient has osteomyelitis or osteoporosis They can reveal the coexistence of a pathological fracture and a destructive bone lesion They can reveal the size and shape of primary bone tumors.
Imaging Studies
Radiography is an extremely precise way to diagnose various arthritic disorders. Rheumatoid arthritis of the hands usually involves the metacarpophalangeal joints, and a radiograph can reveal an incriminating narrowing of the joint space as well as articular surface erosions. Radiographs also reveal arthritic osteophytes (bony outgrowths) and sclerosis (scarring). Additional reasons for spine pain exposed by radiography include: Spondylolisthesis (when one vertebra has slipped over another) Narrowing of disc space Kyphosis (“widow’s hump”) Scoliosis (abnormal curvature of the spine) Osteoporosis Hypertrophic spurs Failed spinal fusions Spondylosis (degeneration of one or more vertebrae) Pars interarticularis defects (a break in the posterior elements of the spine) Zygapophyseal (facet) joint abnormalities One also can use oblique X-rays to expose the neural foramina and flexion/extension views to assess spinal stability. Because this diagnostic tool is noninvasive, most people with chronic pain accept it readily. Myelography Myelography may be used to confirm a diagnosis of a surgically correctable lesion, such as a herniated disk or nerve root impingement, and to pinpoint its exact location. It is less commonly used today, but it is still helpful when primary screening with magnetic resonance imaging (MRI) fails or cannot be used (as is the case when a pacemaker is present). Computed Tomography Scanning Computed tomography (CT) scans are used to evaluate the bony structures and soft tissues of the spine. CT scan is especially good for evaluating bony structures and abnormalities, whereas MRI can better image soft tissue structures. Laterally placed fragments of herniated disc, for example, may be visible on a CT scan but missed on a myelogram. A CT scan provides important additional information when a herniated disc causes radicular pain by compressing a nerve root exiting through its neural foramen. Images of facet joints obtained by CT will reveal the degenerative and/or hypertrophic origin of chronic spinal pain, and axial CT scans provide three-dimensional images of spinal ligaments and discs. CT scans are particularly good for evaluating SI joint abnormalities and more detailed images of discogram injections. Magnetic Resonance Imaging Presently, as the single most important imaging tool for spine pathology, MRI provides a detailed image of the spinal cord, cerebrospinal fluid,
45
46
Chapter 3 Patient Evaluation and Criteria for Procedure Selection
extradural structures (intervertebral discs), and the patency of neural foramina. An MRI reveals: Disc degeneration Herniated discs Facet joint arthropathy Vertebra or disc infection Subluxation Stenosis Fracture Neoplasm Edema Vascular abnormalities. It remains the most accurate study for determining the acuity of vertebral fractures because bone scanning may show increased uptake for up to a year. However, as previously discussed, it cannot be emphasized enough that MRI findings do not tell how much a structure hurts. Disc degeneration or protrusions may not be painful, and assuming the source of pain may lead to improper diagnosis and treatment. Spinal fusion may be performed on a patient with presumed discogenic pain based on MRI findings, when, in fact, they may have had facet joint pain that was not properly diagnosed. It is the job of the physician to use the proper tools, which may include diagnostic blocks or injections, to correctly diagnose the source of pain so the appropriate treatment may be recommended. The practitioner should personally review MRI films whenever possible rather than relying on the radiological report. Just as it is hard to describe a painting in words, it is hard to envision an MRI image by reading a radiological report. Because of the incorrect use of terminology, conditions such as protrusions, herniations, extrusions, and displacements may be inaccurately described and, thus, inaccurately interpreted by the practitioner. Radiologic reports may not give the degree of stenosis or degeneration, or may not describe high intensity zones in the annulus. It is important that the practitioner becomes adept at interpreting MRIs and at correlating the history and physical exam with the MRI findings. Ultrasound Although virtually useless for evaluating musculoskeletal pain, ultrasound is the best way to evaluate suspected gall bladder disease in patients with abdominal pain. Bone Scanning Bone scanning permits detection of the early stages and the course of bone metastasis, osteomyelitis, bone trauma, arthritis, hairline fractures, and all other diseases that involve bone turnover and that can easily be missed by conventional radiography. Bone scan can be used to determine if a vertebral fracture is acute or subacute, or in the diagnosis of loose or infected orthopedic prostheses or hardware.
Electrodiagnostics
Because bone scanning is nonspecific, however, diagnoses based on bone scans must be generally supported by appropriate clinical information and other imaging studies. Thermography Thermography may help diagnose neuromuscular and soft tissue disorders, especially in patients whose abnormalities elude detection during a physical examination. Some clinicians use thermography to evaluate: Neuropathic syndromes (e.g., complex regional pain syndrome, radicular syndromes, peripheral neuropathies, carpal tunnel syndrome and other nerve entrapments, postherpetic neuralgia, thoracic outlet syndrome, and trigeminal neuralgia) Myofascial syndromes (e.g., fibromyalgia and lumbosacral strain) Circulatory disorders (e.g., peripheral vascular occlusive disease, vasospastic disease, and venous insufficiency) Skeletal disorders (e.g., osteomyelitis, lumbar facet syndrome, rheumatoid arthritis, scoliosis, and postfracture extremity pain) Clinicians have not yet agreed upon the clinical applicability of thermography, and its use remains controversial and limited.
Electrodiagnostics Electrodiagnostic studies provide information on how well nerve roots, peripheral nerves, and muscles are functioning. These diagnostic tools thus provide important information in suspected cases of nerve entrapment, radiculopathy, and peripheral neuropathy, to name a few. Specially trained physicians generally perform and interpret electrodiagnostic studies. Electromyography Electrical potentials become abnormal in the presence of a diseased muscle or nerve serving a muscle. To discern the presence of abnormal potentials, one can record changes in intermuscular voltage on an electromyelograph. Fibrillation potentials and positive sharp waves often occur simultaneously in the presence of radiculopathy and peripheral neuropathies, such as a diseased nerve plexus or degeneration of nerve axons, which cause muscle fibers to lose their normal innervation and undergo spontaneous depolarization. Because it can occur in healthy individuals, fasciculation must accompany fibrillation potentials and positive sharp waves to contribute to a diagnosis of neuropathic disease. In the early stages of neural injury, however, neural conduction velocity testing is more sensitive than electromyography (EMG) because EMG changes occur slowly over a period of week. So a normal EMG in the early stage of injury does not rule out the presence of a nerve injury or neuropathy.
47
48
Chapter 3 Patient Evaluation and Criteria for Procedure Selection
Nerve Conduction Studies Nerve conduction studies (NCVs), which use surface electrodes to stimulate a peripheral nerve and evaluate how well it is functioning, may show the abnormal nerve conduction that occurs during neuropathy as well as the location of a nerve lesion and/or nerve entrapment. To perform motor NCVs, one stimulates a nerve to record a target distal muscle’s evoked response (impulse velocity, amplitude, and latency). To perform sensory NCVs, one places both a stimulating and a recording electrode over the target sensory nerve. Sensory NCVs may reveal peripheral neuropathies before a patient experiences significant sensory loss. However, because of the relatively frequent false positive and false negative or inconclusive results, these test should be interpreted accordingly. It is also important to realize that these studies are designed to assess normal, large nerve fiber function, such as A-beta function. The pain fibers, A-delta and C-fibers, are small, slowly conducting fibers. The presence or absence of an EMG or nerve conduction velocity abnormality does confirm or rule out the presence or absence of neuropathic pain. Small nerve fibers are not being measured with EMGs; large nerve fibers are being measured. There are some tools that are less commonly used, including Quantitative Sensory Testing (QST) that can measure the functioning of smaller pain fibers. Regrettably, insurers and even some clinicians have used normal EMG/NCV studies as evidence to say that the person is not experiencing pain. In this respect, there are limitations of electrodiagnostic testing.
Laboratory Tests Laboratory tests can uncover abnormalities associated with many of the neurological diseases that present with pain. Obvious uses of laboratory tests include screening for diabetes, malnutrition, toxins, dysproteinemia, cancer, and the thyroid disorders that can cause compression neuropathies. One can also detect abnormal inflammatory states or autoimmune dysfunction by checking a patient’s erythrocyte sedimentation rate or levels of antinuclear antibodies.
Diagnostic Nerve Blocks Diagnostic blocks, perhaps one of the most valuable tools to the practitioner, help determine whether a particular nerve or structure is causing pain or is the pain generator. Despite conducting an excellent history and an excellent physical examination and evaluating appropriate tests, the exact source of pain still may be uncertain. A correctly performed diagnostic block should help identify the painful structure. To perform a diagnostic block, one injects a local anesthetic around a presumed pain-generating structure. The diagnosis depends upon whether this leads to pain relief. There are many variables to consider in the interpretation of the results of blocks. False positive results, for example, can be due to a placebo response or to the effect of systemically administered analgesics or a systemic
Diagnostic Nerve Blocks
uptake of local anesthetics. A second, confirmatory block often is used or recommended to reduce the placebo effect. Systemic analgesics and anesthetics should not be given in the peri-block period, so that, if the patient reports relief of the pain after the block, it can be assured that the local anesthetic caused analgesia, not related to systemically administered agents. It is also inappropriate to decide that because a patient has responded to a placebo injection, the person’s pain is psychogenic. Diagnostic blocks may include intra-articular or periarticular joint injections, myofascial injections, peripheral nerve blocks, or central nerve blocks. The patient’s response to a block helps diagnose specific joint pain such as hip joint, SI joint, or cervical or lumbar facet joint syndrome. Pain arising from the C2–C3 facet joints generally radiates to the occiput, and pain arising from C5–C6 radiates to the shoulder. One can reproduce this pain with ipsilateral rotation and extension of the cervical spine. Lumbar facet joint syndrome causes constant pain in the lumbar region that may radiate to the hips or even below the knee, and can be elicited by hyperextending the spine ipsilaterally. Facet joint syndrome is difficult to diagnose because it arises from the same types of degenerative change that show up in x-ray images of asymptomatic joints. The diagnosis is further obfuscated because similar symptoms can arise from discopathy, spinal stenosis, nerve root impingement, and/or myofascial disease. One can differentiate facet joint syndrome by the response to radiographically guided injections of local anesthetics into the zygapophyseal joints or around the dorsal medial branches of the posterior primary rami. Peripheral Nerve Blocks To determine whether peripheral nerves are the source of the pain, local anesthetics are injected around a nerve and the response is assessed. A report of a marked reduction in pain indicates that the pain is coming from a location distal to that nerve. (One must be mindful that the test can produce false positive results.) Pain relief after blockade of a peripheral nerve may indicate a peripheral nerve injury or entrapment. A fluoroscopically directed selective nerve root block may help determine if a particular nerve root is causing pain. This may be helpful especially when MRI indicates several levels of nerve impingement, if EMG/NCV tests are inconclusive, or if the diagnosis is unclear. If the block results in pain relief, one presumes that the pain generator is at or distal to the anesthetized site. If the block results in numbness but no pain relief, one presumes that the pain generator is proximal or collateral to the anesthetized site. To block a sympathetic nerve, the local anesthetic is injected onto the sympathetic chain at various sites. The primary sympathetic ganglia involved in pain include the stellate ganglion, the celiac plexus, the lumbar sympathetic ganglion, the superior hypogastric plexus, and the ganglion impar. The stellate ganglion is blocked to diagnose sympathetically mediated pain of the upper thorax, arm, head, or face, and to treat postherpetic neuralgia, sympathetically maintained pain, or vasoocclusive disease. Celiac plexus blocks indicate whether pain is arising from the abdominal viscera and relieve pain caused by upper abdominal malignancies,
49
50
Chapter 3 Patient Evaluation and Criteria for Procedure Selection
including pancreatic cancer. A positive response to a celiac plexus diagnostic block is prognostic of several months of pain relief from celiac plexus neurolysis. Lumbar sympathetic ganglion blocks allows diagnosis of sympathetically mediated pain of the lower extremities. Superior hypogastric plexus blocks uncover any visceral cause of pelvic pain, and ganglion impar blocks shed light on the cause of perineal (rectal, anal, vaginal) pain. Central Nerve Blocks Differential epidural or intrathecal blocks can reveal whether pain is arising from the somatic nerves, the sympathetic nervous system, or the central nervous system. The first injection in a differential epidural is a placebo (saline). If this leads to pain relief, the clinicians halt the injections. If the placebo relief is long lasting, it is possible that the pain is centrally maintained or psychogenic. If the placebo provides no pain relief, three injections of successively higher concentrations of local anesthetic are administered. If the lowest concentration of anesthetic provides pain relief, one considers the pain to be sympathetically maintained. If the next level of anesthetic provides relief, one presumes that the pain is somatosensory. If the pain persists, the highest concentration is injected, which usually causes a temporary loss of motor function. If this fails to provide relief, one presumes that the pain is centrally maintained or psychogenic.
Psychological Evaluation Pain is, by definition, a sensory and emotional experience of actual or perceived tissue damage.1 The biologically oriented clinician may not recognize the impact of depression, anxiety, or other negative affective states on the experience of pain. The experience of pain always involves emotional dysfunction (Figure 3.5). The challenge for the pain
Figure 3.5. Pain treatment strategy (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Criteria for Procedure Selection
practitioner is to differentiate between the component that is biologically driven and the component that is magnified by emotions.2 Patients with severe depression or anxiety should be evaluated to determine the impact of these comorbid psychological states on their pain. This evaluation is an important part of a multidisciplinary medical approach to their pain and is essential before they receive interventional therapies. Most practitioners and some insurance carriers mandate a psychological evaluation prior to implementing any neuromodulation therapies. Patients with major depressed mood, anxiety, or other negative affective states report more pain with noxious stimuli than do controls with positive affective states. Emotionally depressed patients can be appropriate candidates for interventional therapies; it is simply necessary to be especially careful when offering them therapies that carry significant risks. While it may be obvious that patients with severe pain caused by a peripheral pain generator also will experience depression or anxiety, it is less obvious that the same negative affective states actually increase the experience of pain itself. Depressed affective states also can maintain pain and cause it to take on a life of its own by dramatically amplifying what would otherwise be a relatively minor pain generator. Frequently, a physician can determine the severity of emotional dysfunction during an initial encounter. If the patient reports anhedonia, depressed or increased appetite, a history of major depression, or difficulty sleeping, a physician should be alert to the possibility that depressed mood is an exacerbating component of the pain. Simple screening tests such as the Beck Depression Scale may be administered in the office. When a major depression or other psychological disease is suspected, it should be treated, prior to initiating interventional techniques, directly or by referral to a competent physician who can help with this aspect of pain.
Criteria for Procedure Selection While it is not in the scope of this chapter to discuss in detail the criteria for each interventional procedure, it is relevant to discuss the overall approach to the pain patient and consideration for procedure selection. It seems obvious that more conservative procedures and therapies should be initiated prior to more invasive and riskier procedures when appropriate. It is important that a systematic evaluation of the patient occurs, which involves history, physical exam, and the appropriate diagnostic tests outlined previously. Only at that point, once a diagnosis or differential diagnosis has been made, should a procedure be contemplated or performed. Of course, some procedures are performed as a diagnostic procedure, and it is appropriate to perform them as part of the workup. However, a therapeutic intervention or procedure should be performed only after the diagnosis has been made or confirmed. In benign axial back pain, for example, there are several causes or pain generators. Facet joints, intervertebral discs, sacroiliac joints, vertebral bodies, muscles, and ligaments are the typical structures that can cause mechanical low back pain. Depending on the individual patient’s age, symptoms, physical examination, and diagnostic tests, the most appropriate procedure should be performed to help confirm the diagnosis
51
52
Chapter 3 Patient Evaluation and Criteria for Procedure Selection
of the specific pain generator or structure. If the patient is elderly with symptoms of achy, low back pain with insidious onset, no radicular symptoms, evidence of degenerative facet and disc disease, and has failed physical therapy, facet blocks would be most appropriate before discography or spinal cord stimulator is contemplated. If an elderly patient presents with acute history of fall, with severe back pain, and with an acute vertebral compression fracture, it will be appropriate to first consider conservative management or vertebroplasty rather than facet injections. Of course, not all patients present with such clear history and findings, but the thought process or logic should be the same. If the primary diagnosis or pain generator was ruled out by a diagnostic test, then the next most likely diagnosis or pain generator should be evaluated. An algorithmic approach to interventional pain management is a way to assist the physician in clinical practice. Using evidence-based medicine, algorithms and practice guidelines may be established. Probably, the most extensive interventional pain management guidelines utilizing an evidence-based research model are published in Pain Physician3 and are updated regularly and remain a good resource for reference: “An algorithmic approach was developed based on the structural basis of spinal pain, and incorporated acceptable evidence of diagnostic and therapeutic interventional techniques available in managing chronic spinal pain.”3 Figure 3.6 describes a proposed algorithmic approach
Figure 3.6. An algorithmic approach to diagnosis of chronic low back pain without disc herniation (Reprinted with permission from Boswell MV, Trescot AM, Datta S et al. Interventional Techniques: EvidenceBased Practice Guidelines in the Management of Chronic Spinal Pain. Pain Physician 2007; 10: 7–111).
Criteria for Procedure Selection
53
Figure 3.7. A suggested algorithm for therapeutic interventional techniques in management of chronic low back pain (Reprinted with permission from Boswell MV, Trescot AM, Datta S et al. Interventional Techniques: Evidence-Based Practice Guidelines in the Management of Chronic Spinal Pain. Pain Physician 2007; 10: 7–111).
for the diagnosis of chronic low back pain and Figure 3.7 describes an algorithmic approach to management of chronic low back pain. Figure 3.8 describes a proposed algorithmic approach for diagnosis and management of chronic neck pain. With the advances in minimally invasive techniques, patients and physicians have more options in treating spinal and nonspinal pain. Less invasive procedures such as intradiscal therapies and percutaneous disc decompression or discectomy are being used increasingly before more invasive and riskier laminectomies or fusions are performed. By avoiding riskier procedures, the complications associated with the riskier procedure and its compounding a chronic pain problem are avoided. Likewise, treating lumbar facet disease with radiofrequency facet ablation may have better long-term outcomes when compared with treating the patient with opiates on a long-term basis or a lumbar fusion, which have a higher potential risk and financial cost. Similar criteria should be used when contemplating neuromodulation or neurodestructive procedures. Spinal cord stimulation should be considered after more conservative or less invasive therapies have failed. In the United States, spinal cord stimulation is most often used to treat patients with failed back surgery syndrome with radiculopathy. Techniques using multiple leads and contact arrays and more advanced hardware have been developed to treat patients with low back pain more successfully than previously possible. Many new indications for neurostimulation are being treated such as refractory angina, peripheral vascular disease, CRPS, movement disorders, deep brain stimulation, vagus nerve stimulation, and peripheral nerve stimulation, among others.
54
Chapter 3 Patient Evaluation and Criteria for Procedure Selection
Figure 3.8. An algorithmic approach to diagnosis of chronic neck pain without disc herniation. “a” means transforaminal epidural injections have been associated with reports of risk. “b” means not based on evidence synthesis (Reprinted with permission from Boswell MV, Trescot AM, Datta S et al. Interventional Techniques: Evidence-Based Practice Guidelines in the Management of Chronic Spinal Pain. Pain Physician 2007; 10: 7–111).
Intrathecal administration of medications also has advanced. New medications such as clonidine, local anesthetics, opiates, and ziconitide have helped in treating more patients successfully. However, one must be cautious and remember that there are significant long-term risks associated with intrathecal infusions not limited to granuloma formation, physical dependence, hormone depletion, and drug side effects. The age of the patient, psychological status, comorbid medical conditions, and diagnosis must be considered before contemplating or implementing intrathecal therapies. One rarely deliberately destroys primary motor or mixed motor/ sensory nerves. Often, radiofrequency lesioning techniques of the spine, however, are used to treat known facet disease with facet rhizolysis. This technique has an extremely high success rate and very low complications. Generally, use of other neurodestructive techniques in the spine are delayed until all conservative therapies have failed. This is particularly true for patients who have long life expectancies. However, celiac plexus, sympathetic, or peripheral neurolysis may
Conclusion
be extremely effective in patients with malignant pain with short life expectancies. Keep in mind though, that there are still significant risks associated with these neurodestructive procedures. It is this author’s opinion that neuromodulation should be performed before a neurodestructive procedure whenever possible.
Conclusion Pain management is like any other medical specialty in that one must first make a correct diagnosis before initiating or prescribing any treatment. While the focus of this book is on interventional procedures for pain management, it is vital that, before any therapeutic procedures are performed, the clinician must be confident that the treatment is appropriate for that patient’s condition or diagnosis. Some of the interventional procedures are diagnostic, as with facet joint injections, and need to be performed in order to establish the correct diagnosis. In order to establish a correct diagnosis, the practitioner must perform a thorough history and physical examination, and use appropriate diagnostic tests or procedures. A complete understanding of the disease process, the proper diagnostic tools that should be used, and the proper interpretation of those results are the essential first steps in the practice of pain management. References 1. International Association for the Study of Pain. Pain terms: a list with definitions and notes on usage. Pain 1979;6:249; 1982;14:205. 2. Staats P, Hekmat H, Staats AW. Psychological behaviorism theory of pain: a basis for unity. Pain Forum 1996;5:194–207. 3. Boswell MV, Trescot AM, Datta S et al. Interventional techniques: evidencebased practice guidelines in the management of chronic spinal pain. Pain Physician 2007;10:7–111.
55
4 The Surgeon’s Perspective: Image-Guided Therapy and Its Relationship to Conventional Surgical Management F. Todd Wetzel Introduction The indications for surgical management of compressive syndromes such as herniated nucleus pulposus with radiculopathy and lumbar spinal stenosis with neuroclaudication are clear, and outcomes are predictable. For example, in the case of a herniated nucleus pulposus with unilateral radiculopathy, assuming strict concordance between the patient’s clinical presentation and imaging findings, the likelihood of successful outcome is 90–100% following laminotomy discectomy.1 Likewise, the addition of arthrodesis to treat degenerative spondylolisthesis in the setting of stenosis with neuroclaudication has been shown to be the treatment of choice, based on randomized prospective data.2 Outcome data for surgical treatment of low back pain (LBP) per se in the absence of neural compression and referred extremity pain is, however, less promising.3,4 Lumbar discography, for example, has been cited as a reasonable diagnostic technique to identify painful segments and treat them by arthrodesis.3 Not only have the sensitivity and specificity of the diagnostic technique been criticized,5–7 but the surgical treatment itself – arthrodesis – has been studied in only two randomized, prospective studies to date.8,9 While the data of Fritzell and colleagues8 demonstrate significant clinical advantages of fusion over (unspecified) nonoperative care for low back pain, the study suffers from several significant flaws, including imprecise patient selection, heterogeneity of treatments, and low absolute rates of improvement (for LBP, 45% good or excellent in the surgical group versus 18% in the nonoperative group, at 2 years). Even the statistically significant differences reported by Fritzell and colleagues8 are not replicated in a more recent study from Brox and coworkers.9 Sixty-four patients with LBP were randomized to a surgical group consisting of instrumented fusion and postoperative physical therapy or to cognitive intervention and exercises. At 1 year, the outcomes
From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_4, © Springer Science + Business Media, LLC 2010
57
58
Chapter 4 The Surgeon’s Perspective: Image-Guided Therapy
were no different between groups, with a 70% success rate after surgery and a 76% success rate after cognitive intervention and exercise.9 From a surgical perspective, precise, reliable, sensitive, and specific diagnostic techniques are required to identify a lesion amenable to surgery at each step of a treatment algorithm such as that just described (e.g., persistent axial pain → discography → arthrodesis), and conflicting or variable reports of sensitivity, specificity, and efficacy further hinder the clinician in making rational decisions based on acceptable standards of care or scholarly consensus. Invariably, individual surgical philosophy plays a role in patient selection, with some physicians more likely to recommend surgery for indications that would not be accepted by others.4,10 Furthermore, a surgical diagnosis may not represent a true surgical lesion. It is the unfortunate experience of many surgeons to have patients referred to them with a lumbar disc prolapse, apparent radiculopathy, and the expectation of a surgical recommendation, only to discover that the prolapse demonstrated on magnetic resonance imaging or computed tomographic myography is minimal, or that it is not precisely correlated with the patient’s symptoms. The surgical treatment of lumbar spinal stenosis serves as another example. In one report, patients with a higher degree of midsagittal stenosis, including complete myelographic block, had lower functional disability scores at follow-up of 4.5 years. Patients who had a midsagittal stenosis exceeding 12 mm had a poor outcome.11 While one surgeon may wish to treat a patient conservatively until such a critical threshold is reached, in order to maximize surgical outcome, another may offer an earlier decompression based on individual patient characteristics, experience, and expectations. This difference in philosophy may be further compounded by discrepancies in education between patient and physician, and thus, in their respective expectations. Another level of variance is added by pain-based diagnostics, which contain an unavoidable element of subjectivity. Obviously, patient expectations factor into this as well, with most surgeons more likely to offer surgical care to those who appear to have reasonable expectations. Unfortunately, there are no reproducible standards whereby patient expectations can be quantified,12–15 thus adding another layer of individual idiosyncrasy. The goals of this chapter are to review, from a surgeon’s perspective, provocative diagnostic maneuvers, including discography, facet blockade, epidural steroid injections, selective nerve root blockade, and sacroiliac joint injection. Specifically, the results of these diagnostic maneuvers will be scrutinized for their predictive value with regard to current concepts in surgical treatment. Additionally, minimally invasive intradiscal therapy, vertebroplasty, and kyphoplasty will be reviewed.
Discography While the use of discography to diagnose spinal pain syndromes has increased, the practice is not free from controversy. Despite reports of its utility in clinical decision-making3 as well as reports of high sen-
Discography
sitivity and specificity16 (including one report of 100% sensitivity and specificity in distinguishing symptomatic from asymptomatic patients with back pain)17 discography is innately subjective and thus can never be completely controlled. This aspect of the procedure relates to the use of pain provocation, which must be concordant with presenting symptoms. As Saal18 notes, most pain-provocative or ablative tests used in the diagnosis of spinal conditions are closely related to the physical examination. In the case of “nonspecific” low back pain created by degenerative lumbar disc disease, the findings from a physical examination are not as clearly defined as those involved in radicular syndromes; or are they? Recent attention to the classification of diagnostic subgroups of LBP and application of the AssessmentDiagnosis-Treatment-Outcome (ADTO) research model of Spratt19 may influence the selection of patients for discography. This model focuses on the diagnostic aspect of “nonspecific” LBP in particular. Without, for instance, the definition of a diagnostic subgroup such as discogenic pain, the outcome data from this essentially linear process would not be valid. Numerous authors have noted the association of certain physical signs with discogenic pain.20–29 Donelson and colleagues noted an increased incidence of positive concordant discography in patients who failed to centralize30 according to the criteria of McKenzie.31 In a specific subset, discography was more likely to be clinically positive (concordant) in patients with annular incompetence. The clinical correlate of this was failure to centralize. While this is certainly encouraging, it is by no means definitive. One of the key issues in determining the sensitivity and specificity of a particular test is comparison to a recognized and accepted standard of accuracy, a “gold standard.” This further complicates the situation for discography, and for invasive spinal diagnostics in general: since a painful joint or disc may have a variable or wide range of anatomical and clinical features that overlap with an asymptomatic structure, a “gold standard” is difficult to define.18 Furthermore, there is no reliable way to confirm or refute the symptomatic status of a noncompressive degenerative lumbar disc (painful versus painless) at surgery. In an attempt to address false positive findings of lumbar discography, Carragee and colleagues6 studied eight subjects with a total of 24 discograms. None of the patients had a history of low back pain. Patients were scheduled to undergo posterior iliac crest grafting for nonthoracolumbar procedures; 2–4 months after the bone graft, patients underwent lumbar discography by a blinded protocol. Fourteen of 24 discs were painful, with 2 (14.3%) reproducing the pain “exactly.” In this case, the pain was referred to the iliac crest bone graft. Based on these results, Carragee and colleagues6 concluded that the ability of a patient to separate spinal from nonspinal sources of pain may be questionable. This study is important for several reasons. First, it suggests that discography done under blinded conditions, in accordance with accepted protocol, may in fact not be specific to spinal pathology. Second, it suggests that in patients who were free from other potentially confounding influences (all patients passed a standardized psychometric screening battery prior to the test), significant
59
60
Chapter 4 The Surgeon’s Perspective: Image-Guided Therapy
pain can be produced in a clinically irrelevant setting. Since in order to be graded positive; pain must be concordant; by definition all those patients in whom no spinal pain source was being evaluated would have had discordant spinal pain. However, many patients undergoing spine procedures have had iliac crest bone graft harvest. This serves to underscore the possibility that, in patients with previous surgery, the findings from discography may be complicated by a confounding variable: a potential pain generator in close anatomical if not physiological proximity. The technique of discography is of interest as well. While the double needle technique and multiple blinded injections have become the standard of care, the utility of pressure-controlled discography remains unclear. In a multicenter retrospective study of long-term surgical and nonsurgical outcomes, Derby and coworkers32 reviewed 96 patients who underwent first discography and then fusion, or continued nonoperative care. These investigators noted no long-term differences in surgical outcome across the entire sample, with the surgical group as a whole doing better than the patients who did not have surgery. In a specific subset, the data suggested that patients with highly pressuresensitive discs appeared to achieve better long-term outcomes with interbody or circumferential fusion than with intertransverse fusion. For this reason, the authors suggested that there may be a biochemical component to discogenic pain. These results, however, have not been corroborated in prospective studies. Another requirement for successful discography is the study of a large enough number of discs to permit inclusion of a rostral and, possibly, caudal control. Given that many surgeons have empirically limited arthrodesis to two- or three-level disease in the lumbar spine, the presence of appropriate control levels is critical. A final note of concern also must be added regarding surgical treatment for discogenic pain. Whitecloud and Seago33 reported a 70% rate of clinical success for cervical arthrodesis on the basis of discography. While Wood and colleagues34 have noted that thoracic discography may differentiate between symptomatic and asymptomatic degenerated discs (as characterized by the presence of Schmorl’s nodes), the optimal surgical treatment of thoracic discogenic pain remains to be identified. In the lumbar spine, a wide variety of success rates has been reported. In one study, an overall success rate of 46% was identified, with a clinical success rate of 96% in the subset that fused solidly.3 Disc replacement may be another option, which is potentially attractive due to preservation of more physiologic kinematic behavior at adjacent levels.35–38 Clearly, based on data collected to date, there is no role whatsoever for decompressive surgery in the treatment of discogenic (axial) pain syndromes. What then are the criteria for “definitive” discography and its use as an indication for surgery? Strict adherence to technique, including double needle, multiple blinded injections and identification of rostral and caudal controls, is essential, as is insistence on strict concordance with presenting complaints for a study to be considered “positive.” Finally, and most importantly, patient selection is of the greatest importance. Based on the current literature,19–30 one can argue that a
Facet Blockade
mechanical assessment should be an integral part of patient evaluation prior to consideration of discography. In addition, ideally, the patient should be free of confounding organic and psychological pathologies, should have disease limited to one or two levels, and should have reasonable expectations. Perhaps it is in this final area in which the thought processes of the diagnostician and surgeon must be most closely aligned.
Facet Blockade Numerous studies have demonstrated that the zygapophyseal joints, particularly in the lumbar spine, are a source of low back pain with or without referred sclerotomal pain.39–41 Several studies also suggest that the so-called facet pain may have a higher prevalence than previously suspected, with rates reported as high as 40% in older patients.42–44 While few would dispute the existence of posterior mechanical column pain in the presence of a sagittal deformity (e.g., spondylolisthesis), some investigators have disputed its existence without either such a deformity or coexisting degenerative changes in the motion segment.45 The potential clinical utility of a diagnostic response from anesthetic blockade of a suspected pain generator is highest when there is a significant gap between objective data and subjective complaints.21 Obviously, the ability of the block depends on the pharmacology of the agent used, the anatomical accuracy of the needle placement, and, perhaps most significantly, the ability of the patient to accurately report changes in symptoms. Kaplan and coworkers46 characterized the ability of lumbar medial branch blocks to anesthetize the facet joint. In this study, 18 asymptomatic individuals were assigned to L4-5 or L5, S1 facet blocks with radiographic contrast until capsular distention elicited pain. No extracapsular contrast was noted. One week later, 15 of the 18 underwent one of two randomized injections with saline or lidocaine. Thirty minutes after medial branch injections, the same individuals underwent repeat capsular distention of the joints that had been distended the preceding week. All five control individuals who received saline injections experienced pain with repeat capsular distention. Only one of the nine patients who received the active block experienced pain on capsular distention. Thus, with strict attention to technique, including the avoidance of inadvertent venous uptake with medial branch injection, facet blockade was successfully accomplished in 89% of the active treatment group. There are difficulties similar to those discussed for discogenic pain when one is attempting to identify patients who will be candidates for facet block on the basis of physical findings. Several studies to date44,47 have failed to identify the predictive value for any clinical findings or feature that would suggest a positive response to facet blockade. Revel and colleagues47 did note an increased likelihood of response to facet blockade in older patients who were relieved of pain in recumbency and did not have an increase in pain with coughing or use of the Valsalva maneuver. Specificity and sensitivity were increased when range of motion and functional tolerance were included: final sensitivity
61
62
Chapter 4 The Surgeon’s Perspective: Image-Guided Therapy
and specificity were, however, limited at 78 and 80%, respectively. Laslett and coworkers48 studied 151 patients and did identify four Clinically Predictive Rules (CPR), each of which required one or more positive findings from various signs and symptoms. The most useful included age greater than 50, symptoms best when walking, symptoms best when sitting, onset of pain is paraspinal, and positive extensionrotation test. If three of these five were present, there was a 100% chance (100% sensitivity) of 90–95% pain relief. In addition, the absence of centralization30,31 was associated with a positive response. As is the case with discography, there is no “gold standard” from a surgical point of view that can help to refine the diagnostic accuracy of facet blockade. In the lumbar spine, North and colleagues49 found that 42% of patients who had greater than 50% relief after facet anesthetic block had clinical improvement 2 years after facet rhizotomy. However, 17% of block responders who did not have facet rhizotomy were improved as well. In the cervical spine, some evidence exists that intervention for a facet-mediated pain problem may be warranted. Several studies50–53 have investigated the reliability of facet blockade in the cervical spine, as well as the utility of radiofrequency (RF) neurotomy.54 One published report investigated the correlation of facet blocks with lumbar fusion,4 but few meaningful conclusions can be drawn from this study, which was retrospective and did not use facet blockade as the definitive diagnostic procedure for surgical decision making. However, a recent study55 reported on 20 patients who underwent lumbar radiofrequency neurotomy. Mean duration of relief after the first treatment was 10.5 months. All patients had at least two treatments, and, overall, each treatment resulted in a mean duration of relief of 10.5 months and was successful in over 85% of patients. Thus, at the present time, the identification of facet-mediated pain by diagnostic blockade has little meaningful impact on surgical decision making. Based on the literature to date, RF facet rhizotomy may be viable. There are however, no convincing studies in the peer-reviewed literature suggesting that conventional surgical treatment (e.g., arthrodesis) is effective in treating facet-mediated pain syndromes in the absence of sagittal deformity.
Sacroiliac Joint (SI) Injections The difficulties identified in terms of sensitivity and specificity, particularly in comparing diagnostic blockade to a known or reproducible standard, also apply to SI joint blockade. It is generally accepted that the SI joint can be a source of pain owing to posterior ligamentous disruption (secondary to trauma), infection, or tumor. The characteristics of the so-called SI joint pain without these obvious anatomical correlates are, however, controversial. To date, no physical finding has proven to be specific enough to reliably diagnose sacroiliac joint pain.56 Additionally, the sacroiliac joint appears to be relatively immobile, and position has not been shown to be altered by manipulation.57 Technically, the SI joint may be more difficult to access than others, although access is possible with strict attention to fluoroscopic
Selective Nerve Root Blockade
technique.58–60 Several studies have noted that the pain provoked by joint distention may be ablated by anesthetic block.58,59 The clinical significance of this finding is unclear. Unfortunately, many of the appropriate afferent pathways are poorly understood. Additionally, in the presence of capsular incompetence, contrast extravasations may anesthetize nearby neural structures, further compounding the diagnostic difficulties with this particular injection. From a surgical point of view, perhaps the most telling limitation is the lack of any reproducible surgical procedure to treat sacroiliac joint pain. While joint reconstruction or arthrodesis has been demonstrated to restore pelvic stability in traumatic situations, there are no published reports in the peer-reviewed literature of significant pain relief following SI joint fusion for clinical syndromes diagnosed by SI joint blockade. Cavillo and colleagues61 reported two instances of successful treatment of presumptive SI joint pain by neurostimulation. The precise mechanism is unclear and certainly cannot be extrapolated. Thus, from a surgeon’s point of view, sacroiliac joint injections are therapeutic only because no firm recommendations can be made on surgical treatment for these presumed disorders.
Selective Nerve Root Blockade Selective nerve root blockade has received attention as a diagnostic and therapeutic tool in the management of referred pain, presumably of radicular origin. From the surgical point of view, the potential utility of this test lies in diagnostic specificity, not in its ability to identify a radicular etiology as the source of referred pain, but in its ability to localize a symptomatic level. In certain instances, with clinical evidence of radiculopathy and no underlying structural cause, nerve root blockade has been used to guide surgical intervention such as laminectomy or fusion.62 This is particularly distressing, since selective nerve root blockade has been found, in randomized prospective studies, to be neither sensitive nor specific.55,63,64 One report in the literature65 does suggest a more favorable response to decompressive surgery in the face of equivocal MRI findings when SNRB is positive. More compelling evidence for the therapeutic role of SNRB is provided by the studies of Riew and coworkers.66 In this paper, patients with documented lumbar disc prolapse and monoradiculopathy who responded to SNRB of bupiviacine and betamethasone had a 71% chance of avoiding lumbar disc surgery. These results held at one-year follow-up.67 Based on these data, it would appear that these blocks may have a therapeutic role, but the role as a definitive diagnostic maneuver is minimal. A particularly unfortunate clinical situation occurs when a patient who has been diagnosed with radiculopathy is informed that surgery is required for neural compression even though, from a strictly anatomical point of view, no surgical lesion exists. Again, the potential dichotomy between the diagnostician and the surgeon bears scrutiny. Although the patient may in fact have a radiculopathy that is helped
63
64
Chapter 4 The Surgeon’s Perspective: Image-Guided Therapy
by selective nerve root blockade, this may not be amenable to surgical treatment. Selective nerve root blockade has been used in the diagnosis of radicular syndromes.63–69 However, recent reports have called attention to temporary pain relief by reversible anesthetic blocks that failed to yield reliable long-term predictions about interventional results. There have been disappointing results from neuroablation procedures including dorsal rhizotomy59 as well as ganglionectomy,60 when these procedures were selected on the basis of response to selective nerve root blockade. Wetzel and coworkers70 reported a 19% success rate in patients who underwent selective lumbar sensory rhizotomy, with levels being selected on the basis of response to selective nerve root blockade. In this study, the decision to perform rhizotomy was based on the response to selective neural blockade that required reproduction of familiar pain, the disappearance of root tension signs after infiltration of anesthetic, and correlation between clinical and radiographic findings. These criteria were met in 90% of the cases, but satisfactory relief was not reliably obtained by selective sensory rhizotomy of the appropriate root. In addition, results of selective blockade may be confounded by systemic effects of lidocaine. When this is viewed in conjunction with the results of anesthesia of cutaneous nerves in the area of referred pain (i.e., pain relief), a notable lack of anatomical specificity becomes quite evident.71,72 North and coworkers64 performed diagnostic nerve blocks in a randomized prospective manner. In this study, 33 patients underwent a battery of local anesthetic blocks in an attempt to evaluate sciatica. The specificity of sciatic nerve block was 24% immediately and 36% at 1 h. The sensitivity of selective nerve root blockade was 91% immediately and 88% at 1 h. When analyzed in the context of blocks (from proximal to distal), the root block alone yielded significant pain relief in 9% immediately and in 21% at 1 h. The root block yielded greater relief of pain than any other block in 30% of patients immediately and in 42% at 1 h. In all other cases, the sciatic block or facet block yielded equal or better results. With the exception of one report65 to date, there has been no convincing study demonstrating the ability of conventional surgery (i.e., lateral recess decompression or foraminotomy) to reliably treat referred pain diagnosed predominantly on the basis of response to selective nerve root blockade. Selective blockade may, however, be of therapeutic value in the ongoing treatment of radicular pain.
Epidural Steroid Injections Epidural steroid injections should theoretically diminish inflammation in the epidural space and lead to improvement in symptoms resulting from neural compression. Epidural injections are commonly used in the setting of spinal stenosis with neurogenic claudication, and unilateral or bilateral radiculopathy from disc prolapse. A recent study by Wang and coworkers73 suggests that epidural steroid therapy benefits patients with lumbar disc prolapse and radiculopathy. In this
Minimally Invasive Intradiscal Therapy
retrospective review, 69 patients were studied. At an average follow-up of 1.5 years, 77% had resolutions of symptoms significant enough to cause them to decline surgical intervention. However, Carette and colleagues,74 in a randomized prospective double-blind study, examined the effects of epidural steroid injection on sciatica due to lumbar disc prolapse. The authors found no functional benefit in the group who underwent epidural injections. Short-term improvements in leg pain and sensory deficit were noted in the treatment group, but these benefits did not last beyond 3 months. Many patients in the study went on to discectomy within a year. Thus, from a surgical prospective, the diagnostic utility of epidural steroid injection is quite limited. Certainly, there are no convincing data suggesting that a response or lack of response to epidural injection correlates positively or negatively with the outcome of decompressive surgery. From a practical point of view, the use of epidural steroid therapy would appear to be reasonable in the symptomatic management of patients with compressive syndromes. Additionally, from a cost-effective point-of-view, it may be plausible to consider epidural therapy as a first-line intervention.
Minimally Invasive Intradiscal Therapy From a therapeutic point of view, the treatment of discogenic pain appears to be rather limited. On the one hand, appropriate conservative care (e.g., active physical therapy, pharmacological management) should be expected to yield success in the vast majority of cases. However, failing this, the only surgical treatments are arthrodesis or arthroplasty. Clearly, this is a very limited treatment continuum. From a philosophical point-of-view, a minimally invasive intradiscal treatment technique is quite attractive in an attempt to extend that continuum. Recent attention has been focused on the use of thermal energy to treat discogenic pain (Intradiscal Electrothermal Therapy, or IDET). Whether the mechanism of action is deafferentative, biomechanical, or both remains to be elucidated,75,76 although clinical data suggesting a delayed therapeutic effect after the procedure would suggest the latter. Numerous studies suggest a therapeutic effect.77–82 Although most studies have been prospective cohort controlled or retrospective, within the limitations of this study design, a therapeutic effect comparable to that of arthrodesis has been suggested. The two randomized prospective studies to date, however, fail to demonstrate any significant therapeutic effect. Pauza and coworkers,81 in a randomized, prospective, double-blind study (level I) reported on the 6-month results of 37 IDET versus 24 Sham patients. Significant improvements in Visual Analog Scale (VAS), Oswestry Disability Questionnaire (ODQ), and stratified Short Form 36 (Sf 36) Physical Function (PF) scores were noted in IDET versus Sham. The percentage change in pain between groups was significant as well, favoring the IDET group. However, a closer look at the data reveals that the authors did not report significance for subgroups that had >50% pain
65
66
Chapter 4 The Surgeon’s Perspective: Image-Guided Therapy
relief (37.5% IDET versus 33.3% Sham, NS), and that the SF 36 PF score differences were only significant after stratification. The second level I study is that of Freeman and colleagues.83 In this study, 38 were in the IDET group and 19 in the Sham group. Standard outcome scales were used: Low Back Outcome Score (LBOS), ODQ, sf 36 Zung Depression Index (ZDI), and Modified Somatic Perceptions Questionnaire (MSPQ). No subject in either group met the criteria for a successful outcome. Radiofrequency (RF) lesioning has been shown to be helpful in medial branch lesioning54,55 and Van Kleef and coworkers reported an effectiveness of 70% at 8 weeks.84 Once again, however, efficacy is not supported by Level I data. Barendese and colleagues85 studied 28 patients, all of whom had at least 1 year of LBP. Thirteen were randomized into an active lesion group and 15 into a Sham group. Standard outcome measures, including VAS and ODQ, were administered pretreatment and at 8 weeks follow-up. There were no significant differences between active and Sham groups and only two treatment successes in the Sham group versus one in the Active group. Coblation has been shown to decrease intradiscal pressure in vitro, with a proportionately greater change occurring in nondegenerated discs;86 however, the clinical significance of this remains unclear. Thus, from a surgical perspective, effective intradiscal therapy for discogenic LBP is notably absent.
Vertebroplasty and Kyphoplasty Osteoporotic vertebral compression fractures are the leading cause of disability and morbidity in the elderly.87–89 The consequences of these fractures may include pain and, in many cases, vertebral collapse and kyphosis. Traditionally, these fractures have been treated nonsurgically, as by definition no neurological compromise exists. Obviously, surgical reconstruction in the patient with osteoporosis is challenging. From a surgical point-of-view, orthopedic fracture care emphasizes the restoration of anatomy, correction of deformity, and subsequent preservation of function. These goals have not been met in the conservative care of patients with vertebral compression fractures. The ideal treatment should address both the fracture-related pain and the mechanical compromise related to kyphosis. Percutaneous vertebroplasty was described in 1987.90 In this procedure, whereby polymethylmethacrylate is injected into a compressed segment, immediate stability is obtained, but deformity is not corrected. Suggested indications included stabilization of painful osteoporotic fractures, painful fractures due to myeloma, and painful hemangiomata. Reports on clinical outcome for vertebroplasty have been encouraging, with most patients experiencing partial or complete pain relief within 72 h.91–96 Complication rates have been low, with the most significant complications resulting from extravertebral cement leakage causing spinal cord or nerve root compression, or pulmonary embolism.91,92,94–98 Additionally, a higher rate of extravasation has been noted in patients with metastatic disease versus patients with osteoporosis.99,100
Vertebroplasty and Kyphoplasty
Overall, vertebroplasty appears to be a reasonable method by which to treat a symptomatic vertebral compression fracture that has failed to respond to time-limited conservative care. Certainly, in a patient with multiple levels of and significant debility, this may be the procedure of choice. However, a potential theoretical limitation of vertebroplasty is its inability to address the aspect of persistent deformity, which is accompanied by a theoretical increased risk of adjacent segment degeneration, or possible fracture, as well as chronic pain related not to the fracture per se but, rather, to the postural concerns raised by deformity. Nonetheless, good results have been reported. Alvarez and coworkers compared outcomes in 101 patients who underwent vertebroplasty with those in 27 who refused. Significant improvements in pain and function were noted in the treatment group at 3 months, but there were no discernable differences between groups at 6 and 12 months posttreatment.101 Kyphoplasty claims to reduce a fracture via an inflatable bone tamp placed percutaneously into the vertebral body.100–105 Indications for kyphoplasty include painful osteoporotic compression fractures with induced kyphotic deformity. Kyphoplasty has not been investigated in the treatment of nonosteoporotic spinal metastatic disease. Initial reports of pain relief with kyphoplasty are comparable to those for vertebroplasty. In a study by Garfin and colleagues,106 90% of patients reported significant pain relief in the first 2 weeks of the procedure. In the initial series of these investigators, there were four major complications in 340 patients. Overall, serious adverse events occurred in 1.2% of patients.103 Wong and coworkers100 reported one presumed cement embolus to the lung, although this was attributed predominantly to technical issues associated with the use of less viscous cement. Lieberman and colleagues104 had one major and two minor complications while achieving an average of only 2.9 mm height restoration. In addition, Phillips and coworkers105 reported improvement in local kyphosis by a mean of 14°C. Kyphosis reduction may also be seen with vertebroplasty simply as a result of pain relief, so the effect with kyphoplasty may be less significant as an indicator of a procedural advantage. A note of caution must be sounded, however. Lavelle and Cheney reported a 10% incidence of recurrent fractures within the first 90 days following kyphoplasty. In their study group of 109 procedures in 94 patients, recurrent fractures occurred in 15% overall, with no significant difference in incidence noted between adjacent level and distant fractures.107 Patients who had multiple kyphoplasties were more likely to have additional fractures. Based on these data, it is unclear as to whether or not this represents a true effect of treatment or represents the natural history; the risk of a second fracture within one year of the first has been reported to be 19.2%.108 The obvious theoretical advantage of kyphoplasty – namely, an attempt to restore normal anatomy – requires further follow-up and investigation. Certainly, if fracture reduction can be demonstrated to result in a decreased risk of adjacent segment failure, either by a painful degenerative change or subsequent fracture, then the advantages
67
68
Chapter 4 The Surgeon’s Perspective: Image-Guided Therapy
of kyphoplasty would be apparent. However, height restoration, to date, has been meager,104 and the cost and complication rates remain a disadvantage when the bone tamp procedure is compared with vertebroplasty.
Conclusion From the point-of-view of planning surgical intervention, a diagnostic test must be sensitive, specific, and reproducible. The patient’s clinical findings must be precisely supported by the results of the diagnostic intervention. A well-studied surgical procedure to treat the specific pathology must be identified. Clearly, in many of the diagnostic regimens reviewed, the very nature of the tests (especially those involving pain provocation or ablation) may, by their very nature, preclude 100% sensitivity. Thus, the practical utility of a particular study in the matrix of clinical evaluation and subsequent surgical planning is of crucial importance. Appropriate patient selection and education are vital to identify patients who will have a successful surgical outcome. Ideally, the indications and expectations should be identical in the minds of the diagnostician and the surgeon. Finally, in many instances, more rigorous study of both diagnostic and surgical procedures is required. It is perhaps the greatest temptation of the clinician scientist to utilize promising techniques or procedures in an effort to alleviate patients’ suffering before the techniques have been completely evaluated. Thus, the exercise of compassionate restraint may be the greatest challenge facing clinicians today. References 1. Morris EW, DiPaola M, Vallance R, Waddell G. Diagnosis and decision making in lumbar disc prolapse and nerve entrapment. Spine 1986; 11(5):436–439. 2. Fischgrund JS, Mackay M, Herkowitz HN, Brower R, Montgomery DM, Kurz LT. Degenerative lumbar spondylolisthesis with spinal stenosis: a prospective, randomized study comparing decompressive laminectomy and arthrodesis with and without spinal instrumentation. Spine 1997;22:2807–2812. 3. Wetzel FT, LaRocca H, Lowery GL, Aprill CN. The treatment of lumbar spinal pain syndromes diagnosed by discography. Spine 1994;19(7):792–800. 4. Esses S, Moro J. The value of facet joint blocks in patient selection for lumbar fusion. Spine 1993;18(2):185–190. 5. Carragee EJ, Paragioudakis SJ, Khurana S. Lumbar high-intensity zone and discography in subjects without low back problems. Spine 2000;25(23): 2987–2992. 6. Carragee EJ, Tanner CM, Khurana S, Hayward C, Welsh J, Date E, Truong T, Rossi M, Hagle C. The rates of false positive lumbar discography in selected patients without low back symptoms. Spine 2000;25(11):1373–1381. 7. Carragee EJ, Tanner CM, Yang B, Brito JL, Truong T. False-positive findings on lumbar discography reliability of subjective concordance assessment during provocative disc injection. Spine 1999;24:2542–2547. 8. Fritzell H, Hogg O, Wessberg P, et al. Lumbar fusion versus non-surgical treatment for chronic low back pain. Spine 2001;26:2521–2534. 9. Brox JI, Sorensen R, Friis A, Nygaard O, Indahl A, Keller A, Ingebrigtsen T, Erikson HR, Holm I, Koller AK, Riise R and Reikeras, O. Randomized
Conclusion clinical trial of lumbar instrumented fusion and cognitive intervention and exercises in patients with chronic low back pain and disc degeneration. Spine 2003;28:1939–1921. 10. Dreyfuss PA, Schwarzer A, Lau P, Bogduk N. Specificity of lumbar medial branch and L5 dorsal ramus blocks: a computed tomography study. Spine 1999;22(8):895–902. 11. Airaksinen O, Herno A, Turunen V, Saari T, Suomlainen O. Surgical outcome of 438 patients treated surgically for lumbar spinal stenosis. Spine 1997;22:2278–2287. 12. Barnes D, Gatchel RJ, Mayer TG, Barnett J. Changes in MMPI profile levels of chronic low-back-pain patients following successful treatment. J Spinal Disord 1990;3:353–358. 13. Kanner AD, Coyne JC, Schaefer C, Lazarus RS. Comparison of two modes of stress measurement: daily hassles and uplifts versus major life events. J Behav Med 1981;4:1–39. 14. Pheasant HC, Gilbert D, Goldfard J, Herron LD. The MMPI as a predictor of outcome in low-back surgery. Spine 1979;4:78–84. 15. Sternbach RA. Psychological aspects of chronic pain. Clin Orthop 1977;129:150–155. 16. Grubb SA, Lipscomb HJ, Guilford WB. A relative value of lumbar roentgenograms, metrizamide myelography, and discography in the assessment of patients with chronic low back syndrome. Spine 1987;12:282–286. 17. Walsh TR, Weinstein JN, Spratt KF, Lehmann TR, Aprill C, Sayreh H. Lumbar discography in normal subjects. J Bone Joint Surg (Am) 1990;72:1081–1088. 18. Saal JS. General principles of diagnostic testing as related to painful lumbar spine disorders: a critical appraisal of current diagnostic techniques. Spine 2002;27(22):2538–2546. 19. Spratt K. Statistical Relevence. In Fardon DF (ed): Orthopaedic Knowledge Update: Spine 2. Rosemont, IL: The American Academy of Orthopaedic Surgeons, 2002:497–505. 20. Crock H. Internal disc disruption. Spine 1986;11:650–653. 21. Saal JA, Saal JS. Nonoperative treatment of herniated lumbar intervertebral disc with radiculopathy. An outcome study. Spine 1989;14(4): 431–437. 22. Parker L, Murrell S, Boden S, Horton W. The outcome of posterolateral fusion in highly selected patients with discogenic low back pain. Spine 1996;21(16):1909–1917. 23. Donelson R, Silva G, Murphy K. The centralization phenomenon: its usefulness in evaluating and treating referred pain. Spine 1990;15(3):211–213. 24. Karas R, McIntosh G, Hall H, Wilson L, Melles T. The relationship between nonorganic signs and centralization of symptoms in the prediction of return to work for patients with low back pain. Phys Ther 1997;77(4):354–360. 25. Long A. The centralization phenomenon: its usefulness as a predictor of outcome in conservative treatment of chronic low back pain. Spine 1995;20(23):2513–2521. 26. Sufka A, Hauger B, Trenary M, Hagan A, Lozon R, Martens B. Centralization of low back pain and perceived functional outcome. J Orthop Sports Phys Ther 1998;27(3):205–212. 27. Werneke M, Hart DL, Cook D. A descriptive study of the centralization phenomenon. a prospective analysis. Spine 1999;24(7): 676–683. 28. Werneke M, Hart DL. Centralization phenomenon as a prognostic factor for chronic low back pain and disability. Spine 2001;26(7):758–765. 29. Kopp JR, Alexander AH, Turocy RH, Levrini MG, Lichtman DM. The use of lumbar extension in the evaluation and treatment of patients with acute herniated nucleus pulposus, a preliminary report. Clin Orthop 1986;202:211–218.
69
70
Chapter 4 The Surgeon’s Perspective: Image-Guided Therapy 30. Donelson R, Aprill C, Medcalf R, Grant W. A prospective study of centralization of lumbar and referred pain: a predictor of symptomatic discs and anular competence. Spine 1997;22(10):1115–1122. 31. McKenzie R. The lumbar spine: mechanical diagnosis and therapy. Waikanae, New Zealand: Spinal Publications, 1981. 32. Derby R, Howard MW, Grant JM, Lettice JJ, Van Peteghem PK, Ryan DP. The ability of pressure-controlled discography to predict surgical and nonsurgical outcomes. Spine 1999;24:364–372. 33. Whitecloud TS III, Seago RA. Cervical discogenic syndrome. Results of operative intervention in patients with positive discography. Spine 1987;12:313–316. 34. Wood KB, Schellhas KP, Garvey TA, Aeppli D. Thoracic discography in healthy individuals. Spine 1999;24:1548–1555. 35. Goel VK, Grauer JN, Patel T Ch, Biyani A, Sairyo K, Vishnubhotla S, Matyas A, Cowgill I, Shaw M, Long R, Dick D, Panjabi MM and Serhan H. Effects of charite artificial disc on the implanted and adjacent spinal segments mechanics using a hybrid testing protocol. Spine 2005;30:2755–2764. 36. Huang RC, Girardi FP, Cammisa FP, Lim MR, Tropiano P, Marnay T. Correlation between range of motion and outcome after lumbar total disc replacement: 8.6 year follow-up. Spine 2005;30:1407–1411. 37. Le Huec J-C, Basso Y, Aunoble S, Friesem T, Bruno MB. Influence of facet and posterior muscle degeneration on clinical results of lumbar total disc replacement. J Spinal Disord Tech 2005;18:219–223. 38. Cinotti G, David T, Postacchini F. Results of disc prosthesis after a minimum follow-up period of two years. Spine 1996;21:995–1000. 39. Marks R. Distribution of pain provoked from lumbar facet joints and related structures during diagnostic spinal infiltration. Pain 1989;39:37–40. 40. McCall IW. Induced pain referral from posterior lumbar elements in normal subjects. Spine 1979;4:441–416. 41. Mooney V, Robertson J. The facet syndrome. Clin Orthop 1976;115:149–156. 42. Schwarzer AC, Wang S, Bogduk N, McNaught PJ, Laurent R. The prevalence and clinical features of lumbar zygapophysial joint pain: a study in an Australian population with chronic low back pain. Ann Rheum Dis 1995; 54:100–106. 43. Schwarzer AC, Aprill CN, Derby R, Fortin J, Kine G, Bogduk N. The falsepositive response rate of uncontrolled diagnostic blocks of the lumbar zygapophysial joints. Pain 1994;58:195–200. 44. Schwarzer AC, Aprill CN, Derby R, Fortin J, Kine G, Bogduk N. Clinical features of patients with pain stemming from the lumbar zygapophysial joints: is the lumbar facet syndrome a clinical entity? Spine 1994;19:1132–1137. 45. Jackson RP. The facet syndrome: myth or reality? Clin Orthop 1992;279: 110–121. 46. Kaplan M, Dreyfuss P, Halbrook B, Bogduk N. The ability of lumbar medial branch block to anesthetize the zygapophyseal joint. Spine 1998;23:1847–1852. 47. Revel M, Listrat V, Chevalier X, Dougados M, N’Guyen MP, Vallee C, Wybier M, Gires F, Amor B. Facet joint block for low back pain:identifying predictors of a good response. Arch Phys Med Rehab 1992;73:824–828. 48. Laslett M, McDonald B, Aprill CN, Tropp H, Oberg B. Clinical predictors of screening Lumber Zygopophyseal joint blocks: development of clinical prediction rules. Spine J 2006;6:370–379. 49. North R, Kidd D, Zahurack M, Piantadosi D. Specificity of diagnostic nerve blocks: a prospective, randomized study of sciatica due to lumbosacral spine disease. Pain 1996;65:77–85. 50. Barnsley L, Lord SM, Bogduk N. Comparative local and anesthetic blocks in the diagnosis of cervical zygapophysial joint pain. Pain 1993;55:99–106. 51. Barnsley L, Lord SM, Wallis BJ, Bogduk N. False-positive rates of cervical zygapophysial joint blocks. Clin J Pain 1993;9:124–130.
Conclusion 52. Barnsley L, Lord SM, Wallis BJ, Bogduk N. Lack of effect of intra-articular corticosteroids for chronic pain in the cervical zygapophysial joints. N Engl J Med 1994;330(15):1047–1050. 53. Barnsley L, Lord SM, Wallis BJ, Bogduk N. Chronic cervical zygapophysial joint pain after whiplash: a prospective prevalence study. Spine 1995;20:20–26. 54. Lord SM, Barnsley L, Bogduk N. Percutaneous radiofrequency neurotomy in the treatment of cervical zygapophysial joint pain: a caution. Neurosurgery 1995;36:732–739. 55. Schofferman J, Kine G. Effectiveness of radiofrequency neurotomy for lumbar facet pain. Spine 2004;29: 2471–2473. 56. Slipman C, Jackson H, Lipetz J, Chan K, Lenrow D, Vresilovic E. Sacroiliac joint pain referral zones. Arch Phys Med Rehabil 2000;81:334–338. 57. Tullberg T, Blomberg S, Branth B, Johnsson R. Manipulation does not alter the position of the sacroiliac joint: a roentgen stereophotogrammatic analysis. Spine 1998;23(10):1124–1129. 58. Fortin J, Dwyer A, West S, Pier J. Sacroiliac joint: pain referral maps upon applying a new injection/arthrography technique: I: asymptomatic volunteers. Spine 1994;19(13):1475–1482. 59. Fortin J, Dwyer A, West S, Pier J. Sacroiliac joint: pain referral maps upon applying a new injection/arthrography technique: II: clinical evaluation. Spine 1994;19:(13) 1483–1489. 60. Pulisetti D, Ebraheim N. CT-guided sacroiliac joint injections. J Spinal Disord 1999;12(4):310–312. 61. Calvillo O, Esses S, Ponder C, D’Agostino C, Tanhui E. Neuroaugmentation in the management of sacroiliac joint pain: report of two cases. Spine 1998;23(9):1069–1072. 62. Derby R, Kine G, Saal JA, Reynolds J, Goldthwaite N, White AH, Hsu K, Zucherman J. Response to steroid and duration of radicular pain as predictors of surgical outcome. Spine 1992;17:S176–S183. 63. Slosar PJ, White AH, Wetzel FT. Controversy. The use of selective nerve root blocks: diagnostic, therapeutic or placebo? Spine 1998;23: 2253–2256. 64. North RB, Kidd DH, Zahurak M, Piantadosi S. Specificity of diagnostic nerve blocks: a prospective, randomized study of sciatica due to lumbosacral spine disease. Pain 1996;65:77–85. 65. Sasso R, Macadaeg K, Nordmann D, Smith M. Selective nerve root injections can predict surgicaloutcome for lumbar and cervical radculopathy. J Spinal Disord Tech 2005;18:471–477. 66. Riew KD, Yin Y, Gilula L, Bridwell KH, Lenke LG, Lauryssen C, Goette K. The effect of nerve root injections on the need for operative treatment of lumbar radicular pain. J Bone Joint Surg 2000;82A:1589–1593. 67. Riew KD, Park J-B, Cho Y-S, Guila L, Patel A, Lenke L, Bridwell KH. Nerve root blocks in the treatment of lumbar radicular pain. J Bone Joint Surg 2006;88A:1722–1725. 68. Bonica JJ, Buckley FP. Regional analgesia with local anesthetics. In Bonica JJ (ed): The Management of Pain. Philadelphia: Lea & Febiger, 1990:1883–1966. 69. Loeser JD. Dorsal rhizotomy for the relief of chronic pain. J Neurosurg 1972;36:745–754. 70. Wetzel FT, Phillips F, Bernard TN, Aprill CN, LaRocca SH. Extradural sensory rhizotomy in the management of chronic lumbar radiculopathy: a minimum two-year follow-up study. Spine 1997;22:2283–2292. 71. North RB, Kidd DH, Campbell JN, Long DM. Dorsal root ganglionectomy for failed back surgery syndrome: a five-year follow-up study. J Neurosurg 1991;74:236–242.
71
72
Chapter 4 The Surgeon’s Perspective: Image-Guided Therapy 72. Weiss S, Davis D. The significance of the afferent impulses from the skin in the mechanism of visceral pain: skin infiltration as a useful therapeutic measure. Am J Med Sci 1928;176:517. 73. Wang JC, Brodke DS, Youssef JD. Epidural injections for the treatment of symptomatic lumbar herniated discs. J Spine Disord 2002;15:269–272. 74. Carette S, Leclaire R, Marcoux S, Morin F, Blaise GA, St.-Pierre A, Truchon R, Parent F, Lévesque J, Bergeron V, Montminy P, Blanchette C. Epidural corticosteroid injections for sciatica due to herniated nucleus pulposus. N Engl J Med 1997;336:1634–1640. 75. Kleinstueck F, Diederich C, Nan W, Smith JA, Puttlitz C, Lotz J, Bradford DS. The IDET procedure: thermal distribution and biomechanical effects on human lumbar disk. North American Spine Society, 15th Annual Meeting, New Orleans, October 25–28, 2000. 76. Shah RV, Lutz CE, Lee J, Doty SB, Rodeo S. Intradiscal electrothermal therapy: a preliminary histologic study. Arch Phys Med Rehabil 2001;82:1230–1237. 77. Saal JA, Saal JS. Intradiscal electrothermal treatment for chronic discogenic low back pain. a prospective outcome study with minimum 1-year followup. Spine 2000;25(20):2622–2627. 78. Saal JS, Saal JA. Management of chronic discogenic low back pain with a thermal intradiscal catheter. Spine 2000;25(3):382–387. 79. Karasek M, Bogduk N. Twelve-month follow-up of a controlled trial of intradiscal thermal annuloplasty for back pain due to internal disc disruption.Spine 2000;25(20):2601–2607. 80. Thompson K, Eckel T. IDET Nationwide Registry preliminary results: 6-month follow-up data on 170 patients. North American Spine Society, 15th Annual Meeting, New Orleans, October 25–28, 2000. 81. Pauza K, Howell S, Dreyfuss P, Peloza J, Park K. A randomized doubleblinded placebo-controlled trial evaluating the efficacy of intradiscal electrothermal annuloplasty (IDET) for the treatment of chronic discogenic low back pain: 6-month outcome. International Spinal Injection Society, 10th Annual Meeting, Houston, TX, September 7, 2002. 82. Wetzel FT, Anderson GBJ, Peloza J, Rashbaum R, Lee CR, Yuan HK, Phillips FM, An HS. Intradiscal electrothermal therapy (IDET) to treat discogenic low back pain: preliminary results of a multi-center prospective cohort study. North American Spine Society, 15th Annual Meeting, New Orleans, October 25–28, 2000. 83. Freeman BJC, Frasier RD, Cain CMJ, Hall DJ, Chapple DC. A Randomized, Double-blind controlled trial. Intradiscal electrothermal therapy versus placebo for the treatment of chronic discogenic low back pain. Spine 2005;30:2369–2377. 84. Van Kleef M, Barendese GAM, Kessels A, Voets HM, Weber WEJ, de Lange S. Randomized trial of radiofrequency lumbar facet denervation for chronic low back pain. Spine 1999;24:1937–1924. 85. Barendese GAM, van den Berg SGM, Kessels HGF, Weber WEJ, van Kleef M. Randomized controlled trial of percutaneous intradiscal radiofrequency thermocoagualtion for chronic discogenic low back pain: lack of effect from a 90-second 70C lesion. Spine 2001;26:287–92. 86. Chen YC, Lee S-h, Chen D. Intradiscal pressure of percutaneous disc decompression with nucleoplasty in human cadavers. Spine 2003;28:661. 87. Iqbal MM, Sobhan T. Osteoporosis: a review. Mod Med 2002;99:19–24. 88. Verbrugge LM, Lepkowski JM, Imanaka Y. Comorbidity and its impact on disability. Milbank Q 1989;67:450–484. 89. Johnell O. Advances in osteoporosis: better identification of risk factors can reduce morbidity and mortality. J Intern Med 1996;239:299–304.
Conclusion 90. Galibert P, Deramond H, Rosat P, LeGars D. Preliminary note on the treatment of vertebral angioma by percutaneous acrylic vertebroplasty. Neurochirurgie 1987 33:166–168. 91. Cortet B, Cotton A, Boutry N, Flipo RM, Duquesnoy B, Chastanet P, Delcambre B. Percutaneous vertebroplasty in the treatment of osteoporotic vertebral compression fractures: an open prospective study. J Rheumatol 1999;26:2222–2228. 92. 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 surgical cement in 20 Patients. Am J Roentgenol 1999;173:1685–1690. 93. Deramond H, Depriester C, Galibert P, Le Gars D. Percutaneous vertebroplasty with polymethylmethacrylate. Technique, indications, and results. Radiol Clin North Am 1998;36:533–546. 94. Gangi A, Kastler BA, Dietemann JL. Percutaneous vertebroplasty guided by a combination of CT and fluoroscopy. Am J Neuroradiol 1994;15:83–86, 1994. 95. Grados F, Depriester C, Cayrolle G, Hardy N, Deramond H, Fardellone P. Long-term observations of vertebral osteoporotic fractures treated by percutaneous vertebroplasty. Rheumatology (Oxford) 2000;39:1410–1414. 96. Mathis JM, Petri M, Naff N. Percutaneous vertebroplasty treatment of steroid-induced osteoporotic compression fractures. Arthritis Rheum 1998;41:171–175. 97. Hardouin P, Grados F, Cotton A, Cortet B. Should percutaneous vertebroplasty be used to treat osteoporotic fractures? An update. Joint Bone Spine 2001;68:216–221. 98. Padovani B, Kasriel O, Brunner P, Peretti-Viton P. Pulmonary embolism caused by acrylic cement: a rare complication of percutaneous vertebroplasty. Am J Neuroradiol 1999;20:375–377. 99. Watts NB, Harris ST, Genant HK. Treatment of painful osteoporotic ver-tebral fractures with percutaneous vertebroplasty or kyphoplasty. Osteoporosis Int 2001;12:429–437. 100. Wong W, Reiley MA, Garfin S. Vertebroplasty/Kyphoplasty. J Women’s Imag 2000;2:117–124. 101. Alvarez L, Alcaraz M, Perez-Higueras A , Granizo JJ, De Miguel I, Rossi RE, Quinones D. Percutanoues Vertebrolasty. Spine 2006;31:1113–1117. 102. Belkoff SM, Mathis JM, Fenton DC, Scribner RM, Reiley ME, Talmadge K. An ex vivo biomechanical evaluation of an inflatable bone tamp used in the treatment of compression fracture. Spine 2001;26:151–156. 103. Garfin SR, Yuan HA, Reiley MA. New technologies in spine: kyphoplasty and vertebroplasty for the treatment of painful osteoporotic compression fractures. Spine 2001;26:1511–1515. 104. Lieberman IH, Dudeney S, Reinhardt MK et al. Initial outcome and efficacy of “kyphoplasty” in the treatment of painful osteoporotic vertebral compression fractures. Spine 2001;26:1631–1638. 105. Phillips FM, McNally T, Wetzel F T et al. Early clinical and radiographic results of kyphoplasty for the treatment of osteopenic vertebral compression fractures. Eur Spine J 2001;10:(suppl 1)S7. 106. Garfin S, Lin G, Lieberman I, et al. Retrospective analysis of the outcomes of balloon kyphoplasty to treat vertebral body compression fracture (VCF) refractory to medical management. Eur Spine J 2001;10 (suppl 1):S7. 107. Lavelle WF, Cheney R. Recurrent fracture after vertebral kyphoplasty. Spine J 2006;6:488–493. 108. Lindsay R, Silverman SL, Cooper C, Hanley DA, Barton I, Broy SB, Licata A, Benhamou L, Geusens P, Flowers K, Stracke H, Seeman E. Risk of a new vertebral in the year following a fracture. JAMA 2001;285: 320–323.
73
5 Image-Guided Percutaneous Spine Biopsy A. Orlando Ortiz, Gregg H. Zoarski, and Allan L. Brook
Introduction Prior to the development of image-guided percutaneous spine biopsy techniques, an open biopsy procedure was required for definitive diagnosis. The advantage of the open biopsy procedure is twofold. First, under direct visualization, multiple and larger tissue samples can be obtained and made available for frozen histopathological analysis. Second, the open biopsy can be performed as part of a surgical decompression and/or stabilization procedure of the spine. The first report of percutaneous spine biopsy was in 1935 by Robertson and Ball.1 Their procedures, however, did not utilize imaging guidance. Siffert and Arkin utilized a posterolateral approach for spine biopsy using radiographic guidance.2 Fluoroscopy-guided spine biopsy was subsequently reported in 1969, and CT-guided spine biopsy was reported in 1981.3,4 Percutaneous spine biopsy has several advantages over an open biopsy procedure. The percutaneous image-guided procedure is faster and more cost-effective and has an overall lower risk of complications.5 Image-guided spine biopsy procedures are usually performed to diagnose suspected primary or secondary neoplastic processes or to evaluate for the presence of infectious spondylitis.6 These procedures are less frequently performed to assess for other noninfectious inflammatory conditions that can affect the spine. The decision to perform a spine biopsy procedure is made after close communication between the radiologist and the referring clinician. Both individuals must be convinced that the benefit to be gained from the biopsy results firmly outweighs the risks of the procedure. To this end, as a prerequisite, there must be a thorough medical history and a physical examination, combined with a complete review of all prior imaging and laboratory examinations. This consultation will avoid unnecessary spine biopsies (when they are not indicated or when a more accessible bone biopsy site, such as the iliac bone, is available), From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_5, © Springer Science + Business Media, LLC 2010
75
76
Chapter 5 Image-Guided Percutaneous Spine Biopsy
Table 5.1. Indications for spine biopsy 1. Suspected secondary spine tumor (i.e., metastasis) with either a known or an unknown primary tumor 2. Suspected secondary spine tumor with a history of two or more preexisting primary tumors 3. Suspected primary spine or paraspinal tumor 4. Pathological compression fracture 5. Suspected infectious spondylitis 6. Suspected inflammatory condition that involves the spine Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004.
ensure patient safety, and identify the optimal location and level for performing the biopsy procedure. Spine biopsy is often performed to evaluate destructive or spaceoccupying lesions within the spinal axis (Table 5.1). Abnormal foci of marrow replacement within the vertebral column that are detected with noninvasive imaging modalities, such as Computed Tomography (CT) or Magnetic Resonance Imaging (MRI), also are often referred for spine biopsy. In every instance, the decision to proceed with a biopsy procedure is based upon a thorough analysis of risks and benefits. The overall benefit of the information gained by the procedure should always favor its performance. The results of the biopsy will affect the subsequent clinical management of the patient and influence treatment decisions in such areas as surgery, chemotherapy, radiation therapy, and antibiotic therapy. The immediate contraindication to percutaneous biopsy is coagulopathy. Yet, even this condition, when properly anticipated and managed, can be corrected long enough to permit a surgeon to perform the procedure. When a vascular tumor such as a renal metastasis is suspected, a catheter angiogram should be considered in the diagnostic workup. These vascular lesions, however, can be carefully sampled with smaller gauge-core needle biopsy systems and with fine-needle aspiration techniques (Figure 5.1). Informed consent must be obtained prior to the procedure after the patient has received an explanation of the benefits and risks of image-guided percutaneous spine biopsy. The procedure offers the benefit of supplying diagnostic information that will guide subsequent treatment decisions. The alternative procedure is an open spine biopsy. The general risks of percutaneous spine biopsy include bleeding at or deep to the puncture site manifested as active hemorrhage or hematoma formation (Tables 5.2). Infection is another potential complication associated with spine biopsy, hence the requirement for strict aseptic technique when the procedure is performed. The spread of disease by the biopsy procedure, an extremely rare complication that has been described,7 is related to tumor implantation or spread of infection along the biopsy tract.5 The development of coaxial biopsy techniques and transcortical approaches with shorter needle
Introduction
Figure 5.1. Axial CT image shows a lytic lesion (arrows) that is centered primarily within the posterior elements of the thoracic vertebra. Since the patient had a history of kidney resection, this lesion was sampled by fineneedle aspiration with a 22-gauge spinal needle. A single pass showed positive cytology for metastatic renal cell carcinoma. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
trajectories has decreased the incidence of these complications. Site-specific biopsy complications that have been reported are related to the spine level (cervical, thoracic, or lumbar spine) that was sampled and the proximity to critical structures. Pneumothorax can occur not only during thoracic spine biopsy, but also during the attempted biopsy of thoracolumbar or cervicothoracic lesions. Neural injury, particularly to the spinal cord, is a devastating complication that has been reported. Nevertheless, the incidence of reported complications
Table 5.2. Complications associated with spine biopsy 1. Active hemorrhage 2. Hematoma 3. Vascular injury 4. Neural injury (spinal cord or nerve) resulting in transient or permanent paralysis 5. Pneumothorax 6. Infection, including meningitis Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004.
77
78
Chapter 5 Image-Guided Percutaneous Spine Biopsy
in percutaneous skeletal biopsy is low, estimated at less than 0.2%.5 The combination of image guidance, small-gauge biopsy needle systems, and operator experience should result in an overall major complication rate that is much less than 1%.
Patient Preparation Percutaneous spine biopsy can be performed either on an inpatient or outpatient basis. The patient must not eat or drink for a minimum of 8 h prior to the procedure. The following laboratory parameters are assessed at the authors’ institution: hematocrit, hemoglobin, platelet count, Prothrombin Time (PT), Partial Thromboplastin Time (PTT), International Normalized Ratio (INR), Blood Urea Nitrogen (BUN), and creatinine. Patient allergies are recorded, with particular attention to anesthetic agents and imaging contrast agents. Prior to performing the biopsy procedure, the operator should carefully scrutinize all pertinent imaging studies. This will help identify the optimal lesion(s) for biopsy and the safest approach to access the lesion(s). Furthermore, this pre-procedure evaluation can assist in the selection of the appropriate needles and imaging modality. Percutaneous spine biopsy can be performed with local anesthesia, with local anesthesia and conscious sedation, or under general anesthesia. The procedure is often performed with a combination of local anesthesia and intravenous conscious sedation using a shortacting benzodiazepine (Versed ®, Hoffmann–La Roche, Nutley, NJ) and an analgesic such as fentanyl or morphine. While general endotracheal anesthesia often is not utilized owing to the requirement for prone positioning of the patient, general intravenous anesthesia can be performed with propofol. To minimize the possibility of infection, the study should be performed with strict aseptic technique. Patient positioning depends upon the spine level (cervical, thoracic, or lumbosacral) of the lesion and its location (vertebral body versus posterior elements). The prone position is optimal for accessing lesions in the thoracic or lumbosacral spine or, rarely, within the posterior aspect of the cervical spine. The supine position is usually required to access the cervical spine. In certain instances – for example, when a patient cannot lie completely prone – the lateral decubitus or prone oblique position can be helpful (Figure 5.2). Patient monitoring is performed with the help of a pulse oximeter, continuous electrocardiography, and an automated blood pressure cuff. Appropriate placement of the monitoring equipment is required so that it does not obscure the field of view during the procedure and does not contaminate the sterile field. An intravenous catheter should also be in place prior to the procedure to facilitate the intravenous administration of medications, contrast agents, or hydration. The antecubital fossa should be avoided in situations that require prone positioning of the patient: the patient’s elbows are often flexed in this position, and the intravenous catheter function can be compromised.
Equipment Requirements
Figure 5.2. Axial CT image obtained during a thoracic spine biopsy performed with the patient, who could not tolerate the prone position, in a prone oblique position. A transcostovertebral approach (arrow) was used, and the lesion was subsequently shown to be an osteoporotic compression fracture. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Equipment Requirements Image guidance can be accomplished with several different modalities. These include fluoroscopy, computed tomography, computed tomography combined with a multidirectional fluoroscope, computed tomographic fluoroscopy, and magnetic resonance imaging.8 The choice of equipment is determined by its availability, operator preference, and by the location and size of the suspected lesion. A CT-guided spine biopsy can be performed without or with the use of a stereotactic apparatus to guide the insertion of the biopsy needle.9–11 The use of MRI requires the simultaneous use of MR-compatible equipment, both for patient monitoring and for performing the biopsy procedure. The modality selected depends upon its availability and the training and experience of the operator. The cross-sectional modalities afford the advantage not only of precise lesion localization, but also of “critical” structure (e.g., lung, aorta, and carotid artery) identification. In experienced hands, however, fluoroscopy-guided biopsies tend to be performed more quickly and with good patient safety. For cervical spine biopsy, CT, fluoroscopy, or CT with fluoroscopy facilitates the selection of an optimal biopsy trajectory that yields access to the
79
80
Chapter 5 Image-Guided Percutaneous Spine Biopsy
Table 5.3. Some commercially available biopsy systems System
Manufacturer or city
Aspiration 3.5-6 in. 18- to 22-gauge spinal nee- Becton–Dickinson, Rutherford, NJ dles 10-20 cm 22-gauge Chiba needles Cook Co., Bloomington, IN Cutting Tru-cut Trephine Craig Ackermann Elson Franseen Geremia Jamshidi Parallax Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004.
lesion but avoids critical neck structures. Numerous factors influence the total procedure time, but the average time using local anesthesia is approximately 30 min. This assumes that the patient is cooperative and that the radiologist and the radiology technologist are experienced in biopsy procedures.12 Several biopsy needle systems are commercially available (Table 5.3). The system that is utilized depends upon the lesion type (soft tissue or osseous), the lesion location (vertebra, disc space, and paraspinal soft tissues), and the method of specimen acquisition (aspiration biopsy versus core biopsy). Aspiration biopsy can be performed with a 22- or 20-gauge stylet-bearing needle. Core biopsy can be performed with a trephine or beveled tip (usually 11-, 12-, or 14-gauge) bone biopsy needle or a soft tissue-cutting needle (usually 18-gauge) (Figure 5.3). These core biopsy needles can be used as part of either a tandem needle system or a coaxial system. In the tandem technique, the needle that is used in the initial application of local anesthesia both localizes the lesion and serves as a visual guide. In a simultaneous tandem system, the biopsy needle is placed alongside a thin needle that was previously placed to anesthetize the biopsy tract. In a sequential tandem system, the biopsy needle is advanced along a tract previously created by the smaller anesthetizing needle. Coaxial needle systems have increased in popularity.13 The biopsy needle is advanced over the anesthetizing and localizing needle (22-gauge). The localizing needle has a removable hub and serves as a mechanical guide for the biopsy needle. A guiding cannula, through which multiple biopsy needle passes can be made, is left in place. Coaxial biopsy needle systems are particularly helpful for cervical spine biopsies. The major advantages of the coaxial system, therefore, are a decreased procedure time, resulting from better accuracy, and decreased procedure complications. Only a single biopsy tract is used with the coaxial system, thus avoiding the risk of additional soft
Biopsy Techniques
Figure 5.3. An 18-gauge soft tissue-cutting needle (arrow) is used to obtain a core of soft tissue from this large paraspinal mass that erodes the lateral margin of the vertebral body. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
tissue structure injury associated with a second pass. Additionally, the guiding cannula can serve as a guide for fine-needle aspiration prior to core biopsy, or for obtaining multiple core biopsy samples with a soft tissue-cutting needle. An 18-gauge spring-loaded biopsy needle is used to obtain soft tissue cores. Accessory guidance systems have been developed to facilitate needle localization. These vary in complexity and are infrequently used in routine practice.
Biopsy Techniques An important decision that is made before and during spine biopsy is the choice of approach. The determinants for the approach are lesion location and lesion size (Table 5.4).14 A posterior approach is used for thoracic, lumbosacral, and posterior cervical lesions. An anterior approach is used for most cervical spine biopsies. The location of “critical” normal anatomical structures also will modify the approach. Unless the lesion is clearly localized to the left side of the spine, for example, a right-sided approach is preferable to a left-sided approach for accessing thoracic spine tumors without damaging the aorta. In the cervical spine, the critical structures include the great vessels of the neck, the pharynx and hypopharynx, the trachea, the esophagus, the thyroid gland, the lung apices, and the spinal cord. In the thoracic spine, the critical structures are the lungs and the aorta. In the lumbar spine, the critical
81
82
Chapter 5 Image-Guided Percutaneous Spine Biopsy
Table 5.4. Biopsy approaches Location
Approach
Spine level
Bone
Paraspinal oblique Transpedicular Transcostovertebral Posterolateral Anterolateral
Disc
Paraspinal oblique Posterolateral Anterolateral Paraspinal oblique
Thoracic or lumbar Cervical
Posterolateral Anterolateral
Thoracic or lumbar Cervical
Paraspinal Soft tissues
Thoracic or lumbar Thoracic Lumbar > thoracic > cervical Cervical
Reprinted with the kind permission of Springer Science+Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004.
structures are the aorta, inferior vena cava, kidneys, large and small bowel, conus, and exiting spinal nerves. The objective is to choose a trajectory that enables access to the lesion without compromising normal, critical structures (Figure 5.4). The specific location of the lesion within the spine also will influence the approach that is selected. A vertebral body lesion and a posterior element lesion (Figure 5.5) will be approached differently.
Figure 5.4. Axial CT image shows a localizing needle adjacent to the right pedicle (long arrow) of a lumbar vertebra. A transpedicular approach was chosen to access the most proximal (small arrow) of three sclerotic lesions in a patient with a history of breast cancer. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Biopsy Techniques
Figure 5.5. Axial CT image shows an expansile lytic lesion within the right transverse process and posterior vertebral body of this thoracic vertebra. Fine-needle aspiration of the right transverse process (arrow) was therefore performed with a 22-gauge Chiba needle. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
The type of posterior approach (posterolateral, transpedicular, or transcostovertebral) can be tailored to the specific location of the lesion (Figure 5.6). The posterolateral approach can be used to access lesions located within the vertebral body, disc, or paraspinal soft tissues of the lumbar spine (Figures 5.7 and 5.8). The transpedicular
Figure 5.6. Diagram of vertebra indicating the biopsy routes for the posterolateral, transpedicular, and transcostovertebral approaches. (Drawing modified with permission from Dr. Bernadette Stallmeyer; Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
83
84
Chapter 5 Image-Guided Percutaneous Spine Biopsy
Figure 5.7. Axial CT image obtained during a disc and vertebral endplate biopsy (arrow) shows a bone biopsy needle inserted via a left posterolateral approach. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Figure 5.8. Axial CT image shows a left parapedicular approach (arrow) used to sample this destructive vertebral body lesion. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Biopsy Techniques
approach can be used to safely access lesions within the thoracic or lumbar vertebrae. A transcostovertebral approach can be used for thoracic disc space lesions, thoracic paraspinal soft tissue masses, or vertebral body lesions (Figure 5.9). The selected imaging modality is used to identify the lesion level (Figure 5.10). Once a safe path to the target lesion has been chosen, the entry site on the skin surface is marked with an indelible ink marker. The region of interest is then prepped and draped in sterile fashion. A 1 cm wheal is raised at the skin entry site by using a 25-gauge needle and a local anesthetic agent (e.g., 1% lidocaine, 0.25% bupivacaine). A #11 scalpel blade is used to make a dermatotomy incision at the skin entry site. A stylet-bearing thin needle is then advanced by means of image guidance, and the local anesthetic is then administered into the deeper soft tissues. If a vertebra is to be entered, infiltration of the anesthetic agent into the periosteum is extremely helpful in minimizing patient discomfort. With coaxial technique, the position of the needle tip relative to the lesion is adjusted and confirmed by means of image guidance. When the needle tip is in satisfactory position, the needle hub is removed and the needle then essentially serves as a stiff guide wire. A guiding cannula is inserted over the hubless needle and advanced to the desired level under image guidance. Aspiration or core needles can be passed through this guiding cannula to obtain specimens. The needle tip must always be accounted for with respect to the target lesion and to all pertinent critical structures (Figure 5.11). This rule
Figure 5.9. Axial CT image shows a right transcostovertebral approach (arrow) used to sample this destructive vertebral body lesion (fungal osteomyelitis). (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
85
86
Chapter 5 Image-Guided Percutaneous Spine Biopsy
Figure 5.10. Steps in a CT-guided biopsy. (A) To use a right transpedicular approach (long black arrow), the skin entry site must be located near the second skin marker (white arrow). (B) The guide needle is advanced to the posterior margin of the pedicle (arrow).
applies especially to cutting needles; their biopsy chamber requires additional exposure and excursion within the lesion matrix to enable the cutting portion of the needle mechanism to slide over the biopsy
Figure 5.10. Continued. (C) The bone biopsy needle is advanced through the pedicle to sample the right-sided lytic lesion (arrow). (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Figure 5.11. Axial CT image obtained during a bone biopsy shows a guide needle that reaches the anterior vertebral body cortex (large arrow). Note the proximity of this needle to the aorta (arrowhead). The guide needle had been advanced far beyond the target lesion (small arrow). (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
88
Chapter 5 Image-Guided Percutaneous Spine Biopsy
chamber and retrieve the specimen. Moreover, specimen retrieval by means of fine-needle aspiration requires an in-and-out motion within the lesion matrix. Failure to completely account for the position of the needle tip may result in an unsuccessful biopsy and may also injure a critical structure. To access bone marrow or a lytic lesion with an aspiration or cutting needle, a pre-existing bone window must be present within the vertebral cortex, as occurs with a lytic focus, or a cortical window must first be cut with a bone needle. Neither aspiration nor cutting needles will penetrate normal or near normal bone cortex. Cervical spine biopsy often requires an anterolateral approach.14 The neck can be separated into suprahyoid and infrahyoid compartments (Figure 5.12). The location of the carotid space contents within these compartments and the location of the spinal lesion will determine the skin entry site for the biopsy (Figure 5.13). Other important structures that are to be avoided include oropharynx, hypopharynx, and visceral space contents (esophagus, trachea, and thyroid gland). In approaching lower cervical spine lesions, care must be taken to avoid the pulmonary apex. In addition to being constantly aware of the location of the carotid artery and jugular vein, the operator must be cognizant of the location of the vertebral artery. When in doubt about the location or identity of a potentially important vascular structure, administer an intravenous contrast agent to clarify the situation. The trajectory can be anterior or posterior to the carotid space, depending on the location of the great vessels. A 22-gauge needle can be used to go safely beside these structures with CT guidance. Alternatively, some operators prefer to use palpation and carotid displacement during the initial needle placement, to bypass the carotid artery. This maneuver is often performed with fluoroscopyguided biopsy procedures. Once the needle tip has passed beyond the carotid space and is near the target, a coaxial technique can be used to safely obtain multiple biopsy specimens. A posterior approach is occasionally required for accessing posterior element lesions. Given the relatively small size of the posterior elements and the proximity to the spinal cord, it is advisable to utilize CT for safely approaching and sampling lesions in this location.15 For thoracic or lumbar spine lesions, a transpedicular approach is optimal for accessing centrally located vertebral body lesions (Figure 5.14). The pedicle provides a safe passageway to the vertebral body. Special care must be taken to avoid fracturing the pedicular cortex. This complication can cause either direct injury to the spinal cord or exiting nerve root, or can indirectly injure these structures by leading to hematoma formation. The margins of the pedicle should be visualized at all times while the biopsy needle courses through the pedicle. A potential pitfall of the transpedicular approach, which occurs when the pedicle is not involved by tumor, is the possibility of obtaining a false-negative biopsy result. The solution in such cases is to take deeper and multiple samples. The transcostovertebral approach is useful in accessing laterally located thoracic vertebral lesions or in sampling the thoracic disc.12 The posterolateral approach is ideal for accessing laterally located vertebral
Biopsy Techniques
Figure 5.12. Location of the prevertebral and paravertebral spaces within the suprahyoid (A) and infrahyoid neck (B). Note the anterolateral position of the carotid space (arrows) relative to the prevertebral space. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
89
90
Chapter 5 Image-Guided Percutaneous Spine Biopsy
Figure 5.13. Steps in a CT-guided biopsy of the cervical spine. (A) Skin markers (a set of 4 taped 18-gauge 1 in. needles) are placed for an anterolateral approach with the patient in the supine position. The soft tissue window algorithm is used to identify the carotid artery (arrow) and internal jugular vein (arrowhead). (B) Coaxial technique is used to advance a needle with a removable hub, through a short 18-gauge needle, past the carotid artery (white arrow) and adjacent to the abnormal cervical vertebra (black arrow).
Biopsy Techniques
Figure 5.13. Continued. (C) A bone cannula is safely advanced over the wire. A trephine needle (arrow) is advanced into the substance of the vertebral body to obtain a core of bone. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
body lesions or their paraspinal soft tissue components or the intervening disc space within the lumbar spine. Two or preferably three core specimens are obtained and placed in 10% formalin. When bone biopsy cores are obtained, they must undergo a period (approximately 48 h) of decalcification in 7% formic acid, whereupon the specimens are embedded in paraffin for subsequent histological sectioning and staining. The reported diagnostic accuracy of core biopsy ranges from 77 to 97%.16 If the clinical concern is infection, the specimens are placed in sterile containers and immediately brought to the microbiology laboratory for appropriate processing. When aspiration biopsy is anticipated, it should be performed prior to obtaining any core specimens, since the core biopsy can create a hemorrhagic tract that prevents successful aspiration of the desired abnormal tissue.16 Otherwise, a different tract to the lesion must be utilized. Successful aspiration biopsy requires a secure fit between the aspirating syringe and the needle hub to facilitate forceful suction. Full negative pressure is generated by using a 20-mL syringe while the needle is being advanced and retracted within the lesion.17 The distance of the needle excursions depends upon the lesion size; large lesions permit safer, longer excursions, and short excursions are required for small lesions adjacent to critical structures (Figure 5.15). Needle excursions extending
91
92
Chapter 5 Image-Guided Percutaneous Spine Biopsy
Figure 5.14. Steps in a CT-guided biopsy of the thoracic spine. (A) The patient is in the prone position and skin markers (arrow) are placed to determine the optimal skin entry site. (B) A 1.5 in. 22-gauge needle is used to administer local anesthetic along the biopsy tract to the periosteal surface (arrow).
Figure 5.14. Continued. (C) The sequential tandem technique is used to replace the 22-gauge needle with a 12-gauge bone needle, which is gradually advanced through the pedicle (arrow) and into the vertebral body under imaging guidance. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Figure 5.15. Axial CT image shows a large right paraspinal mass (arrows) that erodes into the lumbar vertebra. The size of this mass permits long excursions of the biopsy needle during fine-needle aspiration. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
94
Chapter 5 Image-Guided Percutaneous Spine Biopsy
more than 3–4 mm are required to obtain a specimen.18 Slight adjustments in angulation, when possible, are made with each needle pass to increase the yield of pathological tissue.18 A flash of hemorrhagic fluid within the needle hub usually signals the end point of aspiration. In the ideal situation, the needle and syringe are withdrawn from the spinal lesion, and this ensemble is immediately handed to a cytotechnologist, who prepares slide smears of the specimen. The technologist or a pathologist looks at the slides under a microscope and determines whether abnormal cells are present within the specimen. Alternatively, the biopsy specimen can be placed in 95% ethanol before being sent for cytological analysis. The published accuracy of aspiration biopsy is series dependent and ranges from 23 to 97%.16 When infection is the working clinical diagnosis, the aspirates are not placed in ethanol but instead are submitted in sterile containers to the microbiology laboratory. If fluid cannot be aspirated, a few milliliters of sterile, nonbacteriostatic normal saline can be injected through the biopsy needle and reaspirated for subsequent microbiological analysis. Aspirates obtained following core biopsies also can be sent for microbiological analysis: there is always bleeding at the core biopsy site, so that blood can be aspirated and placed in a sterile container. Alternatively, the aspiration biopsy can be performed prior to the core biopsy procedure. These two techniques have been shown to be complementary and to increase the diagnostic accuracy of the percutaneous biopsy procedure.16 The histological features of cell structure and microarchitecture may provide a specific cytological diagnosis. A positive fine-needle aspirate can obviate a more aggressive biopsy procedure, thereby reducing the likelihood of a procedure-related complication (Figure 5.16). Furthermore, the core biopsy also can be used to produce a touch preparation for immediate cytological analysis.19 These procedures in combination can minimize the possibility of obtaining a specimen that is too small for analysis. A spine biopsy procedure may be discontinued when a positive aspirate is identified by the cytopathologist, or when a set of three fine-needle aspirations and three bone and/or soft tissue cores has been obtained. Other factors, such as small lesion size, limited lesion access, or the occurrence of a complication may require discontinuation of the biopsy procedure at the discretion of the operator. Specific instances do occur in which percutaneous biopsy may be unsuccessful, yielding either no specimen or one that proves to be nondiagnostic. The bony elements of the vertebrae consist of round, hard surfaces. Securing purchase on their normal hard cortex can be difficult when the target lesion lies deep into normal bone. Sclerotic or osteoblastic lesions can be quite difficult to sample (Figure 5.17).7 At the other end of the lesion spectrum are heterogeneous lesions that are predominantly either cystic or necrotic. Despite multiple attempts, it may not be possible to harvest a satisfactory specimen from these lesions. Lesions with variable histology from one area to another, such as cartilaginous tumors, also can cause a diagnostic dilemma. Fortunately, these diagnostic challenges are infrequent. More often, one is unable to retain a specimen within the bone biopsy needle chamber after successful entry into the substance of an osseous lesion. Several maneuvers
Biopsy Techniques
Figure 5.16. Intraspinal biopsy. Fine-needle aspiration technique was used to sample (A) a cystic astrocytoma (arrow) of the spinal cord and (B) a solid astrocytoma drop metastasis within the lumbar spinal canal (arrow). (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
can be attempted to obtain a specimen. Slight, gentle rocking of the needle may allow separation of the specimen from the parent bone. If the lesion is large enough and there is a margin of safety, then advancing the biopsy needle slightly may enable retention of the bone core
95
96
Chapter 5 Image-Guided Percutaneous Spine Biopsy
Figure 5.17. (A) Axial CT image obtained during a cervical spine biopsy of a sclerotic vertebral body lesion (arrow) shows a guide needle in place. (B) The bone needle deflected across the hard surface of this sclerotic lesion and was advanced into the opposite side of the vertebral body. The needle tip is located just medial to the foramen transversarium (arrow) and anterior to the right neural foramen. The patient did not experience any adverse sequelae despite this suboptimal needle placement. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
Biopsy Techniques
Figure 5.18. Axial CT image demonstrates a Craig bone biopsy needle with its tip located in the substance of a lytic endplate lesion (arrow). Smaller gauge needles were unable to provide satisfactory amounts of tissue. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science+Business Media, 2004).
within the chamber of the biopsy needle. Applying suction to the biopsy needle with a 20-mL syringe may also facilitate a successful biopsy. Some single-pass bone biopsy needles come with an inner cannula that is partially truncated near its tip to trap the bone core within the parent needle chamber. Alternatively, if the sample size remains unsatisfactory for diagnostic purposes, a larger gauge needle system such as the Craig system can be used to obtain a specimen (Figure 5.18). Other reasons for a nondiagnostic result include biopsies that are limited either by small lesion size or because too few passes were made with the biopsy needle. Hypervascular lesions can be difficult to sample, since the brisk bleeding that can potentially occur with the initial access to the lesion can terminate the procedure. The intraosseous blood that is often aspirated during bone biopsy is sometimes erroneously discarded. This osseous blood should be considered to be a biopsy specimen and should be submitted for pathological analysis, since it is possible to diagnose malignancy from this tissue.20 Occasionally, a discrepancy in accounting for vertebral levels between different modalities causes the wrong vertebral levels to be sampled. Many spine lesions are identified on MRI, yet the percutaneous biopsy procedure is performed either with fluoroscopy or with CT. In certain situations, lesion conspicuity may be so much decreased with the latter modalities that optimal sampling is compromised. With respect to infectious spondylitis, the common reason for a nondiagnostic biopsy
97
98
Chapter 5 Image-Guided Percutaneous Spine Biopsy
result is that patients are already being treated with antibiotics at the time of the procedure. Other reasons for a nondiagnostic biopsy result in spine infection include a failure to perform the correct microbiological testing (such as not performing an acid-fast bacillus stain or culture), dismissing as contaminants unusual microbes that may in fact be the causative agents, improper specimen handling or transport (e.g., not using anaerobic culture media when these organisms are suspected), or failing to follow specific cultures (e.g., Mycobacterium tuberculosis) for an extended period of observation. To optimize the success of the biopsy procedure, the radiologist must communicate his or her clinical concerns to either the pathologist or the microbiologist. In the case of a suspected neoplasm, the clinical information and the radiological differential diagnosis should be communicated to the interpreting pathologist. The more useful the data shared with the pathologist and/or the microbiologist, the greater the likelihood of arriving at the correct diagnosis. This is the equivalent of a radiologist’s request for the appropriate clinical history from the referring clinician whenever imaging studies are to be performed or interpreted. For instance, if a patient is undergoing a biopsy to test for possible metastatic breast cancer, it is helpful to inform the pathologist that the woman had a mastectomy last year at the same institution. Similarly, it is important to inform the microbiologist whether the patient is already on intravenous antibiotics or that a specific organism, such as Mycobacterium tuberculosis, is causing concern.
Soft Tissue Biopsy Percutaneous image-guided biopsy techniques are also utilized to approach and diagnose spinal and paraspinal soft tissue abnormalities. Whether aspirating a collection or obtaining tissue from a neoplasm, image guidance provides a high margin of safety and diagnostic efficacy. CT guidance is typically used to target soft tissue lesions that would be occult under fluoroscopy. CT has been shown to have a diagnostic yield of about 80% and an accuracy of over 95% for bony lesions, and 70% and 93%, respectively for soft tissue lesions.21 The pre-procedural evaluation and regimen are the same as for vertebral or other deep bone biopsy; pre-procedure clinical assessment and laboratory studies are obtained in the interest of safety. The paraspinal space is large in scope and dimension, and each region poses its unique set of challenges to the approach. Coaxial techniques should be utilized when possible; reports of tract seeding from TB, sarcomas, and other etiologies have been well documented. The use of coaxial needle systems22 and core biopsies23 along with CT guidance has been shown to be safe, effective, and useful. The location of the pathology and the adjacent vital structures will dictate the trajectory of the needle or device that can be used. Innovation is often required, but patient safety should remain paramount. The cervical anatomy is particularly complex and may require a contrast-enhanced CT for planning in order to identify the major vascular structures (Figure 5.19).
Soft Tissue Biopsy
Figure 5.19. Fifty-year-old woman with dysphagia. Carcinoid tumor. Axial CT image (A) demonstrates a well-defined prevertebral mass (arrow) deep to the trachea and between the carotid sheath and esophagus. CT-guidance allows for a direct anterior approach (B) that avoids the carotid, trachea, and esophagus.
99
100
Chapter 5 Image-Guided Percutaneous Spine Biopsy
Figure 5.20. Sacroiliac osteomyelitis and abscess. Posterior approach. Sixtyyear-old man presenting with increasing pain 1 month after a fall that resulted in sacral fractures. Axial T1-weighted image following gadolinium administration (A) demonstrates enhancement within the left sacral ala and iliac bone consistent with osteomyelitis. A low signal pelvic fluid collection anterior to the left SI sacroiliac joint represents abscess surrounded by a thick, enhancing wall (arrow). Axial CT image (B) demonstrates CT-guided posterior transarticular approach to abscess through the SI joint. Drainage of the collection resulted in immediate improvement of symptoms. The patient was pain free at 6 month follow-up. Methacrylate sacroplasty was not required.
Conclusion
In the thoracic region, care must be taken to avoid a pneumothorax. In the lower thoracic and upper lumbar regions, a posterior oblique approach should take into consideration and avoid the kidneys; this is even more important when performing a fluoroscopically-guided biopsy. The vast number of cases reported in the literature and the comprehensive vertebral review of vertebral biopsy techniques provided in this chapter preclude a consideration of all of the potential approaches to soft tissue lesions; however, some unique approaches and techniques are presented (Figures 5.20 and 5.21).
Postoperative Care Immediately following the procedure, a sterile dressing is placed over the skin entry site(s). The patient is observed in recovery for 2–4 h, depending on the type of anesthesia that was used. Monitoring of the patient, including vital signs, is continued during the recovery period. The puncture site is periodically observed for signs of active bleeding or for expanding hematoma. Strict bed rest is maintained throughout the recovery period. When the patient is judged to be stable, either by the radiologist who performed the procedure or by the anesthesiologist who sedated the patient, he or she is discharged from the recovery area: an outpatient goes home, an inpatient to a hospital room. An instruction sheet with attention to wound care and observation should be given to all outpatients. All patients should be informed that the test results might not be available for several days owing to specimen processing requirements. More importantly, patients also should be made aware of the small, but real, possibility that the test results may be nondiagnostic, whereupon a repeat percutaneous biopsy or an open biopsy may be required. Adequate follow-up on all biopsy procedures is essential, and the final results should be communicated to the referring clinician(s).
Conclusion Image-guided percutaneous spine biopsy is a procedure that can be performed safely and efficiently by radiologists. The procedure is performed to determine accurately the composition of abnormal tissue. The information obtained from the biopsy procedure can be used to guide patient management. The radiologist is part of a team that includes the patient, the referring clinician, and a pathologist. Optimal communication among the team members will increase the likelihood of a successful procedural outcome.
101
Figure 5.21. Presacral abscess. Sagittal (A) and axial (B) enhanced T1-weighted MR images demonstrate a large, loculated presacral pelvic abscess. Axial images (C, D, and E) acquired during CT-guided approach utilizing a curved needle to gain access to the loculated collection. Frontal (F) and lateral (G) radiographs demonstrate the trajectory of the curved needle.
Conclusion
Figure 5.21. Continued.
103
Figure 5.21. Continued.
Conclusion
Figure 5.21. Continued.
References 1. Robertson RC, Ball RP. Destructive spine lesions: diagnosis by needle biopsy. J Bone Joint Surg 1935;57:749–758. 2. Siffert RS, Arkin AM. Trephine bone biopsy with special reference to the lumbar vertebral bodies. Am J Bone Joint Surg 1949;31:146–149. 3. Adapon BD, Legada BD, Lim EVA, Silao JV Jr, Dalmacio-Cruz A. CT-guided closed biopsy of the spine. J Comput Assist Tomogr 1981;5:73–78. 4. Ottolenghi CE. Aspiration biopsy of the spine. Am J Bone Joint Surg 1969;51:1531–1544. 5. Murphy WA, Destouet JM, Gilula LA. Percutaneous skeletal biopsy 1981: a procedure for radiologists – results, review, and recommendations. Radiology 1981;139:545–549. 6. DeSantos LA, Lukeman JM, Wallace S, Murray JA, Ayala AG. Percutaneous needle biopsy of bone in the cancer patient. Am J Roentgenol 1978;130: 641–649. 7. Kattapuram SV, Rosenthal DI. Percutaneous biopsy of skeletal lesions. Am J Roentgenol 1991;157:935–942. 8. Froelich JJ, Saar B, Hoppe M, Ishaque N, Walthers EM, Regn J, Klose KJ. Real-time CT-fluoroscopy for guidance of percutaneous drainage procedures. J Vasc Int Rad 1998;9:735–740. 9. Onik G, Cosman ER, Wells TH, Goldberg HI, Moss AA, Costello P, Kane RA, Hoddick WI, Demas B. CT-guided aspirations for the body: comparison of hand guidance with stereotaxis. Radiology 1988;166:389–394.
105
106
Chapter 5 Image-Guided Percutaneous Spine Biopsy 10. Palestrant AM. Comprehensive approach to CT-guided procedures with a hand-held guidance device. Radiology 1990;174:270–272. 11. Rice KM, Cronan JJ. Practical CT stereotactic biopsy techniques. Appl Radiol 1995;24:35–38. 12. Brugieres P, Gaston A, Heran F, Voisin MC, Marsault C. Percutaneous biopsies of the thoracic spine under CT guidance: transcostovertebral approach. J Comput Assist Tomogr 1990;14:446–448. 13. Laredo JD, Bard M. Thoracic spine: percutaneous trephine biopsy. Radiology 1986;160:485–489. 14. Kattapuram SV, Rosenthal SI. Percutaneous biopsy of the cervical spine using CT guidance. Am J Roentgenol 1987;149:539–541. 15. Geremia GK, Charletta DA, Granato DB, Raju S. Biopsy of vertebral and paravertebral structures with a new coaxial needle system. Am J Neuroradiol 1992;13:169–171. 16. Schweitzer ME, Gannon FH, Deely DM, O'Hara BJ, Juneja V. Percutaneous skeletal aspiration and core biopsy: complementary techniques. Am J Roentgenol 1996;166:415–418. 17. Kreula J, Virkkunen P, Bondestam S. Effect of suction on specimen size in fine needle aspiration biopsy. Invest Radiol 1990;25:1175–1181. 18. Kreula J. Effect of sampling technique on specimen size in fine needle aspiration biopsy. Invest Radiol 1990;25:1294–1299. 19. Renfrew DL, Whitten CG, Wiese JA, el-Khoury GY, Harris KG.CT-guided percutaneous transpedicular biopsy of the spine. Radiology 1991;180:574–576. 20. Hewes RC, Vigorita VJ, Freiberger RH. Percutaneous bone biopsy: the importance of aspirated osseous blood. Radiology 1983;148:69–72. 21. Puri A, Shingade VU, Agarwal MG, Anchan C, Juvekar S, Desai S, Jambhekar NA. CT-guided percutaneous core needle biopsy in deep seated musculoskeletal lesions: a prospective study of 128 cases. Skeletal Radiol 2006:35(3):138–43. 22. Yaffe D, Greenberg G, Leitner J, Gipstein R, Shapiro M, Bachar GN. CT-guided percutaneous biopsy of thoracic and lumbar spine: A new coaxial technique. Am J Neuroradiol 2003:24(10):2111–3. 23. Michel SC, Pfirrmann CW, Boos N, Hodler J. CT-guided core biopsy of subchondral bone and intervertebral space in suspected spondylodiskitis. Am J Roentgenol 2006;186(4):977–80.
6 Discography Aaron Calodney and Duane Griffith
Introduction Discography has become an indispensable diagnostic tool since its introduction in the 1940s to diagnose lumbar disc herniation.1,2 Today, discography is used most often to identify painful internally disrupted discs.3 Discography is unique in that it is the only diagnostic technique that directly correlates a patient’s symptoms with disc morphology.4 It can be thought of as similar to palpation: an important component of physical examination.5 Tenderness elicited on palpation is analogous to pain provocation on discography. When evaluating a suspected spinal origin of pain, it is critically important to assess the significance of pathological findings on imaging studies and to determine whether those findings correlate with a patient’s symptoms. Discography is distinctive in its ability to make this determination. Discography is used in the lumbar,6–19 thoracic,20–24 and cervical25–34 regions to assess pain that is suspected to be of discogenic origin. Discogenic pain is mechanical in nature (exacerbated with activity and relieved by rest) and is felt primarily in an axial distribution. Performing discography involves the injection of radiographic contrast into the nucleus of the intervertebral disc in order to visualize disc morphology (Box 6.1).4 The discs are then pressurized and the patient’s response is recorded, a separate process referred to as disc stimulation.17 In theory, the composite of disc stimulation and discography is called provocation discography. However, this combined procedure is commonly referred to simply as discography and will be referred to by this name in the remainder of this text. The results of discography must be interpreted along with the patient’s history, physical exam, imaging studies, and response to previous diagnostic injections when formulating a specific diagnosis and treatment plan. Formal investigations have shown that discography performed by skilled, knowledgeable, and experienced proceduralists can substantially improve both surgical and nonsurgical treatment outcomes.35–40 From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_6, © Springer Science + Business Media, LLC 2010
107
108
Chapter 6 Discography
Box 6.1 Discography is the injection of contrast into the disc nucleus for radiographic evaluation. Disc stimulation involves pressurizing the disc and recording the response of the patient. Provocation discography is the combination of discography and disc stimulation. Improvements in spinal imaging and interventional techniques have led to better understanding of the origins of spinal pain. High quality MR imaging6,7,41,42 is capable of revealing degenerative changes involving multiple spinal structures at multiple segments. Such abundance of information has led to an increasing demand for diagnostic procedures that can determine the significance of these imaging findings relative to clinical complaints. As such, discography has been the focus of increased clinical utilization and scientific investigation.
Indications Indications for discography include, but are not limited to:43 1. Evaluation of abnormal discs to assess the extent of abnormality or to correlate the abnormality with clinical symptoms; 2. Patients with persistent, severe symptoms in whom other diagnostic tests have failed to confirm the diagnosis of discogenic pain; 3. Evaluation of patients who have failed to respond to posterior fusion or to determine if a pseudoarthrosis is present; 4. Assessment of discogenic pain prior to fusion and to determine if adjacent segments are painful; 5. Assessment of discogenic pain prior to minimally invasive procedures.
Complications The two major complications of discography are discitis and neural injury (Table 6.1).44–46 The incidence of discitis is extremely low, particularly with the routine use of intradiscal antibiotics. Discitis was found to be the most frequent and significant complication of discography, with an overall incidence reported as less than 0.15% per patient and less than 0.08% by disc injected.47 Discitis usually presents 2–4 weeks after the procedure with fever, chills, and back pain. Significantly increased pain is the cardinal presenting feature. Intradiscal antibiotics and intravenous antibiotics have both been used prophylactically to prevent discitis. The addition of gentamicin, cefazolin, or clindamycin to isohexol has been proven effective in inhibiting bacterial growth.48 Each antibiotic was proven to retain its efficacy against laboratory strains of Escherichia coli, Staphylococcus aureus, and Staphylococcus
Conscious Sedation
Table 6.1. Complications of discography Major complications Discitis Nerve injury Other reported complications Bleeding Bruising Spinal cord injury Myelopathy Epidural abscess Contrast allergy Subarachnoid puncture Vagal response Bowel perforation Increased pain
Box 6.2. Intradiscal antibiotics Gentamicin 1 mg per ml of contrast can be obtained by mixing 0.375 ml of gentimicin (40 mg/ml) into 15 ml of isohexol.
epidermidis. Appropriate dosing of intradiscal gentamicin and cefazolin ranges from 1 to 10 mg/ml.48 The use of cefazolin carries the additional risk of seizures if injected intrathecally.49 The clinical practice of the authors is to mix 0.375 ml of gentamicin 40 mg/ml with 15 ml of isohexol to result in a final gentamicin concentration of 1 mg/ml (Box 6.2). This solution is then utilized for disc injection. The use of proper technique can avoid neural injury. Placing a needle near the ventral ramus will be painful to a conscious patient. A paresthesia is a clear indication to the discographer to withdraw and redirect the needle. Other complications are similar to those of other needle-based spine procedures and include bleeding, bruising, spinal cord injury, myelopathy, epidural abscess, contrast allergy, subarachnoid puncture, vagal response, bowel perforation, and increased pain. Discography has not been associated with chronic changes to the studied discs or the endplates.43
Conscious Sedation Intravenous conscious sedation is a common method of controlling pain and anxiety in patients undergoing invasive procedures.50 The amount and route of sedation are at the discretion of the discographer;
109
110
Chapter 6 Discography
however, the patient must be capable of responding appropriately during the study. Discography should be performed in a procedure suite with monitoring equipment that includes an electrocardiogram (ECG), a continuous pulse oximeter, and noninvasive blood pressure monitoring. Some proceduralists advocate that discography should be performed on a fully alert, unsedated patient. The authors’ experience has been that intravenous conscious sedation can be used during discography without confounding the test results. The patient is comfortable during the placement of the needles and can still be responsive during the provocation portion of the procedure. The authors’ routine practice involves the use of intravenous midazolam and fentanyl – both short acting agents that can be reversed quickly if oversedation occurs. It is important to titrate sedation on an individual basis; constant verbal communication with the patient can determine the level of sedation. If the patient is sedated to the point that he/she cannot respond during provocation, reversal agents can be used to reduce the sedation. This is rarely necessary in the authors’ practice.
Discogram Reporting The formal reporting of discography should be performed immediately following any procedure so that important details can be recalled. In the authors’ practice, a standardized format is used to aid in the communication of the data between the discographer and other physicians. The information that is critical to report is: 1. Disc levels studied and morphology; 2. Injection volume, opening pressure, pressure that elicited pain, and peak pressure (for lumbar discography; pressure is not generally measured in thoracic or cervical studies); 3. End-point characteristics (firm, moderate, or soft) are reported in cervical and thoracic discography in lieu of manometry; 4. Location and characteristic of the pain provoked, and determination as to concordant or nonconcordant pain; 5. Opinion of the discographer of the validity of the study based upon pain tolerance, behavior during the study, and consistency of pain responses to repeat pressurization of positive discs and lack of pain response to sham injection.
Lumbar Discography Lumbar MRI provides recognizable correlates with discogenic pain in patients with low back pain. This includes the high-intensity zones seen in the annulus fibrosis on T2-weighted magnetic resonance imaging.2,6,7,41 Ito and colleagues42 studied 101 lumbar discs in 39 patients with MRI and discography. A radial tear on MRI was highly sensitive (87%) in detecting a painful disc, but this was a common MRI finding and was not specific, having a positive predictive value (PPV) of only 42.6%. Moderate loss of nuclear intensity and moderate disc space
Lumbar Discography
Figure 6.1. Normal lumbar disc anatomy. (Reprinted with permissions from Raj PP. Intervertebral Disc: Anatomy-Physiology-Pathophysiology-Treatment. Pain Practice 2008 Mar; 8(1): 18–44. Wiley-Blackwell Publishers.)
narrowing was not a reliable predictor of concordant pain; both specificity and PPV were low. Severe loss of nuclear intensity was a relatively reliable predictor of dehydrated discs, while severe disc space narrowing was a reliable predictor of concordant pain. The presence of a high-intensity zone was relatively reliable with high specificity (89.7%). Equally important, normal T-2 signal intensity was 95.7% specific and had a PPV of 97.3% for predicting normal discography. Lumbar discs with normal signal intensity are highly unlikely to elicit pain on discography. Anatomy The intervertebral disc is situated between the cartilaginous endplates of the vertebral body above and below. It is composed of an outer layer, the annulus fibrosis, and an inner gelatinous core, the nucleus pulposus (Figure 6.1).51 The annulus fibrosis is formed by concentric lamellae of fibrocartilage (Figure 6.2). The nucleus pulposus contains primarily collagens, proteoglycans, and water.52 The nucleus functions as a shock absorber during axial loading. The intervertebral disc has a complicated innervation. The outer annulus is innervated to a depth of 3.5 mm.51,53–55 Most of the nerve endings are concentrated dorsally and posterior-laterally.26,52 The lumbar discs are innervated by branches from the lumbar ventral rami, the gray rami communicans, and the sinuvertebral nerve (Figure 6.3).52 The density of receptors found within the lumbar end plates and the annulus are similar.56,57 The endplate innervation is most dense centrally, near the nucleus.
111
112
Chapter 6 Discography
Figure 6.2. Innervation of lumbar disc herniation. (Reprinted with permissions from Raj PP. Intervertebral Disc: Anatomy-Physiology-PathophysiologyTreatment. Pain Practice 2008 Mar; 8(1): 18–44. Wiley-Blackwell Publishers.)
This may be one mechanism for pain reproduction during provocation discography. Painful discs removed during surgery have been compared to control discs.56 In a degenerative disc, this histological arrangement changes and nerve fibers have been observed into the inner annulus. The differences observed in these painful discs included the formation of a zone of vascularized granulation tissue from the outer annulus into the inner annulus along the edges of fissures. Growth of nerves deep into the annulus are along this zone of granulation tissue and could account for pain reproduction during discography. Procedure Prior to discography, a history and a physical exam should be performed. The history should focus on the patient’s usual pain condition, previous diagnostic and therapeutic interventions, past medical history, and drug allergies. The patient’s CT and MR imaging should be reviewed to determine the levels to be studied. The patient needs to be informed of the risks and benefits of discography, and it should be made clear to the patient that his/her response to disc stimulation is the basis for the test results. Patients are generally positioned prone on a fluoroscopy table for lumbar discography. This allows needle entry from either side. Foam pillows or pads are placed beneath the upper abdomen and lower chest to reduce the lumbar lordosis. Some discographers prefer to place the patient in an anterior-oblique position, assuming that the tube is perpendicular and the beam is directed anterior to posterior. This keeps the image intensifier out of the way during initial needle placement. Cleansing the skin is vital to reduce the risk of discitis. A skin preparation is used to reduce the bacterial load since the most common
Lumbar Discography
113
Figure 6.3. Anatomic relationship of the lumbar disc, end plate, and nerve root. (Netter Anatomy Illustration Collection, © Elsevier, Inc. All Rights Reserved.)
bacterial contaminants are S. aureus/epidermidis.44–48 Povidone iodine solution is used unless the patient has a known allergy, in which case a noniodine based soap or alcohol-based solution can be utilized. Surgical draping is required to maintain a sterile field during the
114
Chapter 6 Discography Figure 6.4. AP view with the L4-5 end plates parallel to fluoroscopy beam.
procedure, and a full fenestrated drape should be used. The proceduralist must maintain sterile technique throughout the procedure. At no time should the needle tip be touched with the gloved hand; sterile gauze can be used to manipulate the needle tip. Each disc should be accessed with an unused sterile needle. The discs investigated should include those found to be abnormal on screening MR imaging and at least one control level above and/or below. The patient should be blinded to the level being investigated. The fluoroscope is used to obtain an anterior–Posterior (AP) image of the target level. The fluoroscope is then angled cephalad or caudad until the image beam is parallel to the planes of the inferior and superior endplates that surround the target disc (Figure 6.4). The disc should be approached from the side opposite the patient’s typical pain to reduce confusion as to the source of any provoked pain. If the patient has midline or bilateral pain, the disc can be approached from either side. The fluoroscope is next rotated obliquely toward the side of needle entry until the facet joint line is in the midline of the target disc. The needle is to be passed just lateral to the lateral aspect of the Superior Articular Process (SAP) at the level of the target disc (Figures 6.5 and 6.6). Entry into the L5-S1 disc can be challenging due to the overlying iliac crest. A cephalad tilt to the fluoroscopy beam will aid in displacing the iliac crest inferior to the L5-S1 disc space. The target window is defined by the SAP medially, the iliac crest laterally, the sacrum inferiorly, and the inferior endplate (ring apophysis58) of L5 superiorly (Figure 6.7).
Lumbar Discography
115
Figure 6.5. L3-4 oblique view. The SAP is in the midline of the disc space. The needle is slightly lateral to the SAP.
Figure 6.6. L3-4 oblique view with the needle advanced just lateral to the SAP. The needle is parallel to the fluoroscope beam.
A curved tipped needle can be used to avoid the iliac crest while obtaining disc access (Figure 6.8). A skin wheal is made over the needle insertion site with local anesthetic. Lidocaine or xylocaine has the advantage of a more rapid onset as compared with bupivicaine for localizing the skin. Mathis59 describes a “no-sting” solution of 0.5% lidocaine made by combining lidocaine 4%
116
Chapter 6 Discography
Figure 6.7. L5-S1 oblique view. The fluoroscopy beam has been angled cephalad to displace the iliac crest inferior. The existing needle is placed in the L4-5 disc space.
Figure 6.8. L5-S1 oblique view with needle inserted. Illiac crest is inferior, SAP medial to the needle. The curved tip needle can be used to avoid the iliac crest.
Lumbar Discography
(4 cc), lactated ringer’s (24 cc), and bicarbonate (2 cc) for a total volume of 30 ml of local anesthetic mixture. This can be used to anesthetize the skin entry site during discography or any other injection procedure. Each target disc should be accessed with a fresh sterile needle. Both single needle and two needle techniques have been described.60 The two needle technique may reduce the risk of discitis as the skin is not punctured with the same needle that is placed intradiscally.61 Single needle techniques have been documented to be safe,20 particularly with the use of intradiscal antibiotics, and this is the technique that the authors currently employ. The authors’ standard needle length is 7 inches, although shorter needles can be used in slender patients. A longer needle, up to 10 inches, may be needed to access the L5-S1 disc in a large patient. A 22- or 25-gauge needle can be used to obtain disc access; however, 25-gauge needles can be difficult to manipulate due to their compliance. The needle is inserted through the skin parallel to the fluoroscopic beam and advanced just lateral to the SAP. If bone obstructs needle placement, the discographer must determine if the SAP or an endplate has been contacted and make the proper needle correction. Once the needle has been advanced distal to the SAP, the fluoroscopy beam is rotated to obtain a lateral image (Figure 6.9). Care must be taken when crossing the level of the intervertebral foramen not to strike the ventral ramus. If the patient complains of parasthesia during this portion of the procedure, the needle must be slightly withdrawn and redirected. The needle is then advanced, and, if no parasthesia is elicited, the next structure that the needle will encounter is the disc annulus. A firm
Figure 6.9. Lateral view with needles inserted in the L4-5 and L5-S1 discs. The needle is advanced into the center of the disc as viewed laterally.
117
118
Chapter 6 Discography
Figure 6.10. AP view with needle in L4-5. Needle placement should be midline in the AP and the lateral. Note larger gauge needle used as introducer for FAD catheter.
resistance will be felt by the discographer at this point. It is common for the patient to experience a dull ache in the lower back or buttock as the needle passes through the annulus. The needle is then advanced into the center of the disc and final needle position is confirmed with both lateral and AP imaging (Figure 6.10). After proper placement of the needle into the target disc, the stylette is removed from the needle. The needle is connected to a syringe that will inject contrast mixed with antibiotic. If the patient has a known allergy to contrast, either saline or gadolinium62 mixed with antibiotic can be injected. At least one painless disc must be identified as a control level during provocation discography in order to validate the procedure. If all discs studied are painful, the discogram can be considered invalid, and an adjacent level should be tested in order to identify a control. The diagnosis is stronger if the concordant disc displays a grade 3 fissure or greater on a postdiscography CT scan. The diagnosis is most robust if a single disc demonstrates concordant pain production and the two adjacent discs are nonpainful.63 The volume and pressures are recorded while contrast is injected. The patient’s response to the injection is noted.64 In a normal disc, contrast remains in the nucleus and appears as a “cotton-ball” (Figures 6.11 and 6.12). If the patient experiences pain with injection, the location, severity, and quality are documented. Transient pain can be provoked when fissures are opened. To be truly positive, the pain must be sustained during injection.36 “Concordant pain” is pain during provocation that replicates the patient’s usual pain pattern. “Nonconcordant pain” is pain during
Lumbar Discography
Figure 6.11. Lateral view of a normal L4-5 disc. The L5-S1 disc demonstrates an anterior and posterior tear and posterior bulge.
Figure 6.12. AP view with normal L3-4 disc morphology and degenerative L4-5 disc.
119
120
Chapter 6 Discography
provocation that does not replicate the usual pain pattern. Disc morphology, including disc height, tears, and leaks, are also recorded. A confirmatory repressurization of a concordant disc or indeterminant disc is routinely performed to reconfirm the discographer’s findings. Another method used to verify the consistency of the patient’s response to disc pressurization is the use of a sham injection. The patient is told that the disc is being injected while the syringe is held in the operator’s hands. Any pain response is noted. Patients are expected to survive sham injection without response and to respond consistently to repressurization. Injection is continued until: 1. Pain is reproduced at a level of 6/10 or greater; 2. Intradiscal pressure >50 psi above opening pressure in a disc with a grade 3 annular tear; 3. 4.0 ml of volume is reached; 4. 80–100 psi is reached in a normal appearing disc.17,65 The opening pressure, pressure at onset of pain, and peak pressure are also recorded.62 Injection of local anesthetic into a painful disc can reduce the severity of pain after discography. If the disc has a large tear or epidural leak, it can affect testing at adjacent levels due to spread of local anesthetic. Intradiscal steroids have been injected following discography, but not proven to be of long-term value.36,66 Routine postdiscography CT is ordered if the discogram results are likely to lead to surgical intervention. Postdiscography MRI can be used if gadolinium is employed during the study.62 In the authors’ practice, gadolinium is occasionally used in patients with a strong history of contrast allergy. Intrathecal injection of gadolinium carries the same risks as iodinated ionic contrast materials and must be avoided.66,67 CT can differentiate between annular tears and annular disruption.67 The Dallas Discogram Description is a classification method for standardizing CT discogram results. It has been modified twice: by Bogduk17 in 1992 and by Schellhas and colleagues7 in 1996 (see Figure 6.13). Grade 1 discs are rarely painful, while 75% of grade 3 tears are associated with exact or similar pain reproduction.63,68 There are five categories that define the internal disc structure: Grade 0: Normal disc morphologically; Grade 1: Contrast spreads radially along a fissure to the inner 1/3 of the annulus Grade 2: Contrast spreads into middle 1/3 of the annulus Grade 3: Contrast spreads to the outer 1/3 of the annulus, involving <30° of the disc circumference Grade 4: Contrast spreads to the outer 1/3 of the annulus, involving >30° of the disc circumference Grade 5: Full thickness tear with extra-annular leakage into epidural space. Postprocedure Care Routine discography is an outpatient procedure. After discography, the patient is taken to the recovery room and observed for at least 30 min.
Lumbar Discography
Figure 6.13. Disc Morphology. Grade 0: Normal disc morphologically. Grade 1: Contrast spreads radially along a fissure to the inner 1/3 annulus. Grade 2: Contrast spreads into middle 1/3 of the annulus. Grade 3: Contrast spreads to the outer 1/3 of the annulus, involving <30° of the disc circumference. Grade 4: Contrast spreads to the outer 1/3 of the annulus, involving >30° of the disc circumference. Grade 5: Full thickness tear with extra-annular leakage into epidural space. (Adapted with permission from Falco FJE, Zhu , Irwin L, Onyewu CO, Kim D. Lumbar Discography. In Manchikanti L, Singh V (eds): Interventional Techniques in Chronic Spinal Pain. Paducah, Kentucky: ASIPP, 2007.)
Pain medications can be given if the patient is experiencing significant discomfort. After the patient has recovered from sedation and pain is controlled, he/she can be transferred for a postdiscogram CT. A small supply of oral pain medication can be prescribed, but it is not often needed. The patient is given printed discharge instructions, which include warning signs of discitis. Exacerbation of pain is common and can last for several days. Warning signs of discitis, including fever, night sweats, and significantly increased pain, are discussed. The patient is given contact information for the discographer in case of postprocedural problems. Manometry In the authors’ practice, a manometer is used during lumbar discography to measure intradiscal pressures and is rapidly becoming the standard of care. Manometry is not generally used in cervical or thoracic discography. Manometry is recommended during lumbar studies to improve the consistency of results and provide additional objective data.36 Usual opening pressures, where contrast is first seen entering the disc, are 5–25 psi, depending on the extent of degeneration. An opening pressure greater than 30 psi suggests that the needle tip may be in the annulus and needs to be repositioned. The use of pressure monitoring can reduce the incidence of false positive results by
121
122
Chapter 6 Discography
decreasing the likelihood of overpressurization of the disc.51 O’Neill and Kurgansky69 studied 838 discs retrospectively in regards to pressure and pain responses. The conclusion of this analysis was that pressurization of the disc more than 50 psi above the opening pressure is associated with a very high false positive rate. Many discs can be rendered painful by a powerful squeeze on the syringe. The use of a manometer clearly identifies not only the opening pressure, but also the pressure at which a painful response is produced. If a disc is painful at greater than 50 psi over opening pressure, the response cannot be distinguished from stimulation of a normal disc and should not be considered positive. Discs that fall into this category or discs that produce pain only after injection of >4.0 ml of contrast are considered indeterminate. Carrino and colleagues70 proved that normal discs can be painful during discography, but, as a rule, require high pressures on injection and the provoked pain is mild. Derby and colleagues36 divided discs into four categories: 1. Normal disc, no pain 2. Chemically sensitive disc, pain <15 psi above opening pressure 3. Mechanically sensitive disc, pain >15 psi and <50 psi above opening pressure 4. Indeterminate disc, pain >51 and <90 psi above opening pressure. The chemically sensitive category had better outcomes with interbody fusion when compared with intertransverse fusion or nonoperative treatment. Currently, there are several manometers available for use in the United States. These systems vary from a simple syringe with an analogue pressure gauge, syringes with digital displays of pressure and volume, to a system that mechanically injects contrast at a predetermined rate while measuring volume and pressure changes. Functional Anesthetic Discography Functional Anesthetic Discography71 is an additional test that can be carried out following provocation discography. It is based on analgesic discography, where local anesthetic is injected into a painful disc and pain relief is assessed. The theory is similar to diagnostic injections of facet joints or sacroiliac joints. A 17-gauge introducer needle is used for disc access. After the provocation discogram, a balloon tipped catheter is placed into the center of the disc through the needle (Figure 6.14).72 The balloon is inflated with contrast, and a stopcock is closed to secure the catheter in the nucleus (Figure 6.15). In the recovery area, the patient is instructed to perform specific activities and positions that typically elicit pain, while the Visual
Figure 6.14. The intradiscal catheter is placed through the needle and has a distal balloon that is inflated with contrast to maintain placement. (Courtesy of Kyphon-Medtronic, Sunnyvale, CA. Copyright © Medtronic Spine, LLC, Used with permission.)
Lumbar Discography
Figure 6.15. AP view with FAD catheter inflated with contrast in the L5-S1 disc space. Contrast should be used to confirm balloon inflation. (Courtesy of Kyphon-Medtronic, Sunnyvale, CA. Copyright © Medtronic Spine, LLC, Used with permission.)
Analog Scores (VAS) are recorded. In the authors’ practice, 1 ml of 4% lidocaine is injected into the painful disc through the balloon-tipped catheter after baseline pain scores have been recorded. The same measurements are carried out 8–10 min later, and the change in VAS is recorded. Provocation discography facilitates diagnosis by eliciting pain, and FAD helps clarify the diagnosis by relieving pain and improving mobility during testing. Each painful disc can be studied sequentially and individually. The results of the provocation discography are compared with the FAD results to produce a final report. After the FAD is completed, the balloon on each catheter is deflated and the catheter is removed. Alamin and coworkers reported that 50% of patients (n = 32) had confirmatory findings on FAD after a positive discogram.73 The patients with confirmatory FAD underwent fusion and were followed for at least 3 months. The mean preoperative Oswestry score was 58.5, and the mean postoperative was 26.5. The mean preoperative VAS score for back pain was 7.2, and the mean postoperative VAS was 3.1. Luchs and colleagues74 reported on 19 consecutive patients in whom FAD was used. Nineteen of 29 discs injected (65.5%) showed
123
124
Chapter 6 Discography
a favorable response to anesthetic injection. Of these discs, all 19 had a positive provocation discogram, and 18 had pathology on CT scan. Ten discs demonstrated no pain relief on FAD. Eight of these discs were positive on provocative discography, and six had pathology on CT. Lumbar Discography in Surgical Outcomes Discography is often used to identify target levels prior to open surgical procedures. Several studies have investigated the benefits of preoperative lumbar discography on surgical outcomes. Colhoun and colleagues39 compared the morphological appearance and pain provocation during discography with surgical outcome. Eighty-nine percent of patients with abnormal disc morphology and pain provocation had successful surgical outcomes. Patients with abnormal disc morphology without pain provocation had successful outcomes only 52% of the time. Blumenthal and colleagues37 reported a 73% fusion rate, with a 74% clinical success rate using discography prior to lumbar anterior interbody fusion. Provocation discography utilized prior to anterior lumbar interbody fusion resulted in an 86.1% clinical success rate, with a fusion rate of 88.9% as measured by radiographs in the study by Newman and Grinstead.38 Wetzel and colleagues75 followed patients for 65 months after positive discography and lumbar fusion. Patients with a solid fusion had a success rate of 95.6%. Kozak and O’Brien35 performed ventral and dorsolateral fusions on 27 patients with primary low back pain. Provocation discography was performed on every patient to determine the surgical levels. The surgical interventions included 7 one-level, 17 two-level, and 3-three level fusions. Fusion rate was 94.6%. A good clinical result in this population was achieved in 74.1% of the patients. A good clinical result was defined as 76–100% relief of back and leg pain, return to employment, no or slight restriction to physical activity, and no use of analgesics.
Thoracic Discography Thoracic discography is a diagnostic procedure utilized to determine whether a particular thoracic disc is the cause of a patient’s thoracic spinal pain. During this procedure, contrast medium is injected into the nucleus of the disc with the intent of provoking the patient’s usual pain. Although there are several publications reporting the incidence and clinical features of thoracic disc pathology, these are limited predominantly to studies of thoracic disc herniation.76–78 Axial thoracic spinal pain as the result of a painful thoracic disc has not been well studied. Thoracic discs are innervated structures and are capable of producing pain.79,80 The concept of thoracic discogenic pain is based upon the authors’ experience with lumbar disc pain and discography.23,81 Lumbar MRI provides recognizable correlates with discogenic pain in patients with low back pain including the highintensity zones seen in the annulus fibrosis on T2-weighted magnetic resonance imaging, severe loss of nuclear intensity, and severe disc space narrowing.2,6,7,41 No such correlate has been demonstrated in the
Thoracic Discography
thoracic spine. Although a high rate of abnormal findings can be seen on thoracic MR imaging in asymptomatic individuals, subtle but potentially painful annular abnormalities can be missed by MRI. This is precisely why discography is such a useful test in the workup of patients with axial thoracic spinal pain. Review of previous thoracic MR imaging studies is required prior to the procedure to rule out the presence of either spinal cord compression or cord deformity at or adjacent to any level to be studied. The discographer must have knowledge of spinal canal dimensions prior to undertaking the procedure. The authors avoid the study of any segment in which spinal cord compression and/or deformity exists. On occasion, the authors have declined the procedure altogether when asked to inject discs that are deforming the cord and there is accompanying myelopathy. Each case of cord impingement, with or without myelopathy, must be considered individually. Anatomy The innervation of the thoracic disc is analogous to that of the cervical and lumbar disc. The thoracic disc is wedge-shaped. It is thicker posteriorly than anteriorly, which contributes to the normal thoracic kyphosis. The thoracic disc has a centrally located and well defined nucleus pulposis surrounded by the annulus fibrosis, made up of a series of concentric lamellae of collagen fibers.24,79 Procedure A history and a physical examination should be performed, and imaging should be reviewed prior to the procedure. The procedure should be explained to the patient, including the need for patient participation during the provocation portion of the exam. The patient must understand that he/she must clearly describe any pain produced during disc stimulation and compare this to his or her usual pain. Informed consent should be obtained, including the possibility of pneumothorax. Thoracic discography should be performed in a procedure room that is suitable for aseptic procedures. In addition, the room should contain a table that is X-ray compatible. Needle access to the thoracic disc requires a high resolution C-arm with the capability of storing images.60,82 Monitoring equipment should include ECG, blood pressure cuff, and pulse oximeter. Supplemental oxygen and suction and standard resuscitation equipment should be available. The patient should be positioned prone on the procedure table. Monitoring, sedation, skin preparation, and draping follow the same guidelines as for lumbar discography discussed earlier in this chapter. Antibiotics can be given intravenously prior to the procedure and must be placed intradiscally during the procedure. Although cephazolin is suitable for IV use preoperatively, it should be used carefully, if at all intradiscally, as it is known to cause seizure activity and even death if inadvertently placed intrathecally. The authors recommend the use of gentamicin intradiscally. Gentamicin is mixed with contrast to a concentration of 1 mg/ml. The levels to be studied should be carefully
125
126
Chapter 6 Discography
identified by counting down from T1 in an AP view. The endplates at each level will need to be squared, ensuring that the X-ray beam is parallel to the planes of the inferior and superior vertebral body margins above and below the disc to be entered. Needle entry should be on the side opposite the patient’s pain in order to distinguish procedural pain from pain provoked by disc pressurization. Patients with midline or bilateral pain can be approached from either side. The needle is placed using an oblique view and directed to pass medial to the head of the rib and lateral to the lamina (Figure 6.16). This avoids needle entry into the pleura or lung laterally and the spinal canal medially. The procedure is performed using a “down-the-beam” approach, rotating the C-arm obliquely toward the side of needle entry, some 15–20° from the sagittal plane. This view shows the so-called “box” that forms the entry for the needle as it is directed toward the disc. The needle passes through the box, bound laterally by a line connecting the medial edge of rib heads and medially by a vertical line through the center of adjacent pedicles (Figure 6.17). Superior and inferior borders of the box are endplates of the adjacent vertebral bodies (Figure 6.18). When the C-arm is properly aligned, the skin entry point is directly above the “beam” until contact with the disc annulus is made (Figure 6.19). A longer 25- or 22-gauge needle can be used in larger patients. The needle must stay medial to the rib head and costo-vertebral joint in order to avoid the pleura. The needle must stay lateral to the lamina and interpedicular line to avoid entering the spinal canal. Lower and
Figure 6.16. Needle trajectory remains medial to the rib head and lateral to lamina to avoid pleura and spinal canal. (Reprinted with permission from Bogduk N. Thoracic provocation discography. In Bogduk N (ed): Practice Guidelines for Spinal Diagnostic and Treatment Procedures. San Francisco: ISIS, 2004. Provided by Dr. Claire Tibiletti.)
Thoracic Discography
Figure 6.17. The needle passes through a “box” bounded laterally by ribhead, medially by lamina, amd superiorly and infeiorly by endplates. (Reprinted with permission from Bogduk N. Thoracic provocation discography. In Bogduk N (ed): Practice Guidelines for Spinal Diagnostic and Treatment Procedures. San Francisco: ISIS, 2004. Provided by Dr. Claire Tibiletti.)
midthoracic discs can be easily and safely studied in most individuals. Upper (Figures 6.19, 6.20, and 6.21) thoracic discs may be difficult or impossible to enter. As the needle ascends in the thoracic spine, the route of access disappears, owing to the shorter disc height and close approximation of the ribs and costovertebral joints. Factors such as disc height, spinal deformity, and costovertebral or vertebral body osteophytes will affect the accessibility of individual thoracic discs. Once the needle contacts the annulus, a distinct change in resistance is felt. Lateral and AP fluoroscopic views should be utilized to guide the needle into the center of the disc (Figures 6.20 and 6.21). The volume of the thoracic discs changes from a lumbar-like 1 ml in the lower thoracic area to a cervical-like 0.4 ml in the upper thoracic region. It is important to use a low volume extension tubing and to fill the hub of the needle with a few drops of contrast prior to attaching the syringe and tubing to the needle. Manometry is not yet a standard of care in the thoracic region, and proceduralists rely on “thumb-pressure”
127
Figure 6.18. Note that the endplates are squared and the C-arm is rotated toward the side of entry until the “box” is approximately 25% across the disc space.
Figure 6.19. End plates must be parallel at the level of entry. The needles can be seen passing medial to the rib heads into the thoracic disc.
Thoracic Discography
Figure 6.20. Lateral view with needle in the middle of the disc space. The upper two discs appear normal. The lower two levels demonstrated posterior annular abnormalities and recreated familiar pain on injection of contrast.
Figure 6.21. AP view with contrast in the thoracic discs.
129
130
Chapter 6 Discography
to note the characteristics of the end-point pressure (firm, moderate, or low pressure). The volume injected should be recorded along with the appearance of the contrast injection and, most importantly, the pain response of the patient. The manner in which disc stimulation or provocation is carried out is similar to that described for the lumbar region. The patient must be asked whether pain is provoked on pressurization of the disc and if the pain is the same as the patient’s usual pain. A numerical rating scale should be used to rate pain intensity, and a clear description of the location of the provoked pain and its relation to the patient’s usual complaint must be recorded. AP and lateral views should be saved digitally or spot films taken to record both the needle placement and the contrast pattern following injection. A disc is said to be positive if provocation reproduces the patient’s usual pain at an intensity of >6/10 and if two adjacent discs are nonpainful. Postprocedure Care Postoperative care includes observation for at least 30 min. Shortness of breath may suggest the possibility of pneumothorax. Patients should not operate a motor vehicle on the day of their exam. Some soreness should be expected for a few days, and analgesic tablets may be provided in anticipation of this discomfort. The patient should be encouraged to report any unusual pain or other symptom that develops after the procedure. Discitis is a rare complication. As of publication, no case of discitis has been reported in a patient who received intradiscal antibiotics during discography. Discography response cannot be predicted in the thoracic spine based upon imaging studies.21 The authors have observed that thoracic discography has become an indispensable procedure in the investigation of pain that may have originated in the thoracic spine.
Cervical Discography Cervical discography is a diagnostic test utilized to determine whether a particular cervical disc is painful. During this procedure, contrast medium is injected into the nucleus of the disc, distending the disc and provoking pain. Cervical discography was pioneered by Cloward, who stimulated cervical discs intra-operatively in conscious patients, demonstrating that the disc is a source of pain.83–85 Magnetic resonance imaging cannot reliably identify the source of axial cervical pain. Studies have shown that although MR imaging correlates with pain response on lumbar discography, no such correlation is seen in the cervical spine.2,6,7,25–28 Discography can reveal cervical disc annular lesions that are simply not visible on even the highest resolution MR imaging studies. It is not uncommon to find a positive disc (abnormal appearance on injection of contrast with intense, concordant pain) in a patient with neck pain whose MR imaging is only mildly abnormal. If the discogram itself shows a morphologically normal disc, where contrast is confined to the nucleus, the test should not provoke pain.26 The incidence of pain in these normal-appearing
Cervical Discography
discs is uncommon, as most cervical discs show some degree of annular “abnormality,” or uncovertebral joint filling. These changes are age related and do not indicate symptomatic pathology. The presence of annular disruption has little relevance in the cervical spine. In a study by Schellhas and colleagues,27 many discs proved to have painless annular tears on discography. Discographically normal discs were not painful, whereas all intensely painful discs displayed tears of both the inner and outer annulus. Ohnmeiss and coworkers found that 97% of clinically painful discs showed these abnormalities.85 Anatomy Structurally, the cervical disc differs from the thoracic and lumbar discs. The cervical disc is thicker anteriorly than posteriorly. This wedge configuration produces the normal cervical lordosis that helps distribute compressive force.86 The cervical nucleus pulposis is located more posteriorly, in contrast to a more central location in the thoracic and lumbar discs (Figure 6.22).87 Anatomical studies have demonstrated the innervation of the cervical disc in both operative and cadaver specimens. The innervation is analogous to that of the lumbar disc. The sympathetic chain supplies the anterior annulus, the vertebral nerve supplies lateral annulus, and the sinuvertebral nerve innervates the posterior annulus.84,88–92 Procedure Cervical discography requires a high resolution C-arm with the capability of storing images.93 Review of previous cervical MR images is required prior to the procedure to detect any contraindication to discography. Discography should not be performed at any level where frank spinal cord compression exists with or without myelopathy. Any disc level
Figure 6.22. Note the thicker anterior annulus and posteriorly placed nucleus pulposis. (Reprinted with permission from Mercer S, Bogduk N. The Ligaments and Anulus Fibrosus of Human Adult Cervical Intervertebral Discs. Spine 1999;24:619–628.)
131
132
Chapter 6 Discography
demonstrating spinal cord deformity should be avoided or studied with extreme care, based upon individual circumstances.26,27 A history and a physical examination should be performed, and imaging should be reviewed. Informed consent should be obtained. The procedure should be explained to the patient, including the need for patient participation during the provocation portion of the exam. The patient is instructed to clearly describe any pain produced during disc stimulation and to compare this to his or her usual pain. Cervical discography is technically demanding and should not be carried out without significant previous experience in fluoroscopically guided injections. The procedure should be performed in a procedure room that is suitable for aseptic procedures. In addition to a C-arm, the room should contain a table that is X-ray compatible. Monitoring equipment should include ECG, blood pressure cuff, and pulse oximeter. Supplemental oxygen, suction, and standard resuscitation equipment should be available. The patient is positioned supine on the procedure table. Monitoring, sedation, skin preparation, and draping follow the guidelines suggested for lumbar discography. Meticulous sterile technique and a thorough, wide prep are important. The presence of a beard should be the cause for concern, particularly if C2–C3 is to be studied. The patient may need to be rescheduled and the beard trimmed. The only postdiscography discitis encountered by the authors occurred in an obese, diabetic patient with a full beard, some 15 years ago, prior to the use of intradiscal antibiotics. Now, the authors do not enter a disc for any reason without injecting antibiotic. No subsequent case of discitis has been identified in the authors’ practice in over 5,000 patients and 17,000 discs. Antibiotics can be given intravenously prior to the procedure and must be placed intradiscally during the procedure. The levels to be studied should be carefully identified by counting up from T1 and down from C2 in a PA view. The endplates at each level are squared, insuring that the X-ray beam is parallel to the disc space. This requires tilting the C-arm approximately 20° in a caudal direction. The view requires adjustment at each level studied due to patient positioning and degree of lordosis. Cervical discography is a right-sided procedure by convention due to the left-sided location of the esophagus (Figure 6.23). A left-sided approach is reserved for an unusual situation in which a hypertrophic scar, large spur, or hardware impedes a right-sided approach. The procedure is carried out utilizing a single 25-gauge, 3½ inch spinal needle at each level. There are two commonly used approaches (Figure 6.24): the anterior paratracheal approach and the oblique approach. Figure 6.24. The anterior approach passes between the carotid and trachea into the disc space. The oblique approach passes through the sternocleidomastoid andlongus colli muscles, posterior to internal jugular vein and just anterior to the uncinate. The location of carotid is variable and unless a palpating finger identifies this vessel, it may inadvertently become part of the trajectory. (Reprinted with permission from Slipman CW, Simeone FA, Mayer TG, Derby R. Interventional Spine: An Algorithmic Approach. Philadelphia: Elsevier Health Sciences, 2007. Copyright Elsevier.)
Cervical Discography
Figure 6.23. Cervical discography is typically done from the right side due to the left sided location of the esophagus.
133
134
Chapter 6 Discography
Anterior Paratracheal Approach The anterior paratracheal approach requires the use of two hands. The initial projection is PA, with the patient supine and the X-ray tube beneath the table (Figure 6.25). One hand is used to palpate the anterior cervical structures with the index and middle finger; palpation must be deep enough to feel the disc space in a slender patient. This hand is used to move vital structures out of the path of the needle. Once the fingers contact bone, they are spread to move the trachea and esophagus medially and the carotid artery and internal jugular vein laterally. This can be challenging in a large patient or a patient with an immobile larynx. This provides a direct path for the needle to enter the disc space and avoids the esophagus and great vessels. The initial target is the antero-lateral border of the endplate just below the target disc. This bony landmark is used to prevent overpenetration of the needle, which can travel through the disc and directly into the cervical spinal cord. Upon bony contact, the needle is held firmly and the C-arm rotated into a lateral projection. The needle tip should be against the anterolateral aspect of the superior endplate of the vertebral body below the target disc in both views. The needle is then walked superiorly off of the endplate into the disc annulus. There is a clear difference in feeling from hard bone to rubbery disc. Using PA and lateral fluoroscopic views, the needle can be advanced into the center of the disc space. A gentle curve in the tip of the needle can expedite steering to the
Figure 6.25. AP c-spine with caudal tilt to square vertebral endplates.
Cervical Discography
endpoint in disc center. Experienced discographers often enter the disc directly without bony contact, but always check a lateral view as soon as the firm annulus fibrosis is felt. Oblique Approach The oblique approach uses an oblique view. The C-arm is first positioned for a PA view of the cervical spine and then tilted caudally until the beam is parallel to the target disc space. The C-arm is then rotated toward the side of entry (generally the right side) until the neural foramen can be visualized. The needle is passed down the beam, tapping on the medial edge of the uncinate process just lateral and inferior to the disc and walking off medially and superiorly into the disc annulus (Figures 6.26 and 6.27). Once contact with annulus is felt, PA and lateral fluoroscopic views are utilized to advance the needle into the center (Figures 6.28–6.30). The oblique approach has the advantage of keeping the operators hand out of the fluoro beam. It is also often the only way to enter lower cervical discs in larger patients. The disadvantage of the oblique approach is that it puts the carotid artery at risk of puncture, as the course of the carotid is variable and may be directly between needle entry and target. Using a hand to palpate and divide the anterior cervical structures can mitigate this risk. Once needle placement in the center of the disc has been achieved, the stylette is removed and the needle hub filled with a few drops of contrast.
Figure 6.26. Oblique view with needle tapping the medial edge of the uncinate process.
135
136
Chapter 6 Discography
Figure 6.27. Oblique view with needle walked off uncinate process medially and superiorly into the center of the disc.
Figure 6.28. AP view with the needle placed in the center of the C6–7 disc space.
Cervical Discography
Figure 6.29. Lateral view with needle in the center of the disc space.
Figure 6.30. Lateral view with contrast demonstrating small posterior tear and bulge.
137
138
Chapter 6 Discography
The volume of a normal cervical disc can be as small as 0.1 ml, and even the dead space of the needle hub becomes important. A 3 cc syringe and low volume extension tubing filled with contrast (containing antibiotic) are attached to the needle. The patient is then reminded of how the test will be done. The disc is slowly pressurized by injection of contrast. Onset, severity, location, and concordance of any pain noted by the patient should be recorded. The volume and end-point feel of the disc (firm, medium, or spongy) should also be noted. It often takes significant pressure to reach dye-point (the pressure at which contrast is first seen entering the disc) in the cervical spine, particularly in younger, normal discs. The initial entry of contrast can be sudden and produce what we term an “opening-snap,” which can startle the patient and may be uncomfortable. Do not confuse this with a positive painful response. Slowly repressurize the disc and note the response on reinjection. The typical volume of contrast used for cervical discography is between 0.1 and 0.5 ml. Volumes over 0.5 ml are unusual, but occasionally an incompetent disc with a leak will require up to 1 ml in order to pressurize. Ohnmeiss and colleagues85 note an average cervical disc volume of 0.23 ml in their series, which is similar to the authors’ experience. Injection of contrast should be terminated if firm resistance is reached or if the patient complains of significant pain. One should stop the injection if a large leak is noted and be careful if the volume exceeds 0.5 cc. The disc can be repressurized to determine if the painful response is consistent. Another method to check for consistency is a “sham” injection. The patient is told that an injection is occurring while the syringe is simply held in the operator’s hand and asked if there is an increase in pain. The patient is expected not to respond to a sham injection and to respond consistently to reinjection. It is recommended26,27,85,94,95 that all cervical discs from C3–4 through C6–7 be studied if accessible. Pure imaging studies have been proven to be inaccurate in detecting painful annular lesions in the cervical spine. When headache of suspected cervical origin is a prominent clinical complaint, discography at C2–3 may be indicated.85 In the authors’ experience, postdiscography CT in the cervical region is useful and is expected by our surgical colleagues in patients with positive discography. Nevertheless, some do not feel that a CT scan contributes much information after the discogram.60 Postprocedure Care After completion of each discographic examination, patients are advised to expect some pain and discomfort, especially during the first 36 h and lasting up to a few days. Patients are given printed instructions regarding what to expect. They are warned that if they experience symptoms such as worsening pain, fever, chills, malaise, and night sweats within 1 week of the procedure, a disc infection could be developing. Patients may be given a prescription for a few pain tablets intended to last 3–4 days. Patients are kept at the authors’ facility for at least 30 min after the procedure. All postdiscography patients are called 2–5 days later to check on their status.
Importance of Strict Discography Criteria
Cervical Discography for Surgical Planning Cervical discography has diagnostic and therapeutic utility. By providing a diagnosis of intrinsic disc pain, it obviates the need for further diagnostic studies. It also provides the surgeon with information that can lead to improved outcomes.17 These improvements in outcome are a result of identification of symptomatic levels requiring surgical treatment as well as asymptomatic levels that do not. In addition, cervical discography often precludes surgery by identifying multiple symptomatic discs. This protects the patient from what may have been a surgical failure. It has been said that the best discogram is a negative discogram, which rules out a discogenic source of pain with great certainty. Similarly, a study that demonstrates multilevel internal disc disruption (3 or more positive levels) also generally rules out surgical intervention. Grubb and Kelly28 studied 173 patients with cervical pain. Of those patients, 75% had two positive discs and 54% had three or more positive on cervical discography. Only 20% of patients studied were felt to be appropriate surgical candidates following discography.28,30
Importance of Strict Discography Criteria Discography is a provocational test, and, as such, it is subject to certain inherent challenges. Many of the changes that have come about in the performance of this test over the past 15–20 years have been designed to provide more objective data and to minimize the subjective component of this exam. There are several components to a discogram study that can change the degree to which reliable results are obtained. First, as pain is being provoked by pressurizing the disc, the intensity of stimulation should be measured and carefully recorded. In the lumbar spine, this suggests that measuring pressures is very important and allows for consistency in applying criteria for a positive test. Secondly, the pressure criteria that one uses to determine whether any given disc is “positive” must be defined and strict. Pressure cut-off at <50 psi over opening pressure is good, but using a tighter criteria of <15 psi over opening pressure is even better. This means that if the pressure required to produce a significant painful response is more than 50 psi over opening pressure, the disc must be categorized as indeterminate. If pain is produced at less than 15 psi above the opening pressure, the data are much stronger. Thirdly, the intensity of pain provoked by the disc injection must be significant (>6/10) before a disc can be labeled as positive. Fourthly, volumes of contrast over 4.0 cc are probably not meaningful, and these discs should be considered indeterminate.65 Fifthly, there must be an adjacent control disc that is negative. Careful attention to injection pressure and response intensity can bring the false positive rates down to nearly zero.69 Walsh and colleagues11 studied the specificity of lumbar discography. In asymptomatic subjects undergoing discography, the false positive rate was 0%. In the cervical spine, Schellhas and coworkers26 studied asymptomatic volunteers and neck pain patients. Of the 40 discs studied in the asymptomatic group, there were no pain responses.
139
140
Chapter 6 Discography
The patients with chronic neck pain demonstrated concordant painful responses to 23 of 40 disc injections. Wood and colleagues21 also studied patients with chronic thoracic pain and a group of asymptomatic volunteers. Of the 49 discs studied in patients with thoracic pain, 27 (55%) were concordantly painful. Of the asymptomatic volunteers, 3 of 40 (7.5%) injections were painful. All three of these discs demonstrated prominent Schmorl’s nodes, and the pain produced was unfamiliar and nonconcordant. In all three of the above studies11,21,26 including the cervical, thoracic, and lumbar spine, discs that appeared normal on discography did not produce pain. Other authors have not been as supportive of discography. Carragee has published multiple papers that have posed many questions in regard to the validity of discography.93–99 He has made valid points that can improve the objective of provocation discography. Derby65 has meticulously reviewed the literature in regards to Carragee with the following key points: Carragee has reported high false positive rates in volunteer populations without existing low back pain: 10% of asymptomatic volunteers; 40% of patients with chronic cervical pain; >75% of patients with a somatization disorder; 16.7% of iliac bone graft harvest group. Carragee’s studies demonstrate that discography has false positives just like many other tests. The false positives are more common in patients with abnormal psychometric testing, specifically somatization disorder, and in those with previously operated discs. Bogduk has shown that if IASP criteria are followed strictly, using Carragee’s own data, the false positive results decrease significantly to an acceptable level (Table 6.2).17 ●●
●●
●●
Multiple investigators have demonstrated that provocative discography in asymptomatic discs can provoke pain, but it is usually mild and occurs at higher pressures.11,21,36,55 Patients with low pain tolerance and abnormal psychological profiles are not ideal candidates for discography or invasive surgical procedures. Carragee provoked remembered pain during discography that was similar or exactly like pain experienced during iliac crest bone harvesting (4 of 24 discs studied). There is no mention of the side of needle entry potentially provoking false positive rates during needle placement.
Table 6.2. Comparision of Carragee data vs. pressure controlled provactive discography data Carragee
Pain £ 50 psi
Pain £ 15 psi
Asymptomatic
10%
10%
0%
Chronic pain
40%
10%
0%
Abnormal psych
75%
50%
25%
Patient population
Conclusion ●●
●●
●●
In another study, Carragee suggests that discography caused chronic low back pain in patients without previous low back pain complaints. The subgroup that developed increased low back pain after discography had somatization disorder or chronic neck pain after failed cervical surgery. The sample size was small: six patients, with only four undergoing a complete discogram. Carragee performed discography on patients who had previously undergone discectomy. This included a group of patients with continued symptoms and a group that was asymptomatic. Forty percent of the asymptomatic patients and 63% of the symptomatic patients had a positive discogram at the previously operated level. False positive rates may be higher postdiscectomy, and strict adherence to diagnostic criteria must be followed when studying this population. Single-level discogram-positive patients were compared to patients with grade II spondylolisthesis and radiographic instability. Discectomy and fusion with pedicle screws were used in both populations. The spondylolisthesis patients had a high success rate, while only a third of the patients fused based on discography had significant pain relief. Carragee concluded that, since both groups underwent the same surgery, the poor results were due to the discogram providing an incorrect diagnosis in two-thirds of the patients. The surgeries were for two vastly different pathologies, with spondylolisthesis causing pain related to canal stenosis, instability, and foraminal stenosis.
Vertebrogenic Pain The vertebral body is a possible source of pain elicited by discography. This potential source of pain might explain why some patients fail surgical interventions related to discogenic pain. Vertebral endplates can deflect in response to disc pressure produced during pressurization.100 The endplate is highly innervated;57 the basivertebral nerve (branch of sinuvertebral nerve that enters the vertebral body via the posterior neurovascular foramen) has been shown to be the major source of endplate innervation.101 Disruption of the basivertebral nerve is thought to be a mechanism of pain relief obtained with vertebroplasty.102 The ability to accurately and safely ablate the basivertebral nerve in mature sheep using fluoroscopically guided percutaneous radiofrequency ablation via the pedicles and into the postero-central vertebral body has been established.103 This technique may be useful in treating discogram-positive degenerative discs in humans and needs further research.
Conclusion Discography has become an indispensable assessment tool in the evaluation of spinal pain.8–34 Current imaging techniques are insufficient to determine if a suspected disc is the source of pain.2,25–28 Discography is the only test of the intervertebral disc that is able to
141
142
Chapter 6 Discography
stimulate the disc and recreate a patient’s pain. This is analogous to palpation, which is fundamental to physical diagnosis.61 Disc stimulation and pain reproduction can be correlated with analgesia and function if local anesthetic is introduced into the disc.71,72,74 There is both historical and current controversy surrounding the use of discography.11,36,63,65,93–99 This has promoted healthy discussion and advanced the standards for performance of this examination. The use of manometry, Functional Analgesia (FAD), “sham” injection, and strict criteria for identifying positive discs are all intended to limit the likelihood of a false positive result. It is acknowledged that discography has to be interpreted with caution in patients with significant behavioral pathology. With the continuous evolution of spinal interventions come many new and promising treatments for painful intervertebral discs. This includes a host of emerging biological therapies and surgical treatments. All of these require the ability to safely access the intervertebral disc for both diagnosis and treatment. The demand for this procedure is certain to increase. When discography is performed with appropriate clinical indications by skilled, knowledgeable, and experienced proceduralists, it leads to improved clinical outcomes. References 1. Lindblom K. Technique and results in myelography and disc puncture. Acta Radiol 1950;34:321–330. 2. Aprill C, Bogduk N. High-intensity zone: a diagnostic sign of painful lumbar disc on magnetic resonance imaging. Br J Radiol 1992;65:361–369. 3. Guyer RD, Ohnmeiss DD. Lumbar discography. Position statement from the North American Spine Society Diagnostic and Therapeutic Committee. Spine 1995;20:2048–2059. 4. Lindblom K. Technique and results of diagnostic disc puncture and injection (discography) in the lumbar region. Acta Orthop Scand 1951;20:315–326. 5. Bogduk NC, Aprill C, Derby R. Discography. In White A, Schofferman A (eds): Spine Care. Diagnosis and Conservative Treatment. St. Louis: Mosby, 1995:219–236. 6. Lam KS, Carlin D, Mulholland RC. Lumbar disc high-intensity zone: the value and significance of provocative discography in the determination of the discogenic pain source. Eur Spine J 2000;9:36–41. 7. Schellhas KP, Pollei SR, Gundry CR, Heithoff KB. Lumbar disc high-intensity zone. Correlation of magnetic resonance imaging and discography. Spine 1996;21:79–86. 8. Buirski G, Silberstein M. The symptomatic lumbar disc in patients with lowback pain: magnetic resonance imaging appearance in both symptomatic and control population. Spine 1993;18:1808. 9. Milette PC, Fontaine S, Lepanto L, Cardinal E, Breton G, Haughton V. Differentiating lumbar disc protrusions, disc bulges and discs with normal contour but abnormal signal intensity. Magnetic resonance imaging with discographic correlations. Spine 1999;24:44. 10. Ohnmeiss DD, Vanharanta H, Ekholm J. Relation between pain locations and disc pathology: a study of pain drawings and CT/discography. Clin J Pain 1999;15:210. 11. Walsh TR, Weinstein JN, Spratt KF, Lehmann TR, Aprill C, Sayre H. Lumbar discography in normal subjects: a controlled, prospective study. J Bone Joint Surg 1990;72:1081.
Conclusion 12. Manchikanti L, Singh V, Pampati V, Fellows B, Beyer C, Damron K, Cash KA. Provocative discography in low back pain patients with or without somatization disorder: a randomized prospective evaluation. Pain Physician 2001;4:227–239. 13. Horton WC, Daftari TK. Which disc visualized by magnetic resonance imaging is actually a source of pain? A correlation between magnetic resonance imaging and discography. Spine 1992;17:S164–S171. 14. Kornberg M. Discography and magnetic resonance imaging in the diagnosis of lumbar disc disruption. Spine 1989;14:1368–1372. 15. Manchikanti L, Singh V, Pampati V, Damron KS, Barnhill RC, Beyer C, Cash KA. Evaluation of the relative contributions of various structures in chronic low back pain. Pain Physician 2001;4:308–316. 16. Buenaventura RM, Shah RV, Patel V, Benyamin R, Singh V. Systematic review of discography as a diagnostic test for spinal pain: an update. Pain Physician 2007;10:147–164. 17. Bogduk N. Lumbar disc stimulation (provocation discography). In: Practice Guidelines for Spinal Diagnostic and Treatment Procedures. ISIS, 2004:20–46. 18. Derby R, Eek B, Lee SH, Seo KS, Kin B-J. Comparison of intradiscal restorative injections and intradiscal electrothermal treatment (IDET) in the treatment of low back pain. Pain Physician 2004;7:63–66. 19. Holt EP Jr. Fallacy of cervical discography. Report of 50 cases in normal subjects. JAMA 1964;188:799–801. 20. Schellhas KP, Pollei SR, Dorwart RH. Thoracic discography: a safe and reliable technique. Spine 1994;19:2103. 21. Wood KB, Schellhas KP, Garvey TA, Aeppli D. Thoracic discography in healthy individuals: a controlled prospective study of magnetic resonance imaging and discography in asymptomatic symptomatic individuals. Spine 1999;24:1548. 22. Winter RB, Schellhas KP, Painful adult thoracic Scheuermann’s disease: diagnosis by discography and treatment by combined arthrodesis. Am J Orthop 1996;25:783. 23. Bogduk N. Thoracic provocation discography. In: Practice Guidelines for Spinal Diagnostic and Treatment Procedures. ISIS, 2004:287–294. 24. Singh V. Thoracic discography. Pain Physician 2004;7:451–458. 25. Parfenchuck TA, Janssen ME. A correlation of cervical magnetic resonance imaging and discography/computed tomographic discograms. Spine 1994;19:2819. 26. Schellhas KP, Smith MD, Gundry CR, Pollei SR. Cervical discogenic pain: prospective correlation of magnetic resonance imaging and discography in asymptomatic subjects and pain sufferers. Spine 1996;21:300. 27. Schellhas KP, Garvey TA, Johnson BA, Rothbart PJ, Pollei SR. Cervical discography: analysis of provoked responses at C2–C3, C3–C4, and C4–C5. Am J Neuroradiol 2000;21:269. 28. Grubb SA, Kelly C. Cervical discography: clinical implications from 12 years of experience. Spine 2000;25:1382. 29. Fortin JD. Cervical discography with CT and MRI correlations. In Lennard TA (ed): Pain Procedures in Clinical Practice. Philadelphia: Hanley and Belfus, 2000:230–240. 30. Bogduk N. Cervical disc stimulation (provocation discography). In Practice Guidelines for Spinal Diagnostic and Treatment Procedures. ISIS, 2004:95–111. 31. Roth DA. Cervical analgesic discography. A new test for the definitive diagnosis of painful-disk syndrome. JAMA 1976;235:1713–1714. 32. Whitecloud TS III, Seago RA. Cervical discogenic syndrome. Results of operative intervention in patients with positive discography. Spine 1987;12:313–317. 33. Motimaya A, Arici M, George D, Ramsby G. Diagnostic value of cervical discography in the management of cervical discogenic pain. Conn Med 2000;64:395–398.
143
144
Chapter 6 Discography 34. Zeidman SM, Thompson K, Ducker TB. Complications of cervical discography: analysis of 4400 diagnostic disc injections. Neurosurgery 1995;37:414–417. 35. Kozak J, O’Brien J. Simultaneous combined anterior and posterior fusion. Spine 1990;15:322–328. 36. Derby R, Howard MW, Grant JM, Lettice JJ, Van Peteghem PK, Ryan DP. The ability of pressure-controlled discography to predict surgical and non-surgical outcomes. Spine 1999;24:364–371. 37. Blumenthal SL, Baker J, Dossett A, Selby DK. The role of anterior fusion for internal disc disruption. Spine 1988;13:566–569. 38. Newman MH, Grinstead GL. Anterior lumbar interbody fusion for internal disc disruption. Spine 1992;17:831–833. 39. Colhoun E, McCall IW, Williams L, Cassar Pullicino VN. Provocation discography as a guide to planning operations on the spine. J Bone Joint Surg 1988;70B(2):267–271. 40. Cohen S, Hurley R. The ability of diagnostic spinal injections to predict surgical outcomes. Anesth Analg 2007;105:1756–1775. 41. Saifuddin A, Braithwaite I, White J, Taylor BA, Renton P. The value of lumbar spine magnetic resonance imaging in the demonstration of annular tears. Spine 1998;23:453–457. 42. Ito M, Incorvaia K, Yu S, Fredrickson BE, Yuan HA, Rosenbaum AE, Bogduk N. Predictive signs of discogenic lumbar pain on magnetic resonance imaging with discography correlation. Spine 1998;23:1252–1258. 43. Guyer RD, Ohnmeiss DD, NASS. Lumbar discography. Spine J 2003;3: 11S–27S. 44. Grogan JP, Hemminghytt S, Williams AL, et al. Another strategy recommended by Balderston et al. J Spinal Disord Tech 2004;17:248–250. 45. Arrington JA, Murtagh FR, Silbiger ML, Rechtine GR, Nokes SR. Magnetic resonance imaging of postdiscogram discitis and osteomyelitis in the lumbar spine: case report. J Fla Med Assoc 1986;73:192–194. 46. Guyer RD, Collier R, Stith WJ, Ohmneiss DD, Hochschuler SH, Rashbaum RF, Regan JJ. Post-discography discitis. Spine 1988;13:1352–1354. 47. NASS position statement on discography. Spine 1995;20:2048–2059. 48. Klessig HT, Showsh SA, Sekorski A. The use of intradiscal antibiotics for discography: an in vitro study of gentamicin, cefazolin, and clindamycin. Spine 2003;28:1735–1738. 49. Boswell M, Wolfe J. Intrathecal cefazolin induced seizures following attempted discography. Pain Physician 2004;7:103. 50. Mathis J. Procedural techniques and materials: tumors and osteoporotic fractures. In Mathis J (ed): Percutaneous Vertebroplasty. New York: Springer, 2002. 51. Cavanaugh JM, Kallakuri S, Ozaktay AC. Innervation of the rabbit lumbar intervertebral disc and posterior longitudinal ligament. Spine 1995;20:2080–2085. 52. Raj P. Intervertebral disc. Pain Pract 2008;8:18–44. 53. Heggeness MH, Walters WC III, Gray PM Jr. Discography of lumbar discs after surgical treatment for disc herniation. Spine 1997;22:1606–1609. 54. Ahmed M, Bjurholm A, Kreicbergs A, Schultzberg M. Neuropeptide Y, tyrosine hydroxylase and vasoactive intestinal polypeptide-immunoreactive nerve fibers in the vertebral bodies, discs, dura mater, and spinal ligaments of the rat lumbar spine. Spine 1993;18:268–273. 55. Palmgren T, Gronblad M, Virri J, Kääpä E, Karaharju E. An immunohistochemical study of nerve structure. Spine 1999;24:2075–2079. 56. Peng B, Wu W, Hou S, Li P, Zhang C, Yang Y. The pathogenesis of discogenic low back pain. J Bone Joint Surg 2005;87:62–67. 57. Fagan A, Moore R, Roberts V, Blumbergs P, Fraser R. The innervation of the intervertebral disc: a quantitative analysis. Spine 2003;28(23):2570–2576.
Conclusion 58. Pauza K. Nomenclature and Terminology for Spine Specialists (Appropriate Words Meant to Replace the Most Commonly Misused Words of the Spine Specialists). PASSOR Educational Guidelines Task Force, 2005. 59. Mathis J. Percutaneous vertebroplasty: complication avoidance and technique optimization. Am J Neuroradiol 2003;24:1697. 60. Schellhas KP. Discography. Spinal Interv (Neuroimaging Clin N Am) 2000;10:579–596. 61. Derby R. Discography. In Slipman E, Derby R, Simeone FA, Mayer TG (eds): Interventional Spine: An algorithmic approach. Philadelphia: WB Saunders, 2008:291–302. 62. Khot A, Bowditch M, Powell J, Sharp D. The use of intradiscal steroid therapy for lumbar spinal discogenic pain: a randomized controlled trial. Spine 2004;29:833–836. 63. Derby R, Lee SH, Kim BJ, et al. Pressure-controlled lumbar discography in volunteers with low back pain symptoms. Pain Med 2005;6:213–221. 64. Zhou Y, Abdi S. Diagnosis and minimally invasive treatment of lumbar discogenic pain: a review of the literature. Clin J Pain 2006;22:468–481. 65. Wolfer LE, Derby R, Lee J-E, Lee S-H. Systematic review of lumbar provocation discograpy in asymptomatic subjects with a meta-anaylysis of false-positive rates. Systematic review. Pain Physician 2008;11(4):513–538. 66. Bull T, Sharp D, Powell J. The efficacy of intradiscal steroid injection compared to modic changes in degenerative lumbar discs. J Bone Joint Surg Br 1998;80:40. 67. Bogduk N, Modic MT. Lumbar discography. Spine 1996;21:402–404. 68. Vanharanta H, Sachs BL, Spivey MA, et al. The relationship of pain provocation to lumbar disc deterioration as seen by CT/discography. Spine 1987;12:295–298. 69. O’Neill C, Kurgansky M. Subgroups of positive discs on discography. Spine 2004;29:2134. 70. Carrino JA, Swathwood TC, Morrison WB, Glover JM. Prospective evaluation of contrast-enhanced MR imaging after uncomplicated lumbar discography. Skeletal Radiol 2007;36:293–299. 71. Discyphor Procedural Technique Guide. Sunnyvale, California: Kyphon Inc, 2006. 72. Functional Anesthetic Discography [Package Insert]. Sunnyvale, California: Kyphon Inc, 2006. 73. Alamin T, Arawal V, Carragee E. FAD versus provocative discography: comparative results and postoperative clinical outcomes. Spine J 2007;7:39S. 74. Luchs J, DeMoura A, Cho M, Ortiz A. Preliminary experience with functional anesthetic discography. Interv Spine 2007;7(5):63S. 75. Wetzel FT, LaRocca SH, Lowery GL, Aprill CN. The treatment of lumbar spinal pain syndromes diagnosed by discography. Spine 1994;19(7):792–800. 76. Albrand OW, Corkill G. Thoracic disc herniation: treatment and prognosis. Spine 1979;4:41–46. 77. Epstein JA. The syndrome of herniation of the lower thoracic intervertebral discs with nerve root and spinal cord compression. J Neurosurg 1973;11:525–538. 78. Vanichkachorn JS, Vaccaro AR. Thoracic disc disease: diagnosis and treatment. J Am Acad Orthop Surg 2000;8:159–169. 79. Bogduk N. Intervention and pain patterns of the thoracic spine. In Grant R (ed): Physical Therapy of the Cervical and Thoracic Spine, 3rd ed. New York: Churchill Livingstone, 2002:73–81. 80. Wyke B. The neurological basis of thoracic spinal pain. Rheumatol Phys Med 1970;10(7):356–367. 81. Slezak J, Stojanovic M. Injection procedures. In Slipman E, Derby R, Simeone FA, Mayer TG (eds): Discography in Interventional Spine: An Algorithmic Approach. Philadelphia: WB Saunders, 2008:787–792.
145
146
Chapter 6 Discography 82. Kapural L, Goyle A. Imaging for provocative discography and minimally invasive percutaneous procedures for treatment of discogenic lower back pain. Tech Reg Anesth Pain Manag 2007;11:73–80. 83. Cloward RB. Cervical discography: technique, indications and use in diagnosis of rupture cervical disks. AJR Am J Roentgenol 1958;79:563–574. 84. Cloward RB. The clinical significance of the sinu-vertebral nerve of the cervical spine in relation to the cervical disk syndrome. J Neurol Neurosurg Psychiatry 1960;23:321–326. 85. Ohnmeiss DD, Guyer RD, Mason SL. The relation between cervical discographic pain responses and radiographic images. Clin J Pain 2000;16:1–5. 86. An H, Simpson M. Surgery of the Cervical Spine. Baltimore: Williams and Wilkins, 1994:1–39. 87. Sylven B. On the biology of nucleus pulposis. Acta Orthop Scand 1951;20:275–279. 88. Bogduk N, Windsor M, Inglis A. The innervation of the cervical intervertebral discs. Spine 1989;13:2–8. 89. Groen GJ, Baljet B, Drukker J. Nerves and nerve plexuses of the human vertebral column. Am J Anat 1990;188:286–296. 90. Mendel T, Wink CS, Zimny ML. Neural elements in human cervical intervertebral discs. Spine 1992;17:132–135. 91. Ferlic DC. The nerve supply of the cervical intervertebral discs in man. Bull Johns Hopkins Hosp 1963;113:347–351. 92. Chabot MC, Montgomery DM. The pathophysiology of axial and radicular neck pain. Semin Spine Surg 1995;7:2–8. 93. Carragee EJ, Chen Y, Tanner CM, Hayward C, Rossi M, Hagle C. Can discography cause long-term back symptoms in previously asymptomatic subjects? Spine 2000;25:1803–1808. 94. Carragee EJ, Chen Y, Tanner CM, Truong T, Lau E, Brito JL. Provocative discography in patients after limiting lumbar discectomy: a controlled, randomized study of pain response in symptomatic and asymptomatic subjects. Spine 2000;25:3065–3071. 95. Carragee EJ, Tanner CM, Yang B, Brito JL, Truong T. False-positive findings on lumbar discography: reliability of subjective concordance assessment during provocative disc injection. Spine 1999;24:2542–2547. 96. Carragee EJ, Alamin TF, Miller J, Alamin TF, Miller J, Grafe M. Provocative discography in volunteer subjects with mild persistent low back pain. Spine 2002;2:25–34. 97. Carragee EJ, Tanner CM, Khurana S, Hayward C, Welsh J, Date E, Truong T, Rossi M, Hagle C. The rates of false-positive lumbar discography in select patients without low back symptoms. Spine 2000;25:1373–1380. 98. Carragee EJ. Psychological and functional profiles in select subjects with low back pain. Spine J 2001;1:198–204. 99. Carragee EJ, Barcohana B, Alamin T, van den Haak E. Prospective controlled study of the development of lower back pain in previously asymptomatic subjects undergoing experimental discography. Spine 2004;29:1112–1117. 100. Heggeness MH, Doherty BJ. Discography causes end plate deflection. Spine 1993;18:1050–1053. 101. Antonacci MD, Mody DR, Heggeness MH. Innervation of the human vertebral body: a histologic study. J Spinal Disord 1998;11:526–351. 102. Niv D, Gofeld M, Dever M. Causes of pain in degenerative bone and joint disease: a lesson from vertebroplasty. Pain 2003;105:387–392. 103. Hoopes J, Eskey J, Attawia M, Patel S, Ryan P, Pellegrino R, Bergeron J. Radiofrequency ablation of the basivertebral nerve as potential treatment of back pain: pathologic assessment in an ovine model. Proc Soc Photo Opt Instrum Eng 2005;5698:168–180.
7 Percutaneous Lumbar Discectomy Stanley Golovac
Introduction Discogenic leg pain is a primary cause of healthcare expenditure. The two entities of back pain and discogenic leg pain produce more days lost than any other combined illness and injuries, costing the US healthcare system over $20 billion per year.1,2 Pain from discogenic sources typically is produced from annular breakdown and annular tears.3,4 This is commonly treated with a microdiscectomy by orthopedic surgeons and neurosurgeons. Open discectomy has been the “gold standard” for relieving pressure on nerve roots. By decompressing the nerve root from the disc itself, neurologic function is usually restored, and pain is relieved. Because of the annular violation that happens from the surgical procedure, recurrent disc herniations may occur and typically do.5 Because of that, a number of percutaneous procedures have been developed over the last several decades specifically focused on this disc pathology. This includes chemonucleolysis, automated/manual percutaneous nucleotomy, laser treatments, intradiscal thermal annuloplasty, and, more recently, nucleoplasty and dekompressor. All of these are designed to reduce intradiscal pressure and can allow the protruded disc area to retract back into place as long as there is enough elastogenicity to allow recovery.
Anatomy The intervertebral disc is anatomically composed of a central cartilaginous fibrous ring with sensitive nerve endings surrounding the outer rim.6 The nucleus is bordered superiorly and inferiorly by dense cartilage and bony endplates from the vertebral surfaces, respectively (Figure 7.1). The annulus itself is composed of an inner layer and an outer layer. The annulus is loosely attached to the anterior longitudinal ligament and strongly attached to the posterior longitudinal ligament. This is
From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_7, © Springer Science + Business Media, LLC 2010
147
148
Chapter 7 Percutaneous Lumbar Discectomy
Figure 7.1. Sagittal diagram showing the anatomy of the lumbar intervertebral disc. The soft inner nucleus pulposus is encircled by the fibrous bands of the annulus, which are thinner posteriorly. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004.)
one reason why, when a “central” disc protrusion occurs, pain is experienced as axial back pain. Because of the ligamentous displacement from the annulus and the vertebral body, pain is elicited from the fibers pulling away from their attachments. The nucleus is a notochordal remnant that is relatively avascular in the adult and is not significantly innervated. The nucleus acts as a cushion. The annulus is a highly innervated structure in which the sinuvertebral nerve wraps around the posterior aspect. Substance P and unmyelinated C fibers have been demonstrated in the annulus (Figure 7.2). The intervertebral disc functions to provide a surface area to distribute stress absorption and motion restriction.2 The annulus serves to contain the nuclear material and to restrict longitudinal and rotational motion between spinal segments. Fibers in the annulus are arranged in variable directions in each fibrous layer – approximately 20 anteriorly and approximately 12–15 posteriorly – providing support in multiple directions.
Figure 7.2. Axial cross-section showing the innervation of the intervertebral disc. There is no significant innervation of the nucleus, while the annulus is innervated with unmyelinated fibers, primarily by way of the sinovertebral nerve. Pain fibers are present throughout the disc but most densely in the posterior annulus. (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004.)
Historical Perspective
Intervertebral Discs in Spinal Pain While not completely understood to its fullest degree, discogenic pain originates from the sinuvertebral nerves that innervate the outer annulus of the discs. Whether it is pressure or chemical sensitivity, discogenic pain is provoked by any factor that produces an inflammatory response from disc chemical mediators. Phosphlipase A and prostaglandins are known as irritants to the outer nerve supply, thereby producing pain and inflammation. Discogenic pain is typically seen when axial loading causes pain during standing and extension of the spine. The pain is described as stabbing, knifelike, or burning in nature. Theories for the exact pathophysiology of the pain mechanism abound, but most revolve around pathological tears of the posterior annulus of the disc and mechanical or chemical stimulation of nociceptive receptors located in and around the posterior annulus fibrous and relayed through the sinuvertebral nerve. Present therapy has not been questioned or changed in many years. Rest, relaxation, antiinflammatory agents, and ice to the affected areas are commonly used. Then, limited exercise, stretching, and aquatic body posturing are introduced. If the pain increases after conservative measures have failed, then interventional therapies such as trigger injections, nerve conduction blockade, and central nervous system blocks (i.e., epidural steroids, radiofrequency ablation, and sympathetic blockade) may have a place in the treatment algorithm. Surgical intervention should be considered when a neurologic deficit exits, causing loss of reflexes, sensory deficits, and weakness. Surgical intervention is needed to restore normal nerve function. This may mean a minimally invasive microdiscectomy, laminectomy, foraminotomy, or an extensive fusion. The unfortunate occurrence of morbidity always is a factor that needs to be considered. Additional considerations include cost, loss of work time, rehabilitation, the possibility of developing chronic pain, and medication dependency.
Historical Perspective Percutaneous procedures were developed in the 1990s as a minimally invasive treatment to reduce morbidity and relieve discogenic and radicular pain. Percutaneous needle insertion was initiated by Gary Onik, an interventional radiologist who believed that by introducing a trocar into the nucleus one could extract disc material and reduce outer bulges that compressed or irritated nerve roots.7 In 2000, the Saal brothers invented the Intradiscal Electrothermy procedure (IDET) by placing a heating element into the posterior nuclear annulus interface; the theory of regelatinizing the proteoglycan fibers would allow the annular tears to seal off. Several studies have shown an improvement of 40%.8 In 2000, ArthroCare® Spine introduced the Nucleoplasty® (ArthoCare®, Austin, TX) wand that would allow a practitioner to create channels in the disc, and thereby create voided spaces where retraction of the disc bulge could occur. It is still not proven whether the radiofrequency energy or the physical characteristic change in the
149
150
Chapter 7 Percutaneous Lumbar Discectomy
Figure 7.3. Dekompressor® device (Stryker, Kalamazoo, MI). (Used with permission of Stryker, Kalamazoo, MI.)
disc reduces the overall pain. In 2002, Stryker® Corporation introduced the Dekompressor® (Stryker®, Kalamazoo, MI) device (Figure 7.3) to percutaneously remove disc material. This yields a specimen for quantitative and qualitative analysis and allows for proof of what was removed and how much.9,10
Indications and Procedure Needle Placement Inclusion and exclusion criteria are very important for the desired outcome. MRI findings need to be verified first with an AP and then with a LAT view (Figure 7.4a, b). Once a selective nerve root block has been performed, pain relief should be confirmed prior to proceeding with a percutaneous discectomy. The patient should be in a prone position when one begins the procedure. Minimal sedation such as Versed® (Hoffman-LaRoche, Inc, Nutley, NJ) and fentanyl can be titrated to allay anxiety and to help control blood pressure related changes. A betadine and alcohol prep are crucial to help cleanse and sterilize the targeted area. The fluoroscope is positioned to view the spine in an AP view. The endplates and pedicles are aligned. Next, one should oblique the scope approximately 30° or until the SAP (Superior Articulating Process) bisects the endplates. Once the disc is clearly seen lateral to the SAP, the objective target is in sight. The skin and deeper fascial plane should be anesthetized by using a 3.5-in. spinal needle to further place local anesthetic solution toward the SAP. One then should introduce the 17-gauge Crawford needle into the outer annulus and puncture the disc. The needle should enter the disc and stay in the posterior aspect of the disc itself. The first view should be an AP view. This will allow one to visualize the entrance point of the needle in order to determine whether it has crossed the midline mark. The best way to know whether the needle is in an optimal position is to visualize the pedicle and spinous process. Drawing a line between each will allow for a clear reference that the needle tip is in the middle of the disc (Figure 7.5). This is visualized in a lateral view that illustrates the needle tip in the posterior one-third of the vertebral body. It will indicate that the distal tip or auger will be located in the nuclear–annular junction. On an AP view, the needle tip should not pass beyond the spinous process. See Figure 7.6a, b.
Indications and Procedure Needle Placement
Figure 7.4. (A) Contained bulged disc at lumbar vertebra L 4. (B) Axial view of right paravertebral disc bulge.
Next, one turns on the Dekompressor® (Stryker®, Kalamazoo, MI) and slightly pushes the unit in and out approximately 1 cm (Figure 7.7). After running the device for 2 min, the probe should be removed, and the distal portion of the probe should be cut off and sent to pathology for quantitative and qualitative analysis.
151
152
Chapter 7 Percutaneous Lumbar Discectomy
Figure 7.5. P–P proper needle placement midway between the spinousprocess and pedicle denoting correct placement of the introducerneedle.
The lower back area is then cleaned with alcohol. Neosporin® (Johnson & Johnson, New Brunswick, NJ) ointment can be applied over the puncture site. A bandage is used to cover the needle site. The patient is then transferred to a prep and discharge area. After an uneventful recovery with no neurological symptoms and neurologic insults, the patient can be discharged with instructions to not bend, twist, flex, or squat for the next 24 h. Physical therapy is an integral part of rehabilitation, consisting of aqua exercises and dynamic progressive strengthening. After 4 weeks of continued rehabilitation, patients are required to be re-evaluated, and a determination of total improvement can be made. Should pain and radicular symptoms not improve to the physician’s satisfaction, a surgical consult can be requested to determine surgical candidacy.
Complications and Consent Each and every procedure may produce a set of complications that needs to be addressed prior to the procedure. Infection, bleeding, nerve root injury, disc infection (discitis), blood vessel injury, and the remote chance of death: all of these items should be covered in the risks, alternatives, and benefits section of the consent forms.
Figure 7.6. (A, B) Lateral view of needle placement showing the introducer in the posterior 1/3 of the vertebral body.
154
Chapter 7 Percutaneous Lumbar Discectomy
Figure 7.7. Dekompressor® (Stryker, Kalamazoo, MI) device in oblique view noted to enter the affected side. (Used with permission of Styker, Kalamazoo, MI.)
Conclusion When strict criteria are followed, the expected outcome is measured at approximately 75%. This includes postprocedural rehabilitation of 2 weeks of aqua therapy and then 2 weeks of progressive functional exercises. At the completion of 4 weeks of rehabilitation, a reevaluation is performed, and definitive outcome is then measured. Should a persistent mechanical pain exist, then either a discogram or selective nerve root sleeve should be performed. References 1. Lipetz JS. Pathophysiology of inflammatory, degenerative and cooperative radiculopathies. Phys Med Rehabil Clin N Am 2002;13:439–449. 2. Carey TS, Garrett J, Jackman A, Mc Laughlin C, Fryer J, Smucker D. The outcomes and costs of care for acute low back pain among patients seen by primary care practitioners, chiropractors, and orthopedic surgeons. The North Carolina Back Pain Project. N Engl J Med 1995;333:913–917. 3. Prescher A. Anatomy and pathophysiology of the aging spine. Eur J Radiol 1998;27:181–195. 4. Coppers MH, Marani E, Thomeer RT, Groen GJ. Innervation of painful lumbar discs. Spine 1997;22:2342–2350. 5. Carragee EJ, Hahn M, Suen P, Kim D. Clinical outcomes after lumbar discectomy for sciatica: the effects of fragment type annular competence. J Bone Joint Surg 2003;85-A:102–108. 6. Haines SJ, Jordan N, Boen JR, Nyman JA, Oldridge NB, Lindgren BR. Discectomy strategies for lumbar disc herniations: results of the LAPDOG trial. J Clin Neurosci 2002;9(4):411–417. 7. Onik GM, Helms CA, Ginsberg L, et al. Percutaneous lumbar discectomy using a new aspiration probe. AJNR Am J Neuroradiol 1985;6:290–293.
Conclusion 8. Saal JS, Saal JA. Management of chronic discogenic low back pain with a thermal intradiscal catheter. A preliminary report. Spine 2000;25(3):382–388. 9. Alo KM, Wright RE, Sutcliffe J, Brandt SA. Perc lumbar discectomy. Clin response in an Initial cohort of fifty consecutive patients with chronic radicular pain. Pain Pract 2004;4(1):19–29. 10. Amoretti N, David P, Gimaud A, Flory P, Hovorka I, Roux C, Chevallier P, Bruneton JN. Clinical follow-up of 50 patients treated by percutaneous lumbar discectomy. Clin Imaging 2006;30:242–244.155 Chapter 8 Epidural Injections for the Treatment
155
8 Epidural Injections for the Treatment of Spine-Related Pain Syndromes Matthew T. Ranson and Timothy R. Deer
Introduction The use of epidural injections has become a common procedure for the treatment of spinal diseases causing pain. In some settings, such as that of acute lumbar radiculitis, the uses of these techniques have substantial benefits. In other clinical scenarios, such as chronic axial spinal pain, the utility of these procedures is debatable. Epidural steroids are an option in the broad algorithm of pain treatment, but they are a more conservative option than many more permanent techniques.1–3 This chapter will review the proper use of epidural injections with a focus on patient selection, technical aspects of the procedure, risk and complication management, and outcomes.
Historical Perspectives Epidural injections were first reported in the literature more than a century ago when Sicard injected cocaine to treat acute pain. In this case, he placed the drug, a potent local anesthetic, by the caudal approach in an attempt to treat sciatica.4 Shortly thereafter, cocaine was injected by Cathelin to create surgical anesthesia.5 The first reports of epidural steroids being used to treat spinal-related pain disorders surfaced in a 1952 article by Robecchi and Capra.6 Like many of today’s pain publications, these authors presented anecdotal information of unknown clinical significance. Subsequently, there have been many studies performed to evaluate the efficacy of epidural steroids in the management of low back pain and radiculopathy.7–9 Some studies have reported little or no benefit of epidural steroid injections compared with surgery,9 but these studies have been flawed by the lack of controls, absence of patient and physician blinding, and absence of randomization. Unfortunately, many of the studies in the current literature involve injection of patients by a variety of practitioners with
From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_8, © Springer Science + Business Media, LLC 2010
157
158
Chapter 8 Epidural Injections for the Treatment
no standardized training. In many cases, the “epidural” injections were performed without fluoroscopic guidance, and the true location of the needle is unknown, making the studies of questionable significance.
Demographics Back pain is estimated to affect an estimated 70–80% of adults at some point during their lifetimes.10,11 Between 2 and 5% of adults seek medical treatment annually for back pain.12 Approximately 20% of adults are estimated to experience neck pain over a 1-year period and more than 60% at least once during their lifetimes.13,14 The total cost of back pain exceeds 100 billion dollars a year in the US alone.15 These costs may be markedly underestimated based on the cost of treating addiction, depression, and other co-morbidities associated with spinalrelated pain syndromes. Additionally, as managed care attempts to contain the costs associated with the treatment of back pain, it is clear that access to specialists and treatment modalities will be reduced. This may increase costs in the long term, secondary to an increased number of patients with prolonged chronic pain. In addition to these confounding variables, the publication of “guidelines” by groups with poor specialty representation and questionable recommendations may limit appropriate patient access. A significant percentage of spine-related pain patients will eventually be treated with spine surgery, which can create a new class of pain disorders often referred to as the “failed back surgery syndrome” or the “post laminectomy pain syndrome.” The numbers for spine surgery in the United States far exceed those of other industrialized nations. In 2005, the number of fusions in this country was 332,000 and the number of inpatient spinal decompressions was 256,000. The annual costs for these surgeries alone are 22 billion dollars. There is no objective analysis to determine what percentage of these surgeries was medically necessary and to determine the appropriateness of surgery. The percentage of patients in this group that was initially treated with more conservative measures, such as spinal injections and physical medicine, prior to surgery is also unknown. Due to the high cost of surgery and the potential to achieve the same outcome with injections, physical therapy, and other conservative measures, these modalities should be tried initially in most instances.16
Patient Selection It is of primary importance to first perform a detailed history and a physical exam when evaluating a patient presenting with spinal-related pain complaints, whether cervical, thoracic, or lumbar in origin. The temporal nature, the frequency, and the quality of the symptoms should be documented. Exacerbating and alleviating factors should be determined. It is extremely important to determine the presence of any focal deficits such as bowel or bladder incontinence, limb paresthesias or paralysis, and extremity weakness. Additionally, it is important
Patient Selection
to consider constitutional illness such as tumor or infection as the etiology in back pain. The presence of unexplained weight loss, fever, new onset night sweats, or chills may necessitate additional workup, including lab studies such as CBC and ESR. The patient’s history may indicate vascular claudication as a possible cause of lower extremity pain instead of the pain having a spinal origin. The patient should be encouraged to follow up with a primary care specialist if any of these symptoms, signs, or laboratory abnormalities is discovered. The physical exam should focus on extremity strength, range of motion, sensory and proprioceptive deficits, deep tendon reflexes, coordination, asymmetry of muscle girth, areas of tenderness, gait, fine motor skills, and changes on topographical observation such as swelling around a joint, scoliosis, or kyphosis. In addition to the history and physical exam, imaging studies and other diagnostic studies may be helpful. In the setting of neurological changes, the MRI is often the test of choice. In those who are not candidates for this study, a CT or CT myelogram may give additional information. Plain films are helpful in evaluating joint disease, degenerative disc disease, spondylolithesis, spondylosis, stenosis, bony lesions, and spine orientation. Bone scans are helpful when there is concern regarding an intraosseous process such as cancer, infection, unrecognized fracture, or benign tumors. A 3-phase bone scan can be useful when considering complex regional pain syndrome. Electromyelograms and nerve conduction studies can be helpful when the differential diagnosis includes radiculopathy, neuropathy, and referred pain of nonneurogenic origin. Another important facet of the interview of a patient presenting with back pain is the previous treatment history. It is important to obtain previous imaging studies, neurological evaluation, medications, and treatment modalities, including spinal injections. If spinal injections were performed, it is useful to know what type of practitioner performed the procedure and whether the injections were fluoroscopically guided. In one analysis, it was estimated that, even in the hands of experienced practitioners, approximately 40% of nonfluoroscopically guided epidurals were not in the epidural space.17 In some settings, physicians take brief courses and then perform “spinal injections” in their office with no fluoroscopic guidance. This practice is certainly worrisome and should be discouraged. When selecting patients for a trial of epidural steroid injection, one should focus on determining which patients have symptoms consistent with cervical or lumbar radiculopathy. Epidural steroid injections appear to have a higher rate of success in patients with shorter period of symptoms and with radicular symptoms. Isolated axial back pain has not consistently responded to epidural steroid injection, and in these cases other techniques should be considered. Coagulation and Epidural Injections It is important to consider coagulation disorders and the use of anticoagulants and other agents that effect blood clotting before any spine interventions are considered. The incidence of spinal hematoma after
159
160
Chapter 8 Epidural Injections for the Treatment
epidural and spinal anesthetics has been estimated to be 1 in 150,000 and 1 in 220,000, respectively.18 It has been recommended that antifibrinolytic medications be avoided for at least 7–10 days prior to puncture of a noncompressible vessel, but data are not available to clearly delineate the waiting period for neuroaxial techniques following discontinuation of these medications.19 Anticoagulation with heparin usually results in an elevated aPTT for 2–4 h after discontinuation. Administration of unfractionated SC heparin results in a 1–2 h delay in anticoagulation and can produce unpredictable heparin levels within 2 h of administration. However, ASRA guidelines do not list low-dose SC heparin (5,000 units) as a contraindication to performing neuroaxial techniques.19 Additionally, there is a small subset of patients that may develop heparin-induced thrombocytopenia when receiving heparin for greater than 4 days. Therefore, it is advisable to obtain a platelet count in this patient population prior to performing neuroaxial procedures. Patients receiving twice daily dosing with LMWH are at increased risk of spinal hematoma compared with once daily dosing regimens.19 Therefore, many practitioners prefer to wait at least 24 h after discontinuation of LMWH prior to performing neuroaxial techniques in patients receiving the twice-daily regimen. Patients receiving single daily dosing of LMWH are generally considered safe for neuroaxial techniques 12 h after the last dose.19 Additionally, an accurate test of anticoagulation for patients receiving LMWH has not been developed. For patients receiving warfarin, it is recommended that the medication be discontinued for a minimum of 3 days and that PT and INR be normal prior to performing neuroaxial procedures.19 Antiplatet medications, such as NSAIDs, thienopyridine derivatives, and GP IIb/IIIa platelet receptor antagonists, also require special consideration when performing neuroaxial techniques in patients treated with one of these medications. Platelet function is irreversibly inhibited with aspirin and affected by other NSAIDS such as ibuprofen and Naprosyn® (Hoffmann-La Roche Inc. [Roche], Nutley, NJ) for approximately 3 days.20 Thus, patients receiving aspirin could theoretically have affected platelet function for up to 10 days. However, studies of patients undergoing epidural steroid injections while taking aspirin and other NSAIDS have failed to demonstrate any increased risks of epidural bleeding in these patients compared to patients not receiving those medications.19 Additionally, bleeding time has not been demonstrated to effectively predict the degree of hemostatic compromise in patients taking NSAIDS.21 COX-2 inhibitors do not affect platelet activity because platelets do not express COX-2. To date, there are no studies evaluating the safety of thienopyridine derivatives and GP IIb/IIIa platelet receptor antagonists. However, there are case reports of epidural hematomas with neurological compromise in patients taking thienopyridine derivatives and GP IIb/ IIIa platelet receptor antagonists; these case reports have led to the recommendation of discontinuing these medications 5–7 days before performing epidural injections.22 ASRA guidelines suggest waiting 14 days after discontinuing ticlopidine and 7 days after discontinuing
Use of Fluoroscopic Guidance
clopidogrel before performing any neuroaxial injections.19 It is important that all patients consult with the physician prescribing the anticoagulation medications to obtain permission prior to discontinuing these medications. These medications often are used as a prophylaxis for stroke, myocardial infarction, or other vascular event, so the decision to discontinue them should not be in the hands of the injecting physician unless they are also treating the vascular lesion. Contraindications Contraindications to epidural steroid injection include localized infection, sepsis, anticoagulation or coagulopathy, conditions associated with increased ICP, and patient refusal. It is important to consider the systemic effects of epidural steroids in patients with poorly controlled diabetes, and patients should be advised to check their blood sugars and consult with the physicians treating their diabetes if there is an elevation of glucose levels.
Use of Fluoroscopic Guidance It has been reported that between 25 and 40% percent of epidural injections performed without fluoroscopy are not placed in the proper place even in the hands of experienced physicians.23,24 Aside from the obvious lack of efficacy that would result from intramuscular injection of steroids meant for epidural delivery, inadvertent intrathecal injection can result in possible complications such as arachnoiditis, possible nerve injury from preservative containing solutions, aseptic meningitis, cerebral vein thrombosis, adhesive arachnoiditis, and pneumocephaly.25,26 Additionally, it is possible that intravascular placement of the needle in radicular arteries can lead to spinal cord injury secondary to a thrombotic event or vasospasm. Although many radiologists employ computer tomography when performing epidural injections, fluoroscopy-guided injections require less time to perform, are less expensive, and are equally precise.27 Johnson and colleagues reported the use of fluoroscopy-guided epidural steroid injections in a large cohort of more than 5,000 patients with a complication rate of 0.07%.28 In the authors’ experience, fluoroscopy also increases efficiency by allowing the physician to clearly delineate the bony landmarks necessary for correct placement of the epidural needle irrespective of the patient’s body habitus. When utilizing fluoroscopy, epidurography confirms epidural placement and reduces the incidence of intravascular injection by observing contrast washout. The clinician should weigh the benefit of using contrast against the risk of adverse reactions. Selection of contrast medium should be based upon the specific patient and institution preference. Typically, a preservative-free, nonionic iodine containing contrast medium such as Ominpaque™ 300 (GE Healthcare, Waukesha, WI) is used. Alternatively, one may use gadolinium in patients who are at risk for reaction to iodine-containing contrast solutions. A recent report
161
162
Chapter 8 Epidural Injections for the Treatment
by Safriel and colleagues demonstrated the safety of gadolinium when used during spinal pain management procedures such as epidural steroids, facet blocks, and diskography.29 Recent questions regarding gadolinium excretion in those with renal disease has led to caution in this patient group.30
Epidural Techniques Intralaminar The level and side of approach for an intralaminar epidural steroid injection are chosen based on the dermatomal distribution of pain and radiographic evidence from plain radiographs, MRI, or CT scan. It is sometimes necessary to perform the injection at a level that is different from the level of pathology because of factors such as surgical scarring, hardware placement, osteophytes, and critical spinal stenosis. Intralaminar injections are routinely performed by two basic methods. The first method employs a 22-gauge Quinke spinal needle (Becton Dickinson and Co, Franklin Lakes, NJ) that is often angulated approximately 20° at the tip. After sterile prep with betadine or the sterilizing agent of choice and draping of the appropriate area, fluoroscopy is brought to the field. Standard AP view of the spine at the level to be injected is obtained. The endplates of the vertebrae are then aligned, and the spinous processes are placed in the midline. A paramedian approach may be preferred in some cases as this method has been reported to be faster, has resulted in less paresthesias and dural punctures, and avoids many of the bony structures that may interfere with needle advancement in patients with narrow disc spaces.31 The insertion site of the needle is usually chosen over the laminae of the vertebrae just below the level of the interspinous space desired for medication delivery. The insertion site is usually located on an imaginary line connecting the pedicles of the vertebra and just lateral to the spinous process of the vertebrae. The skin and subcutaneous tissues overlying the needle insertion site are then anesthetized with 1% lidocaine. Through the anesthetized tract, the angulated needle is then advanced under intermittent fluoroscopy using the angulated tip to make small corrections in the direction of the needle until the superior endplate of the laminae is contacted. Rotation and advancement of the angulated needle toward the interspinous space allows for passage through the ligamentum flavum using saline or air loss of resistance or hanging drop technique. After negative aspiration, the clinician may choose to inject a small volume (1–5 cc) of preservative-free nonionic contrast medium for epidurography. A washout of contrast should suggest vascular uptake and necessitates repositioning of the needle. Contrast can be helpful when there is doubt about the proper epidural placement of the needle, and use of lateral fluoroscopic views can be obtained to demonstrate anterior and posterior deposition of the contrast. Instillation of the steroid solution is then performed in 0.5 cc increments, as injection can be painful, especially in patients with
Epidural Techniques
Table 8.1. Commercially available steroid preparations Steroid
Available concentrations
Dose
Depo-medrol® (Pfizer, New York, NY) (methylprenisolone)
40 mg/ml, 80 mg/ml
40–80 mg
Celestone® (Schering, Kenilworth, NJ) (betamethasone)
6 mg/ml
6–12 mg
Kenalog® (Bristol-Myers Squibb, New York, NY) (triamcinolone)
25 mg/ml, 40 mg/ml
50–80 mg
Decadron® (Merck, Whitehouse Station, 4 mg/ml, 10 mg/ml NJ) (dexamethasone)
4–10 mg
Refer to steroid monographs such as those available on www.mdconsult.com for relative potencies.
spinal stenosis. It is also prudent to instruct the patient to inform the physician if any paresthesias or pain is perceived during injection, as this may indicate intraneural injection. Options for injection vary as to type of steroid and dosing options (Table 8.1). In most cases, the steroid is delivered in a 2–5 cc solution in 0.5 cc increments with intermittent aspiration. The steroid can be delivered by extension tubing or by using the dorsum of the noninjecting hand to stabilize the needle. Alternatively, a similar procedure is employed that utilizes a 20- or 22-gauge Tuohy needle. After sterile prep and drape, the skin and subcutaneous tissue overlying the interspinous space desired for delivery of medication is anesthetized just lateral to midline. The anesthetized skin is then nicked with an 18-gauge angiocath needle to allow for easier introduction of the Tuohy needle. The Tuohy needle is then passed through the anesthetized tract under intermittent fluoroscopy until the characteristic resistance is felt of the ligamentum flavum. This technique is usually performed without contact of the lamina due to the increased discomfort of the larger Tuohy needle and the decreased ability to make direction changes with Tuohy needle compared with the angulated Quinke (Becton Dickinson and Co, Franklin Lakes, NJ) needle. Entry into the epidural space is confirmed by saline or air loss of resistance or hanging drop technique. Some physicians prefer to advance the Tuohy under fluoroscopic guidance in the lateral view to avoid inadvertent dural puncture. Epidurography and instillation of the medication are then performed as previously described. Caudal Caudal epidurals are typically performed for patients with degenerative lumbar spine stenosis,32 low back pain with radiculopathy, symptoms consistent with involvement of the sacral roots, pelvic pain conditions such as orchialgia or vulvodynia, and in patients with anatomical pathology or hardware that would preclude intralaminar or transforaminal approach. Additionally, caudal epidurals are technically
163
164
Chapter 8 Epidural Injections for the Treatment
easier in most patients compared with the intralaminar or transforaminal approaches as the sacral hiatus is usually easily palpated. Most physicians advocate the use of either intermittent or real-time fluoroscopic imaging.32,33 The incidence of incorrect needle placement in caudal epidurals has been reported to be between 25 and 38.5% when performed without fluoroscopic guidance17,24 Additionally, intravenous leakage has been reported to be as high as 40% during caudal epidural placement.33 After identification of the sacral cornua, the sacral hiatus, which is formed from incomplete fusion of the S4 and S5 vertebrae, is located in the midline. The sacral hiatus can be located by using the index finger of the nondominant hand to locate the tip of the coccyx. The sacral hiatus is approximately 4 cm cephalad from the tip of the coccyx and is usually easily palpated by gently rocking the index finger. Fluoroscopy aids in identification of the sacral hiatus, especially in obese patients. Next, the natal cleft and area surrounding the sacral hiatus is prepped with betadine or other preferred agent and draped in a sterile fashion. The skin and subcutaneous tissues overlying the sacral hiatus are then anesthetized with 1% plain lidocaine. Depending upon the physician’s preference, either a Quinke (Becton Dickinson and Co, Franklin Lakes, NJ) spinal needle or a Tuohy needle may be advanced under intermittent fluoroscopy through the sacrococcygeal ligament. As the needle penetrates the sacrococcygeal ligament, a characteristic pop is usually felt. The needle is then advanced 0.5 cm. Injection of preservativefree saline and observation of the subcutaneous tissue for expansion ensures that the needle is not in the subcutaneous tissue. In the absence of a contrast allergy, the use of this injectate can be helpful in identifying spread to the nerve roots, the presence of intravascular needle placement, and leakage out of the ligament suggesting the need to advance the needle. Spread in the cerebral spinal fluid is possible and would be identified by the use of contrast. A lateral fluoroscopic view is often helpful in delineating this occurrence (Figure 8.1). The total volume of injectate delivered during caudal epidural steroids is controversial. Some physicians believe that volumes of 15–20 cc are required to reach the lumbar roots, but others have documented proper spread to the lumbar nerve roots with volumes of 5 cc or less. Catheter Techniques Epidural steroid administration by a catheter-directed technique can be performed via a caudal approach or intralaminar lumbar, thoracic, cervical, and retrograde sacral approaches. Advantages of this technique include ability to deliver medication continuously over an extended period of time, selective delivery of medication at multiple nerve root levels, and the capability to access the epidural space below the conus medularis and deliver medication to the thoracic and cervical levels by an indirect catheter method. This method employed to deliver epidural steroids is similar to the intralaminar technique. The epidural space is accessed with either a midline or paramedian approach with a Tuohy or RK epidural needle.
Epidural Techniques
Figure 8.1. Caudal epidurogram lateral view.
After negative aspiration for CSF and blood, epidurography is performed to confirm accurate needle placement. An epidural catheter is then introduced in a modified Seldinger technique and advanced under intermittent fluoroscopy to the desired level for treatment. Contrast media may be injected prior to medication delivery to ensure preferential spread of the medication to the desired area. Additionally, it may be helpful to obtain both AP and lateral fluoroscopy views to ensure posterior placement of the catheter in the epidural space. The cylindrilical shape of the space can lead to placement of the catheter in the ventral spinal column. Epidural steroid delivery by catheter is more time-consuming than both the transforaminal and intralaminar techniques and carries some additional risks. Inadvertent dural puncture is possible during advancement of the catheter to the desired level of treatment. This complication is more serious when a tear or rent occurs, which may require neurosurgical correction. Additionally, placement of an epidural catheter, for prolonged periods and in close proximity to the anus, increases the risk of contamination of the epidural space and, thus, the risk for epidural abscess. It is only utilized in approximately 9% of private practices and 11% of academic practices.34 The use of this approach in the cervical spine is being advocated by some clinicians to avoid the cervical transforaminal approach, which has been associated with catastrophic neurological outcomes.35 There are no peer-reviewed data to suggest that this improves safety when compared to the other technique.
165
166
Chapter 8 Epidural Injections for the Treatment
Transforaminal In select patients, transforaminal epidural steroid administration may be more efficacious than traditional intralaminar approach.36 In fact, the first reported use of epidural steroids by Robecchi utilized the transforaminal approach.6 Patients presenting with radiographic evidence of one or two lesions and a corresponding dermatomal distribution of radicular symptoms may be suitable candidates for the transforaminal technique. Transforaminal epidural steroid injection can be performed at the cervical, thoracic, lumbar, and sacral levels. Several nerve roots can be injected by dividing the total dose of medication between the targeted levels.
Cervical Recent case reports and literature reviews have questioned the safety of performing cervical transforaminal epidural steroid injections.37,38 Various theories have been postulated to account for the more severe complications occurring after cervical transforaminal epidural injections, which have been reported to include cerebellar herniation, and anterior spinal syndrome resulting in paralysis, stroke, and death.37 Proposed mechanisms accounting for anterior spinal syndrome include vasospasm, embolism from particulate containing steroids, and direct needle trauma to a radicular artery that supplies the anterior spinal artery or the posterior spinal artery.38,39 Huntoon performed an elegant anatomical study of 10 cadavers that demonstrated the presence of ascending and deep cervical artery branches in the posterior intervertebral foramen that would be in the path of the classic approach to cervical transforaminal epidural injections.40 These arteries supply the anterior radicular and segmental medullary arteries to the spinal cord in a subset of the cadavers studied and could explain some of the cases of anterior spinal syndrome as well as other catastrophic outcomes.40 Furman and colleagues demonstrated intravascular uptake of contrast medium during cervical transforaminal epidural injections in 19.4% of 504 subjects.41 Interestingly, the presence of blood in the needle hub during aspiration in the Furman study was observed to be only 45.9% specific.41 The use of nonparticulate steroids has been postulated to be safer than particulate containing steroids when performing cervical transforaminal epidural steroid injections as the theoretical precipitation and embolism of particulate containing steroids is avoided.42 Dreyfuss and colleagues found no statistical difference in efficacy between the nonparticulate dexamethasone and triamcinolone.42 If careful needle position is acquired with the use of fluoroscopy, then the needle approach is the following: The angle should be approximately 25–30° oblique toward the affected side; the needle should enter the inferior aspect of the foramen, then carefully enter the foramen; visualize this under an AP view first, then under an oblique view; then place a small amount of contrast 240 Omnipaque™ (GE Healthcare, Waukesha, WI) or Isovue® (Bracco Diagnostics, Inc, Princeton, NJ) 300 (Figures 8.2 and 8.3).
Cervical
Figure 8.2. Cervical transforaminal epidural steroid injection AP view.
Figure 8.3. Cervical transforaminal epidural steroid injection lateral view.
167
168
Chapter 8 Epidural Injections for the Treatment
Lumbar As with cervical transforaminal injections, there are recent case reports of paralysis after lumbar transforaminal epidural steroid injection.43 The mechanism is not clear, but likely relates to either direct trauma to a radicular artery or embolism from steroid injection into a radicular artery that directly supplies the spinal cord. The largest of these radicular arteries, the artery of Adamkiewicz, enters the spinal cord between T9 and T12 in 75% of cases and L1 and L2 in 10% of cases.44 However, the artery of Adamkiewicz can be found as high as T5 and will arise from the left in 78% of cases.44 In the case of a high thoracic origin, there may be an iliac artery derived radiculomedullary artery that supplies the lower cord that may enter the spinal cord as far caudad as the S1 foramen.43 As in cervical transforaminal injection, use of nonparticulate steroid solutions may be considered as an alternative to particulatecontaining steroid solutions. Additionally, some practitioners advocate using blunt tip needles to theoretically decrease the incidence of vascular penetration.45
Procedure Needle selection is dictated by the need for precise targeting of the intervertebral foramen. Thus, either a straight or angulated 22-gauge Quinke needle (Becton Dickinson and Co, Franklin Lakes, NJ) is often preferred when performing lumbar transforaminal epidural steroid injections. Alternatively, one may choose a small diameter blunt tip needle to offer the theoretical advantage of decreasing vascular puncture, although there may be more difficulty in proper needle placement. After sterile prep and drape of the area around the level to be targeted for medication delivery, the C-arm is obliqued approximately 15–20° so that the facet joints and pars interarticularis are clearly visualized. Next, the skin and subcutaneous tissues overlying the desired pedicle and nerve root are anesthetized with lidocaine. The needle is then advanced under intermittent fluoroscopy into the neural foramen. Some practitioners prefer to initially place the needle at the junction of the pedicle and transverse process. The needle is then directed inferiorly and advanced under intermittent fluoroscopy in the lateral view to prevent inadvertent dural puncture or shearing of the nerve root. After negative aspiration for cerebrospinal fluid and blood, place a small amount of contrast 240 Omnipaque™ (GE Healthcare, Waukesha, WI) or Isovue® (Bracco Diagnostics, Inc, Princeton, NJ) 300 (Figures 8.4 and 8.5). It is extremely important to keep the needle as superior and medial as possible to avoid damage to the nerve root and damage or injection into a radicular artery.39 One method to reduce the risk of intravascular injection is to place a small IV extension to the needle to allow for continuous fluoroscopy during injection of the contrast. Selection of medication for the transforaminal technique is centered on safety concerns and small volumes. As previously mentioned, some practitioners believe that reports of paralysis after transforaminal
Procedure
Figure 8.4. Lumbar transforaminal view right oblique view.
Figure 8.5. Lumbar transforaminal view right left oblique view.
169
170
Chapter 8 Epidural Injections for the Treatment
epidural injection result from precipitation of particulate steroids in a radicular artery and thus advocate using nonparticulate or low particulate steroid solutions such as betamethasone.46 Typically, small volumes of 1–2 cc are injected at each level when delivered by the transforaminal approach.
Selective Nerve Root Blocks Selective nerve root blocks were first reported by Macnab in 1971.47 To date there have not been any randomized, double-blind, prospective controlled studies to determine the therapeutic and diagnostic efficacy of the procedure.47,48 The procedure is performed intraforaminally for therapeutic interventions or extraforaminally for diagnostic purposes.48 Typically, selective nerve root blocks are considered in patients who have nondiagnostic physical findings, physical findings inconsistent with radiographic studies, radiographic imaging demonstrating multilevel degenerative changes, or in patients with radiographic evidence of isolated diseases that may benefit from selective administration of medication to a specific nerve root.49 Although there is some controversy over the terminology associated with selective nerve root blocks, it is generally accepted that this procedure refers to selective delivery of medication to the nerve root distal to the division of the ventral and dorsal rami.50 The selective nerve root injection, when performed for diagnostic purposes, is considered positive if the pain elicited by contact of the ventral ramus is concordant with the patient’s pain preprocedure and is relieved by injection of local anesthetic with or without steroid.48 The procedure is most commonly performed in the lumbar spine due to the same safety considerations – such as injection into the Artery of Adamkiewicz or a cervical radicular artery – that dictate caution in the cervical and thoracic levels when performing transforaminal epidural injections. Caution also should be taken in the upper lumbar spine due to anatomical variations of the Artery of Adamkiewicz, which can arise as low as L2 and be on either side of the spine. The procedure is similar to that performed for transforaminal epidural injection. After sterile prep and drape of the area around the level to be targeted for medication delivery, the C-arm is obliqued approximately 15–20° so that the facet joints and pars interarticularis are clearly visualized in a view classically referred to as the “Scotty Dog.” Next, the skin and subcutaneous tissues overlying the desired pedicle and nerve root are anesthetized with lidocaine. The needle, typically a 22-gauge Quinke spinal needle (Becton Dickinson and Co, Franklin Lakes, NJ), is then advanced under intermittent fluoroscopy into the neural foramen. The needle tip will lie a few millimeters inferior and lateral to the pedicle.49 After negative aspiration, typically 0.5 ml of nonionic contrast media is instilled under intermittent or live fluoroscopy to exclude intravascular uptake (Figure 8.6). Next, either 1 ml of a local anesthetic or a combination of local anesthetic and steroid solution is instilled with intermittent aspiration. Examination of the patient is then performed to evaluate the
Conclusion
Figure 8.6. Selective nerve root block.
results of the selective nerve root block. When using the selective nerve root approach, attention should be given to the preprocedure and postprocedure pain levels and to the neurological changes after injection. This is crucial when using the procedure as a diagnostic technique.
Conclusion Transforaminal epidural steroid injections have been purported to be more efficacious than interlaminar and caudal epidural approaches.36,38 The vast majority of literature supporting transforaminal epidural steroid injection is confined to the lumbar spine due to reports of catastrophic complications occurring in the cervical spine. Scanlon and colleagues recently reported a survey in which 78 complications were reported with the transforaminal approach to cervical epidural steroid injection.37 The most serious complications included 16 vertebrobasilar brain infarcts, 12 cervical spinal cord infarcts, and 2 combined brain/ spinal cord infarcts. Epidural steroid injections are efficacious and safe when performed by physicians who have been properly trained in the use of fluoroscopy in neuroaxial interventions in select patient populations. Studies which suggest that epidural steroids are not efficacious in patients with cervical radiculopathy, lumbar radiculopathy, and postlaminectomy
171
172
Chapter 8 Epidural Injections for the Treatment
syndrome have been primarily based on studies that were performed by inexperienced physicians, did not employ fluoroscopy, and were not double blind. There are no well designed quality studies to support abandoning epidural steroid injections in patients with axial back pain. Epidural steroid injections should be considered as an integral part of conservative treatment options in patients with cervical radiculopathy, lumbar radiculopathy, and postlaminectomy syndrome that do not require immediate surgical intervention. References 1. Deer TR, Raso LJ. Spinal cord stimulation for refractory angina pectoris and peripheral vascular disease. Pain Physician 2006;9:347–352. 2. Deer TR, Caraway DL, Kim CK, Dempsey CD, Stewart CD, McNeil KF. Clinical experience with intrathecal bupivacaine in combination with opioid for the treatment of chronic pain related to failed back surgery syndrome and metastatic cancer pain of the spine. Spine J 2002;2:274–278. 3. Boswell MV, Trescot AM, Datta S, Schultz DM, Hansen HC, Abdi S, Sehgal N, Shah RV, Singh V, Benyamin RM, Patel VB, Buenaventura RM, Colson JD, Cordner HJ, Epter RS, Jasper JF, Dunbar EE, Atluri SL, Bowman RC, Deer TR, Swicegood JR, Staats PS, Smith HS, Burton AW, Kloth DS, Giordano J, Manchikanti L, American Society of Interventional Pain Physicians. Interventional techniques: evidence-based practice guidelines in the management of chronic spinal pain. Pain Physician 2007;10:107–111. 4. Sicard MA. Les injections médicamenteuses extradurales par voie sacrococcygienne. C R Soc Dev Biol 1901;53:396–398. 5. Cathelin F. Une nouvelle voie d'injection rachidienne. Methodes des injections epidurales par Ie precede du canal sacre. Applications a l'homme. C R Soc Dev Biol 1901;53:452–453. 6. Robecchi A, Capra R. Hydrocortisone (compound F); first clinical experiments in the field of rheumatology. Minerva Med 1952;43:1259–1263. 7. Landau WM, Nelson DA, Armon C, Argoff CE, Samuels J, Backonja MM. Assessment: use of epidural steroid injections to treat radicular lumbosacral pain: report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology 2007;69:614, author reply 614–615. 8. Bush K, Hillier S. A controlled study of caudal epidural injections of triamcinolone plus procaine for the management of intractable sciatica. Spine 1991;16:572–575. 9. Carette S, Leclaire R, Marcoux S, Morin F, Blaise GA, St-Pierre A, Truchon R, Parent F, Lévesque J, Bergeron V, Montminy P, Blanchette C. Epidural corticosteroid injections for sciatica due to herniated nucleus pulposus. N Engl J Med 1997;336:1634–1640. 10. Frymoyer JW. Back pain and sciatica. N Engl J Med 1988;318:291–300. 11. Andersson GB. Epidemiology of low back pain. Acta Orthop Scand Suppl 1998;281:28–31. 12. Rubin DI. Epidemiology and risk factors for spine pain. Neurol Clin 2007;25:353–371. 13. Croft PR, Lewis M, Papageorgiou AC, Thomas E, Jayson MI, Macfarlane GJ, Silman AJ. Risk factors for neck pain: a longitudinal study in the general population. Pain 2001;93:317–325. 14. Cote P, Cassidy JD, Carroll L. The Saskatchewan Health and Back Pain Survey. The prevalence of neck pain and related disability in Saskatchewan adults. Spine 1998;23:1689–1698.
Conclusion 15. Katz JN. Lumbar disc disorders and low-back pain: socioeconomic factors and consequences. J Bone Joint Surg Am 2006;88(Suppl. 2):21–24. 16. Weinstein JN, Tosteson TD, Lurie JD, Tosteson AN, Hanscom B, Skinner JS, Abdu WA, Hilibrand AS, Boden SD, Deyo RA. Surgical vs nonoperative treatment for lumbar disk herniation: the Spine Patient Outcomes Research Trial (SPORT): a randomized trial. JAMA 2006;296:2441–2450. 17. Renfrew DL, Moore TE, Kathol MH, el-Khoury GY, Lemke JH, Walker CW. Correct placement of epidural steroid injections: fluoroscopic guidance and contrast administration. Am J Neuroradiol 1991;12:1003–1007. 18. Tryba M. Epidural regional anesthesia and low molecular heparin. Pro Anasthesiol Intensivmed Notfallmed Schmerzther 1993;28:179–181. 19. Horlocker TT, Wedel DJ, Benzon H, Brown DL, Enneking FK, Heit JA, et al. Regional anesthesia in the anticoagulated patient: defining the risks (the second ASRA Consensus Conference on Neuraxial Anesthesia and Anticoagulation). Reg Anesth Pain Med 2003;28:172–197. 20. Cronberg S, Wallmark E, Soderberg I. Effect on platelet aggregation of oral administration of 10 non-steroidal analgesics to humans. Scand J Haematol 1984;33:155–159. 21. Rodgers RP, Levin J. A critical reappraisal of the bleeding time. Semin Thromb Hemost 1990;16:1–20. 22. Benzon HT, Wong HY, Siddiqui T, Ondra S. Caution in performing epidural injections in patients on several antiplatelet drugs. Anesthesiology 1999;91:1558–1559. 23. White AH. Injection techniques for the diagnosis and treatment of low back pain. Orthop Clin North Am 1983;14:553–567. 24. White AH, Derby R, Wynne G. Epidural injections for the diagnosis and treatment of low-back pain. Spine 1980;5:78–86. 25. Abram SE. Intrathecal steroid injection for posttherapeutic neuralgia: what are the risks? Reg Anesth Pain Med 1999;24:283–285. 26. Mateo E, Lopez-Alarcon MD, Moliner S, Calabuig E, Vivo M, De Andres J, Grau F. Epidural and subarachnoidal pneumocephalus after epidural technique. Eur J Anaesthesiol 1999;16:413–417. 27. Hodge J. Facet, nerve root, and epidural block. Semin Ultrasound CT MR 2005;26:98–102. 28. Johnson BA, Schellhas KP, Pollei SR. Epidurography and therapeutic epidural injections: technical considerations and experience with 5334 cases. Am J Neuroradiol 1999;20:697–705. 29. Safriel Y, Ali M, Hayt M, Ang R. Gadolinium use in spine procedures for patients with allergy to iodinated contrast – experience of 127 procedures. Am J Neuroradiol 2006;27:1194–1197. 30. FDA: Public Health Advisory – Gadolinium-containing Contrast Agents for Magnetic Resonance Imaging (MRI): Omniscan, OptiMARK, Magnevist, ProHance, and MultiHance, 2007. 31. Leeda M, Stienstra R, Arbous MS, Dahan A, Th Veering B, Burm AG, Van Kleef JW. Lumbar epidural catheter insertion: the midline vs. the paramedian approach. Eur J Anaesthesiol 2005;22:839–842. 32. Botwin K, Brown LA, Fishman M, Rao S. Fluoroscopically guided caudal epidural steroid injections in degenerative lumbar spine stenosis. Pain Physician 2007;10:547–558. 33. Ergin A, Yanarates O, Sizlan A, Orhan ME, Kurt E, Guzeldemir ME. Accuracy of caudal epidural injection: the importance of real-time imaging. Pain Pract 2005;5:251–254. 34. Cluff R, Mehio AK, Cohen SP, Chang Y, Sang CN, Stojanovic MP. The technical aspects of epidural steroid injections: a national survey. Anesth Analg 2002;95:403–408, table of contents.
173
174
Chapter 8 Epidural Injections for the Treatment 35. Amar AP, Wang MY, Larsen DW, Teitelbaum GP. Microcatheterization of the cervical epidural space via lumbar puncture: technical note. Neurosurgery 2001;48:1183–1187. 36. Ackerman 3rd WE, Ahmad M. The efficacy of lumbar epidural steroid injections in patients with lumbar disc herniations. Anesth Analg 2007;104:1217– 1222, tables of contents. 37. Scanlon GC, Moeller-Bertram T, Romanowsky SM, Wallace MS. Cervical transforaminal epidural steroid injections: more dangerous than we think? Spine 2007;32:1249–1256. 38. Rathmell JP, Aprill C, Bogduk N. Cervical transforaminal injection of steroids. Anesthesiology 2004;100:1595–1600. 39. Baker R, Dreyfuss P, Mercer S, Bogduk N. Cervical transforaminal injection of corticosteroids into a radicular artery: a possible mechanism for spinal cord injury. Pain 2003;103:211–215. 40. Huntoon MA. Anatomy of the cervical intervertebral foramina: vulnerable arteries and ischemic neurologic injuries after transforaminal epidural injections. Pain 2005;117:104–111. 41. Furman MB, Giovanniello MT, O'Brien EM. Incidence of intravascular penetration in transforaminal cervical epidural steroid injections. Spine 2003;28: 21–25. 42. Dreyfuss P, Baker R, Bogduk N. Comparative effectiveness of cervical transforaminal injections with particulate and nonparticulate corticosteroid preparations for cervical radicular pain. Pain Med 2006;7:237–242. 43. Huntoon MA, Martin DP. Paralysis after transforaminal epidural injection and previous spinal surgery. Reg Anesth Pain Med 2004;29:494–495. 44. Alleyne Jr CH, Cawley CM, Shengelaia GG, Barrow DL. Microsurgical anatomy of the artery of Adamkiewicz and its segmental artery. J Neurosurg 1998;89:791–795. 45. Rathmell JP, Benzon HT. Transforaminal injection of steroids: should we continue? Reg Anesth Pain Med 2004;29:397–399. 46. Benzon HT, Chew TL, McCarthy RJ, Benzon HA, Walega DR. Comparison of the particle sizes of different steroids and the effect of dilution: a review of the relative neurotoxicities of the steroids. Anesthesiology 2007;106:331–338. 47. Macnab I. Negative disc exploration. An analysis of the causes of nerve-root involvement in sixty-eight patients. J Bone Joint Surg Am 1971;53:891–903. 48. Gajraj NM. Selective nerve root blocks for low back pain and radiculopathy. Reg Anesth Pain Med 2004;29:243–256. 49. Fenton DS, Leo F. Image-Guided Spine Intervention, 1st ed. Philadelphia: Saunders, 2003:298. 50. Datta S, Pai U. Selective nerve root block – is the position of the needle transforaminal or paraforaminal? Call for a need to reevaluate the terminology. Reg Anesth Pain Med 2004;29:616–617, author reply 617.
9 Pulsed Radiofrequency Procedures in Clinical Practice Richard M. Rosenthal
Introduction Pulsed Radiofrequency (PRF) is a relatively new use of an older procedure, conventional thermal radiofrequency. Because PRF is reputed to be nondestructive to neural tissue, it has created a lot of interest in the pain management community. Its effect seems to occur as a result of the electrical field generated by radiofrequency current, rather than relying on thermal injury to nervous tissue, which allows for new applications in areas in which tissue heating would be contraindicated.1–3 For example, heating of the dorsal root ganglion has been associated with hyperalgesia, allodynia, dysesthesias, and deafferentation pain.3 Without the requirement to heat the dorsal root ganglion, the procedure can be applied for a wide range of uses with very minimal or no side effects. The list of uses for this valuable procedure (PRF lesioning) has grown in recent years to include acute and chronic radiculopathies, complex regional pain syndrome, shingles/post herpetic neuralgia pain, and post mastectomy/sternotomy pain, to name a few.3 Another application of PRF that would be contraindicated with conventional RF is in the treatment of peripheral neuropathies. This chapter will elucidate the use of PRF for the treatment of lumbar radicular pain in detail, and then give a brief explanation of the uses of PRF for peripheral neuropathies. Sluijter developed pulsed radiofrequency (PRF) in response to a curious phenomenon he noticed while performing a thermal procedure. During a thermal lesion, the majority of the heat surrounds the side of the needle, and very little is directed forward of the needle tip. Sluijter noticed that his patients reported good results to a heat procedure that positioned the needles in such a way that only the tips came in contact with the nerve. From this observation, he wondered whether the effect from radiofrequency current might be due to something other than heat. In addition to heat, the tissue surrounding a radiofrequency needle is exposed to an electromagnetic field.1 Sluijter From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_9, © Springer Science + Business Media, LLC 2010
175
176
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
tested the hypothesis that the electromagneticfield might be responsible for the effect by applying short bursts of current twice per second for 20 ms each. The 480 ms between bursts allowed for heat dissipation so that the needle tip temperature never exceeded 42°C. Since neural destruction occurs within about 20 s at temperatures above 45–50°C,2 he concluded that the electromagnetic field may be the cause of the clinic effect seen with RF lesioning.1 It was later reported that the electrical field was specifically responsible for the effect.1,2 The pulsed radiofrequency (PRF) mode was then added to commercially available lesion generators, and the use of PRF took off. Initially, the effect of radiofrequency current was thought to be due to thermocoagulation of nerve fibers blocking nociceptive transmission to the spinal cord. This concept was re-evaluated in light of certain findings that were inconsistent with this theory. For example, it was known that heating of the dorsal root ganglion for the treatment of acute radicular pain, due to a disc protrusion, produced only a shortterm stunning effect on sensory fibers, while the pain relief seemed to last much longer. In addition, the application of heat to the dorsal root ganglion was distal (peripheral) to the nociceptive input. Thus, the production of pain relief was not dependent on the production of a lesion between the source of nociception and the spinal cord. These elements were additional supporting evidence that led to the development of the pulsed radiofrequency procedure. As a result of the energy being delivered in pulses, the radiofrequency current no longer resulted in the development of heat. Also, pulsing the current allowed the output of the generator to be substantially increased. The usual output of voltage in the continuous mode is 15–25 V while a pulsed radiofrequency lesion is performed at 45 V.4 Controversies exist surrounding the use of PRF. Detractors are opposed to using PRF for the treatment of medial branches on the grounds that “No observational study of pulsed RF has matched the benchmarks of efficacy established for thermal RF in this arena.”5 This publication questions why it is needed when there are already proven applications with the use of conventional RF. The thermal application of radiofrequency for the treatment of painful zygapophysial joints has been well documented. Randomized double-blind studies have shown the efficacy of this modality in the treatment for cervical facet joint syndrome.6 There is also moderately supportive evidence for the lumbar procedure.7 In regards to the development of clinical proof of pulsed RF procedures, it is well known that medical evidence to prove the efficacy of any new procedure takes years to develop. Cahana and colleagues, in a recent article on PRF, points out that “New treatments evolve slowly in clinical medicine, and it usually takes a 10-year delay to accumulate sufficient clinical evidence in order to confirm or refute the value of the new treatment….”8 These authors would point out that the use of thermal RF for the treatment of lumbar z-joint pain, a well-accepted procedure, has been studied for more than 25 years, yet definitive evidence of efficacy is not robust. As to the proper and best use of PRF, Van Zundert and colleagues advocate its use at the dorsal root ganglion for the treatment of radicular pain. A prospective double-blind, sham controlled, randomized clinical trial
The Dorsal Root Ganglion Procedure
of cervical dorsal root ganglion lesioning yielded promising results.9 In a recent publication by Bogduk, he agrees that if there is any clinical utility for PFR, it should be directed at the dorsal root ganglion. He further states that he understands that investigation of its use in this area may be warranted on the basis that “there is little to offer these patients.”5 Finally, detractors have argued that PRF should not be used at all in clinical medicine until further studies demonstrating efficacy can be completed. Those in support of the procedure would disagree. Rolland Gallagher, the editor of Pain Medicine, has independently weighed the evidence in the literature and believes that currently there is enough evidence in support of the procedure to recommend its use in clinical practice if used cautiously based on algorithms and science. Gallagher also points to the fact that, unlike thermal RF, PRF does not destroy targeted nerves, thereby offering a safe non-destructive method of treating a variety of painful conditions.10 This author would point out that, if the use of pulsed RF is not promulgated in clinical practice, the studies supporting efficacy will never be done, and we will have lost a potentially valuable method of helping our patients who have chronic pain. Key Points • When radiofrequency current is delivered, it produces both heat and a powerful electric field. • The risks of damage to neural tissue is reduced or eliminated by pulsed RF due to the lower temperatures that are generated. • The use of pulsed RF offers the hope of treatment for certain painful conditions that are not amenable to conventional neurolytic procedures (heat, ice, or chemical). There are a large number of supporters of this modality who have anecdotally seen good results. • The first prospective double-blind, sham controlled, randomized clinical trail evaluating the use of PRF for the treatment of cervical radicular pain yielded promising results. • Given its positive risk/benefit ratio, there is no reason to withhold the procedure from those patients who might benefit from it until those studies are completed.
The Dorsal Root Ganglion Procedure Introduction Radicular pain is produced when the dorsal root ganglion is compressed or chemically irritated. Chemical irritation can be caused by disc protrusion, but this is not a requirement. It is believed that when the nerve is exposed to material leaking from the nucleus pulposus, a chemical inflammation is produced, resulting in sustained discharge from the nerve. While most patients will exhibit some form of nerve compromise on imaging studies, there is a subset of patients that do not. These patients will complain of radicular symptoms, yet will have a “normal” MRI. In this group of patients, it is believed that the radicular pain arises from chemical irritation of the nerve root as a result of inflammogens leaking from a damaged disc.11
177
178
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
The pain is typically described as shooting or lancinating rather than dull or aching. The patient often feels both a deep and cutaneous sensation in the affected extremity. It is more often felt below rather than above the knee. The physical examination of patients with radicular pain is nonspecific. Bogduk states, “Studies have shown that a diagnosis of lumbar disc herniation cannot be made on the basis of clinical examination.” He goes on to say that the likelihood ratios of various physical examination tests used in the clinical diagnosis of radicular pain are low. MRI testing for disc herniation is fairly reliable in identifying whether or not a herniation is present; however, it does not correlate well with the cause of the symptoms.12 Therefore, diagnostic spinal nerve block has been advocated as a means of specifically diagnosing the cause of the patients’ pain. If a specific nerve is responsible for the pain complaints, then anesthetizing that nerve should temporarily relieve it. This allows the patient to confirm or refute the diagnosis by experiencing pain relief or not, when the alleged pain generator is anesthetized.13 According to Weber and colleagues, the natural history of radicular pain indicates that 51% of males and 66% of females will have continued leg pain one year after injury.14 Saal and Saal quoted a much higher rate of recovery, up to 90% remission after one year.15 Conservative treatment options for radicular pain or radiculopathy include rest, physical therapy, other physical modalities (massage, acupuncture, TENS, heat, and cold), and medications such as anti-epileptic medications, tri-cyclic anti-depressants, anti-arrhythmics, and opiates. If these fail, the traditional approach has been to perform a series of epidural steroid injections at the involved level and, if necessary, proceed to surgery.16 Surgery has traditionally been considered the backup treatment for radiculopathy.12 The rationale in performing surgery is that the removal of the lesion causing the pain should relieve the symptoms; however, surgery is not without attendant risks. Epidural injections have been advocated as a means of obviating the need for surgery. Long-term results have been reported, but not substantiated, by studies meeting the highest level of evidence.17–19 Reports of serious complications in the literature have been widely promulgated throughout the pain management community. There have been reports of spinal cord injury and death after transforaminal epidural steroid injections in both the cervical and lumbosacral areas. Patients have been left with persistent paraplegia and quadriplegia as a result of these procedures.20–22 The theorized cause of these incidents is particulate steroid injected into an end artery that serves to support perfusion of the anterior spinal or vertebral arteries.20 There have also been reports of temporary loss of vision with incomplete recovery thought to be due to either retinal or vitreal hemorrhage.23 Aspiration to identify when the needle has been inadvertently placed into a vessel is not a reliable test.24,25 When the local anesthetic and steroid are injected in separate syringes with the operator waiting 90 s between the injections, serious permanent complications have been avoided.26 Recently, Teixeira and colleagues have suggested that pulsed radiofrequency may be a viable alternative to epidural steroid injections
The Dorsal Root Ganglion Procedure
and surgery in the treatment of acute radicular pain or radiculopathy due to a herniated intervertebral disc. They published a retrospective review of 13 patients treated only with pulsed radiofrequency who did well after their treatment: all professional patients went back to work, neurological abnormalities resolved, and numeric rating scale score fell from 7.83 to 0.27 at the final follow-up at 15.8 months. This promising result warrants further prospective study.27 Key Points • There is no history, physical exam, or anatomic study (CT, MRI) that can reliably diagnose radicular pain. • A well-performed selective nerve block (using a limited volume of a higher concentration of local anesthetic) that gives greater than 80% pain relief on two occasions may be the best method of diagnosing this condition. • Due to the potential risks associated with steroid use, an argument could be made to perform the selective nerve block without steroid, especially if a pulsed RF is planned as a longterm treatment. • A recent retrospective study in subjects diagnosed with radicular pain and scheduled for surgery showed excellent outcomes when patients were first treated with pulsed RF: the majority of patients had resolution of pain avoiding the need for surgery. None of these patients had a preliminary ESI injection. History of Pulsed Radiofrequency Lesioning The widespread use of RF current for the treatment of spinal pain began in 1980 when Sluijter and Mehta introduced a 22 g cannula through which a thermocouple probe could be inserted.28 This allowed the procedure to be performed percutaneously with minimal discomfort. The mode of action of radiofrequency current was initially attributed to the thermocoagulation of nerve fibers. This was brought into question when Sluijter performed the first pulsed RF lesion in February, 1996. He suggested that the electric field rather than temperature was responsible for the analgesic effect.1 Since thermal destruction of nervous tissue takes place at 45°C, Sluijter was careful to choose lesioning parameters that would not produce a temperature greater than 42°C (38°C is more typical). This meant that the dorsal root ganglion might be treated without the risk of deafferentation pain.29 Traditional thermal lesioning of the dorsal root ganglion has also been associated with neuroma formation, allodynia, and dysesthesias.3 Given the potential for nerve damage by heating the dorsal root ganglion, PRF introduced a method of treatment for radicular pain that had potential therapeutic efficacy without the attendant risks. Pulsed RF has found a wide application in clinic practice since its introduction. It has been used to treat patients with chronic radiculopathies without concern of producing sensory or motor loss. It has also been useful for treating painful peripheral neuropathies by performing a pulsed RF lesion directly on the affected nerve. Since there is no satisfactory surgical solution for patients with chronic radicular pain or peripheral neuropathy, this mode of treatment offers an exciting
179
180
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
and potentially valuable alternative to maintaining patients on chronic opiate therapy.30 Other targets of pulsed RF described in the literature include the sphenopalatine ganglion for neuralgia, cluster headache, and some forms of migraine headache; splanchnic nerves for the treatment of chronic abdominal pain; and the treatment of the sympathetic chain for vascular insufficiency and sympathetically mediated pain syndromes.31 Further double-blind placebo controlled studies will be needed to fully assess the usefulness of this modality of treatment in clinical practice. However, given the benign nature of the modality and reports of clinical success, it should not be withheld until such studies are completed. Key Points • Pulsed RF may treat radicular pain without causing the types of nerve damage seen with heating of the nerve. • The main applications of pulsed RF is in the treatment of radicular pain (both acute and chronic) and peripheral neuropathies. Mechanism of Action Conventional heat radiofrequency causes thermal damage to the offending nerve by creating an electrical field between the small, uninsulated electrode needle tip connected to a voltage generator and a large inactive dispersion electrode. The current flow induces the movement of ions within the tissue that alternates at the same frequency as the radiofrequency current.13 The result of this current flow is friction, which produces heat in the tissues surrounding the electrode tip. The resultant lesion is spheroid in shape with the long axis parallel to the needle tip (Figure 9.1). The size of the lesion is fairly predictable and is based on tip temperature, lesion duration, needle diameter, length of the active tip, tissue vascularity, and heat conductivity of the surrounding tissue.13,29,32 This type of lesion is placed between the nociceptive focus and the spinal cord.31 When performing a pulsed lesion, the greatest current density is projected from the tip. Thus, when performing a thermal lesion, the needle should be positioned parallel to the targeted nerve while a pulsed lesion requires perpendicular positioning.5,29,32 Pulsed radiofrequency involves delivering a burst of radiofrequency (RF) current at 500 kHz for 20 ms with a quiet period of 480 ms to allow for heat dissipation. Using 45 V or 100–150 milliamps ensures that the tip temperature stays below 42°C.29 The application time varies between 2 and 6 min depending on the targeted tissue. It is also important to keep the impedance as low as possible in order to facilitate current flow: 250 ohms is ideal, but 400 ohms is acceptable. The mechanism of action of pulsed RF is currently under investigation. Animal studies have shown that exposure of the Dorsal Root Ganglion (DRG) to pulsed RF current causes both early and late induction of a protein, C-fos, in layers 1 and 2 of the dorsal horn bilaterally. These effects are not temperature dependent33–35 and seem to occur as a result of current fluctuations, not because of tissue heating.1,2,36 This implies that exposure of the DRG to pulsed RF current induces changes
The Dorsal Root Ganglion Procedure
181
Figure 9.1. Different shapes of a pulsed vs. heat lesion. (A) Shape of lesion generated by continuous radiofrequency current. Notice that the largest part of the lesion is located around the side of the needle. This requires the needle to be placed parallel rather than perpendicular to the nerve. (B) Shape of pulsed radiofrequency lesion. Notice the greatest current density is generated from the tip of the needle. This requires perpendicular rather than parallel needle placement.
in the structure or processing of the dorsal horn neurons.1,29,32–34,36 There may also be a local neuromodulatory effect produced by the high electrical fields, resulting in neuronal membrane destruction.2 Edrine and colleagues compared the effects of thermal RF at 67°C to PRF at 42°C on the morphology of the dorsal root ganglion in rabbits. Specimens were examined 2 weeks after exposure. They found that both RF modalities resulted in increased numbers of cytoplasmic vacuoles and enlarged endoplasmic reticulum cistern compared to sham RF and control groups. However, thermal lesioning induced mitochondrial degeneration, nuclear membrane disorders, and loss of neurolemma integrity, whereas PRF did not. They concluded that PRF is less destructive to cellular morphology than heat at clinically used doses.37 This evidence would dispute the currently held belief that pulsed RF does not cause a lesion. Instead, pulsed RF currents seem to cause a small area of neurodestruction directly around the electrode tip.2,31,37 In a histological study done by Podhajsky and colleagues, the authors attempted to show that PRF delivered to the dorsal root ganglion is a safe method for the treatment of pain. To do this, they exposed rat DRG
182
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
to PRF, continuous RF, and conductive heat at a tissue temperature of 42°C for 120 s. They also exposed the sciatic nerve to 42°C in continuous mode. As a control, rat sciatic nerves were treated with continuous RF at 80°C. Animals were killed at 2, 7, and 21 days after treatment. They found minor structural changes in all groups that returned to normal after 7 days in the nerve and 21 days in the DRG. The nerves treated with continuous RF at 80°C showed expected thermal injury characterized by Wallerian degeneration of nerve fibers. They noted that rats treated with either continuous or pulsed RF at 42°C showed no signs of sensory or motor deficits, whereas, when the sciatic nerve was exposed to 80°C, rats showed immediate signs of foot drop.36 Cosman and Cosman looked at electric and thermal field effects in tissue surrounding radiofrequency electrodes. They measured both average and transitory temperature effects in liver and egg-white models. They observed rapid, transitory temperature spikes during pulsed RF bursts. They concluded that PRF produces heat bursts into the neurodestructive range based on measured heat spikes exceeding 45–50°C. They also found that PRF produced high electrical fields that may be capable of disrupting neuronal membranes. They state that these effects may be at subcellular levels that have not been well examined to date.2 In a prospective clinical study by Slappendel and colleagues, patients were treated with continuous RF current of the cervical DRG at either 67°C or 40°C. The authors noted no difference in the outcome between groups at 3-month follow-up. This seems to imply that heat is not necessary to produce the effect.38 Key Points • A heat RF lesion is spheroid along the long axis of the needle. • A pulsed RF generates an electrical field that projects from the tip of the needle. • The mechanism of pulsed RF remains unknown, but it is thought to have both local and central effects. • PRF is less destructive to cellular morphology than heat at clinically used doses. • PRF produces high electrical fields that may be capable of disrupting neuronal membranes. Anatomy There are five paired nerves that exit their respective intervertebral foramina from L1-2 to the L5-S1 levels. Just as the orientation of the lumbar zygapophyseal joint differs from L1-2 to L5-S1, the lumbar nerves exit their respective foramina at different angles from L1 through L5. At L1, the nerves exit downward and forward at an acute angle, whereas at L5, the nerves exit more horizontally and at a more obtuse angle.29 The lumbar ventral roots find their cell bodies of origin within the spinal cord at the T9-11 vertebral level.39 Rootlets come off the dorsal and ventral surface of the spinal cord to form the dorsal and ventral roots. The dorsal root ganglion contains cell bodies that provide sensation, proprioception, and pain. The dorsal and ventral
The Dorsal Root Ganglion Procedure
183
Figure 9.2. Spinal nerve, DRG, dorsal and ventral horns of spinal cord, dorsal and ventral rooflets, dorsal and ventral rami of spinal nerve.
roots then join to form the spinal nerve root.40 The location of the DRG varies (Figure 9.2). In the lumbar spine, it is located from the mid to the anterior aspect of the foramen in a saggital plane 98% of the time and is extra-foraminal in 2% of cases. At the L4 level and above, the DRG was usually located within the foramen, but at L5, 5.7% of DRGs studied were intraspinal, and 77.3% were intraspinal at S1.3,41 In the frontal plane, the DRG is located at the junction of the upper one third with the lower two thirds of the foramen. The anatomy of the arterial system is important when performing spinal injections because some of the vessels that supply the spinal nerve roots also reinforce the blood flow to the anterior spinal artery (Figure 9.3). A branch of each vertebral artery combines to form the anterior spinal artery. At several levels throughout the spine, medullary arteries form to reinforce the blood flow to the anterior spinal artery.40 On the left side of the spine, between T8 and L2, the largest medullary artery, termed the artery of Adamkiewicz, is found. There have been reports of central nervous system sequelae after the performance of cervical, lumbar, or sacral nerve root blocks.20–23 In one report, three cases of paraplegia were thought to be due to needle perforation of the artery of Adamkiewicz that originated at an unusually low location. All three cases occurred after injection of local anesthetic and steroid combined in the same syringe.22 The exact cause for this devastating outcome was not known. One theory is that particles in the steroid embolized within the territory supplied by the artery of Adamkiewicz producing spinal cord infarction.22 It is for this reason that injection
184
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
Figure 9.3. Anterior medullary artery. (Reprinted with permission from Landers MH, Jones RL, Rosenthal RM, Derby R. Lumbar Spinal Neuroaxial Procedures. In Raj PP, Lou L, Erdine S, Staats PS, Waldman SD, Racz G, Hammer M, Niv D, Ruiz-Lopez R, Heavner JE (eds): Interventional Pain Management: Image-Guided Procedures, 2nd edition. Philadelphia: Saunders, 2008. Copyright Elsevier.)
of contrast dye under live fluoroscopy is recommended. One should monitor for a flash of dye seen within the central canal, indicating the uptake of dye by the artery. A second physiological test is to inject local anesthetic first, wait 90–120 s, and then make certain the patient can move all extremities prior to injection of the steroid).26 Key Points • Knowing the anatomy is important in targeting the dorsal rot ganglion. • The arterial supply of the anterior part of the spinal cord is tenuous, and the anatomy must be known and respected. Indications The general indication for PRF is neuropathic pain that is confined to the distribution of a known nerve. The specific indication for pulsed radiofrequency treatment of the dorsal root ganglion is radicular
The Dorsal Root Ganglion Procedure
pain or radiculopathy that is completely but temporarily relieved by transforaminal injection of local anesthetic injection done on two separate occasions. The local anesthetic injections are done diagnostically to identify the location of the origin of the pain and confirm the nerve levels involved. The procedure has been used for both acute and chronic radicular pain and radiculopathy.15,32,42,43 Another indication is the treatment of peripheral nerve damage by applying pulsed radiofrequency energy to the nerve roots forming the origin of the damaged nerve. One such study reported on the use of PRF for this purpose. This study reported a case series of five patients who, after responding positively to selective nerve root blocks, underwent PRF of the T12, L1, and L2 nerve roots. All five patients reported pain reduction of 75–100% lasting 6–9 months.44 Contraindications3,13 Absolute • Patients who are unwilling or unable to give consent for the procedure. • Local or systemic infection: this could put the patient at risk for the spread of infection to the central nervous system. • History of anaphylactic reaction of contrast dye: reports of spinal cord injury after injection of local anesthetic and steroid mixed in the same syringe gave raise to concern that some component of the injection may have clogged a radicular artery, reinforcing blood flow to the anterior spinal artery. Use of contrast injected under live fluoroscopic X-ray is thought to be essential to monitor for the spread of dye into the central canal. • Bleeding disorders from disease or due to concurrently used medications: medications that may cause bleeding must be stopped at the appropriate time prior to the procedure. • Uncooperative patient: although this may not be discovered until attempting the procedure, it is worth discontinuing the procedure rather than causing permanent nerve damage. • Needle phobia that is not treatable by psychological interventions. Relative • Allergy to any medication planned to be used during the procedure. • Pregnancy. • Anatomical derangements that could prevent successful conduct of the procedure: patients with previous spinal fusion should have an Xray to evaluate access. • Coexisting disease causing cardiovascular or respiratory compromise. • Patients with immune suppression in whom the procedure could result in the introduction of infection to the central nervous system. Equipment and Drugs The operator should be immediately prepared to treat all complications that might arise as a result of the procedure. Medications and equipment should be immediately available to provide resuscitation of the patient, should the need arise. These include:
185
186
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
• Solutions for aseptic preparation of the skin: • 20 g 10–15 cm (depending on the size of patient) RFK needle with a 5mm active tip. • 5-ml syringe with 27g needle to anesthetize skin and deep tissues. • 3-ml syringe with 2% Xylocaine used to numb the targeted nerve. • 3-ml syringe with nonionic contrast dye. • Sterile gloves. • Peripheral IV cannula (optional). • Physiologic monitors are optional based on planned level of sedation and cardiopulmonary compromise of individual patient. • Preservative-free local anesthetic such as 2% lidocaine or 0.25% bupivicaine to anesthetize the targeted nerve. • 1% lidocaine to anesthetize the skin and needle path. • Informed consent should be performed prior to the procedure. History Reported by the Patient The patient will commonly complain of pain, numbness, tingling, or paresthesia in the involved extremity confined to one or two dermatomes. They may also complain of weakness. The pain is felt in the lower extremity if the lower lumbar or upper sacral nerve roots are involved. It is never felt in the back.12 Physical Exam The following exam findings may be seen alone or in combination: • • • •
Dural tension signs (positive straight leg raise, femoral stretch, etc.) Weakness in the involved muscle groups Numbness or hypoesthesia to touch or noxious stimuli Decreased deep tendon reflexes
Laboratory MRI should be done in all cases of suspected radicular pain or radiculopathy. Usually, no blood work is required, but consider CBC and/or bleeding time in selected cases. X-rays of spine are usually not required but should be considered in patients with a history of possible spinal instability such as spinal fusion and spondylolisthesis. The series should include flexion/extension views. EMG can be helpful in localizing pathology if positive, but does not exclude the diagnosis if negative. Pre-Operative Medication An oral benzodiazepine such as Valium or Xanax given 45–60 minutes prior to the procedure may be all that is required. When correctly performed, the procedure takes no more than 5–10 min, and these medications assist in providing the patient an anxiolytic and muscle relaxation. If IV sedation is considered, it should be given according to ASA guidelines.
The Dorsal Root Ganglion Procedure
Procedure Positioning The goal in positioning is to provide maximum relaxation of the back muscles, making passage of the needle easier and less painful. The patient is positioned prone on the X-ray table with the head turned to the side and arms hanging over the edge of the table (Figure 9.4a). The patient is given soft rubber balls to squeeze to transfer tension from the back muscles to the hands (Figure 9.4b). The back muscles are then further relaxed by teaching the patient deep breathing exercises.
Figure 9.4. (A) The patient is positioned prone on the table with head to the side and arms over the edge. (B) They are given a ball to squeeze. The back muscles are relaxed by assisting the patient with deep breathing. The needle is passed during the expiratory phase.
187
188
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
The patient is instructed to breathe in to the count of five, hold the breath for three counts, and then breathe out. This is done three times; then the needle is passed in time with the expiratory phase (as this is when the back is most relaxed). All of this preparation usually allows the injectionist to pass the needle directly to the target in one or two passes without misdirection by tight muscles. It also causes considerably less pain to the patient. Technique DRG-RF Although this procedure can be done at all spinal levels, this discussion will be limited to the lumbar dorsal root ganglion. The approach is well described in the ISIS guidelines in the chapter on lumbar spinal nerve block. It is referred to as a retroneural approach.13 In this case, the target lies at the intersection of two lines. When viewed in a lateral view, the first line runs longitudinally between the posterior and anterior half of the foramen, bisecting the foramen into two equal halves. The second line runs in a transverse direction between the superior one third and the inferior two thirds of the foramen.13 The intersection serves as a starting point for locating the dorsal root ganglion. However, as noted previously, the dorsal root ganglion (DRG) can lie anywhere between the mid aspect to the most anterior aspect of the foramen in the anterior–posterior plane. To achieve this target point, an oblique view is obtained identical to that used for a transforaminal procedure (Figure 9.5a). After squaring the superior endplate at the involved level, the image is obliqued until the superior articular process is projected one third of the distance across the image of the vertebral body (approximately 15–20 degrees). In this view, the starting point of the needle is slightly inferior and lateral to that used for a transforaminal injection. Aim for a point just underneath the pedicle that is one third of the way down of foramen. As the needle is advanced, rotate it into an AP projection to assess the depth of insertion (Figure 9.5b). If further insertion is required, rotate it back to an oblique view and advance. When the needle tip approaches the lateral aspect of the vertebral body (Figure 9.5c), it is the best to advance further in an AP view. Do this very slowly by pinching the needle shaft at the point of skin entry. This allows only slight advancement (a millimeter at a time) of the needle. Warn the patient that they will feel a paresthesia and so not to jump. Because the greatest current density is projected from the tip of the needle, it is imperative that the needle be pointed directly towards nerve tissue and not against the bone of the vertebral body (thus, placing the needle perpendicular to the target vs. parallel when the needle touches vertebral body). Once the paresthesia is felt, place the electrode into the needle and begin testing. The operator must be very careful in handling the needle at this time since any movement risks spearing the nerve. Electrical stimulation tests are used to determine proximity between the needle tip and the nerve. Adequate placement requires that the patient feel reproducible stimulation (tingling of the stimulated dermatome in
The Dorsal Root Ganglion Procedure
189
Figure 9.5. (A) Oblique view – note the needle position just below the pedicle and about 1/3 of the way down the cephalo-caudal extent of the foramen. Also note the position of the SAP about 1/3 of the way across the vertebral body. (B) AP view – note the location of the needle tip at approximately the six-o’clockposition relative to the pedicle. Also note that it is lower than normal location about 1/3 of the way down the foramen. (C) Lateral view – note the location of the needle tip about 1/3 of the way across the foramen in the cephalo-caudal direction and about 1/3 of the way across the foramen in the dorsal/ventral direction.
190
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
Figure 9.5. Continued.
the lower extremity) at less than 0.2 V, with the ideal stimulation less than 0.1 V. Stimulation is done twice: once to determine the minimum sensory threshold (the lowest voltage at which the patient can still perceive a sensation) and a second time to determine reproducibility. When stimulating the second time, the voltage is slowly increased until the patient reports that they perceive a stimulus. This should be within 0.05 V of the first stimulation test; if not, reposition the needle. If the patient does not feel the current at the required level of less than 0.2 V, the physician is required to advance the needle slightly (no more than 1 mm) or reposition the needle altogether and re-test. The patient should be warned that this adjustment might cause considerable pain. If the operator is having difficulty in locating the nerve, the needle tip is usually too medial and should be corrected in an oblique view such that the tip is located directly beneath the image of the pedicle. There is no need for motor stimulation when performing pulsed radiofrequency since the modality does not damage motor fibers. At this point, turn on the power in the pulsed RF mode. Slowly increase the voltage until the patient can verify that he/she feels the pulsing. The needle must be close enough to the target nerve tissue to produce a perceptible electrical discharge in the treated extremity with each pulse. In other words, during the time the pulsed RF mode is operative, the patient must report a sensation of pulsing in the appropriate dermatome;
The Dorsal Root Ganglion Procedure
otherwise, the needle is not close enough to the targeted nerve tissue to produce an effect. If no pulsing is felt, the needle needs to be repositioned prior to treatment. Although no study has demonstrated this step to be necessary, the author feels it to be another test to verify proximity to the targeted neural tissue. Once the operator has verified that the patient feels pulsing, the pulse mode may be temporarily turned off. The next step is to check impedance. The maximum impedance should be less than 400 ohms and ideally less than 250 ohms. To achieve this, inject a small amount (1 ml) of local anesthetic (2% xylocaine) very slowly, keeping in mind the needle tip may actually be slightly penetrating the axonal fibers. Intraneural injection must be meticulously avoided as it can cause devastating consequences, including irreversible sensory and motor loss. If there is any resistance during injection, stop, back the needle up slightly, and inject again; then replace the needle to its original position. Liquid should flow easily through a 20-g needle. The initial minimal stimulation threshold (MST) and impedance (I) are recorded prior to treatment. Next, precede with pulsed radiofrequency treatment for 3–4 minutes at 200 milliamps or 45 V (as long as the temperature does not exceed 42°C), 2 pulses per second with current applied for 20 ms each pulse. Treatment protocols may vary with different operators, and the author recommends consulting the literature for other examples of lesion parameters (Figures 9.6a, b and 9.7a–d). Post Treatment Patient Advice The patient often feels better immediately after the completion of the procedure due to the injection of the local anesthetic. This effect wears off in 24 h, and they begin to feel sore. Advise the patient that he/she will continue to feel sore for the first week, that he/she will feel better the second week, and that the full effect will take three weeks to develop. During this time, the patient does not need to restrict activities except as needed due to pain. Deep tissue massage once a week for the first three weeks following the procedure is recommended. Complications Bleeding and infection are commonly listed, but are rare complications. Hematoma may occur just under the skin or in the deeper muscle layers. Vaso-vagal reaction may occur during procedure performance. Mechanical nerve root damage can occur from needle trauma, especially if fluid is injected into the axon bundle. Starting in too oblique a position can result in intra-thecal placement resulting in dural puncture and/or injection of local anesthetic into the intrathecal space. This could cause a post dural puncture headache or lower extremity paralysis. Intra-thecal placement may also occur if the needle is advanced beyond the six o’clock position in a patient with long dural root sleeves. All of the complications are extremely rare. There has not been a serious long-term complication due to pulsed RF reported in the literature.
191
Figure 9.6. (A) The procedure can be done at any spinal level. Here is an open mouth view of a needle placed at C2 for the treatment of occipital headache. The patient had previously responded positively on two occasions to occipital nerve blocks. (B) Here is the initial view showing correct needle placement for a C2 PRF. The needle is pointed directly at the targeted nerve.
Figure 9.7. (A) AP view of L5 DRG dye pattern. (B) Lateral View of L5 DRG dye pattern.
194
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
Figure 9.7. Continued. (C) Starting view of L5 DRG. D: Oblique view of L5 nerve.
The Dorsal Root Ganglion Procedure
Efficacy In a prospective placebo-controlled trial by Van Zundert and colleagues, 23 patients underwent either a pulsed RF procedure adjacent to the cervical DRG or a sham lesion. Eleven patients were in the PRF group and 12 in the sham group. At the 3-month followup, 83% of patients in the PRF group reported at least 50% improvement in global perceived effect compared with 33% in the sham group. Eighty-two percent of patients reported at least a 20 point decrease in VAS vs. 25% in the sham group. PRF patients also showed a decrease in medication use. At the 6-month follow-up, 64% of PRF patient reported at least 50% improvement in global perceived effect compared with 17% in the sham group. Again, the PRF patients decreased medication use. Overall, the researchers felt that PRF provided better relief than sham intervention at 3 and 6 months after the procedure.9 A prospective pilot study on the use of pulsed RF in cervical pain syndromes showed 13 of 18 patients treated with pulsed RF to the dorsal root ganglion reported 50% relief at 8 weeks post treatment, but only 6 patients continued to report relief at one year.1 The study could be criticized for having a mixed patient group with diverse pain syndromes. There have also been multiple retrospective trials reporting various results. The most impressive of these is reported next. Teixeira and colleagues reported on the use of PRF of the DRF for the treatment of radicular pain and radiculopathy.27 They retrospectively studied 13 consecutive patients with pain originating at the L3 to S1 spinal levels. Of the 13 patients, 12 were professionals who had stopped working due to the pain. The average numeric pain score on presentation was 7.82. All patients had failed various conservative treatments, while 9/13 patients had slight to moderate loss of motor function, 9 patients had sensory loss, and 12/13 patients had paresthesias. A diagnosis of herniated intervertebral disc, concordant with clinical presentation, had been made with MRI or CT scan in all patients. All patients had been seen by a neurosurgeon and were scheduled for surgery. Prior to treatment, proximity to the targeted neural tissue was verified by sensory stimulation at 50 Hz. The mean stimulation threshold was 0.22 V with a range of 0.08–.03 V. Patients were treated with pulsed RF for 3 min at 45 V, and two pulses of RF current for 20 ms each. At the final follow-up examination, 15.8 months after the initial procedure, one patient had undergone disc surgery due to continued leg pain. The remaining 12 patients did not require surgery for leg pain, but one patient did undergo spinal fusion for back pain. Of the remaining patients, average pain score was 0.27, and all 12 patients with professions had returned to work two weeks after the procedure. Neurological abnormalities resolved in 11 of the 12 patients; one patient had decreased sensation in a small area of the thigh innervated by the
195
196
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
L3 dermatome. The average Numeric Rating Scale (NRS) score fell significantly one week after the procedure. The results were statistically sinificant at 4 weeks post procedure. Motor and sensory loss improved significantly 2 weeks post treatment. Another retrospective study looked at the clinical effect of PRF lesioning of the DRG in patients with radicular pain of varying etiologies. There were three patient groups divided according to etiology of the radicular pain: Herniated Disc (HD), Spinal Stenosis (SS), and Failed Back Surgery Syndrome (FBSS). The results showed a decrease in numeric rating scale in patients with HD and SS but not FBSS. They concluded that PRF of the DRG was significantly more effective in patients with HD and SS than with FBSS. Of note: they do not mention their stimulation parameters prior to lesioning. They reported that 40% of patients in the HD and SS group reported “successful treatment” at 180 days post treatment. 42 Ahadian reported on the use of PRF for a variety of painful disorders in a retrospective study. At three month follow-up, he found fair to excellent relief in 40% of patients with cervical radiculopathy, 64% of patients with lumbar radiculopathy, and 56% of patients with peripheral nerve lesions of varying sorts.43 Cahana and colleagues performed an extensive literature search on the topic of pulsed RF.8 They found 58 reports on the clinical use of the procedure noting mixed results. Currently, there is a great deal of interest in the topic, and further studies are in progress. At present, the trend seems to support the efficacy of the procedure, but, as previously noted, further studies need to be done.
Author’s Opinion Looking at the data presented in this chapter and at personal experience, the patients with stimulation parameters between 0.05 and 0.15 V have the longest lasting success. Values below 0.05 V may reflect intraneural placement. 4 Since there is evidence that a small lesion does occur around the electrode tip, it may be unwise to allow the electrode to penetrate neural tissue.4 For pulsed RF to be effective, the electrode must be carefully and meticulously positioned, pointing directly perpendicular and very close to the targeted nerve. The operator should have very soft hands and excellent needle handling skills before attempting this procedure. The patient must also be warned just prior to eliciting a paresthesia and told not to make any sudden moves. When injecting fluid through the needle, its position must be secured by one hand at the skin and not be allowed to move, lest it cause a severe pain and possible needle trauma to the nerve. Also, no injection should be made if there is even slight resistance to the flow of the fluid. Instead, pull back slightly, complete the injection, and then reposition the needle.
Peripheral Nerve PRF Procedures
Billing During the performance of the procedure, local anesthetic and steriod are injected to anesthetize the nerve prior to lesioning. The CPT code 64483 reads “Injection, ... of diagnostic or therapeutic substance(s) (including anesthetic, ....).” Therefore, this code can be appropriately billed in this situation. Since there is no CPT code for dorsal root ganglion lesion, the code 64999 must be used. If payment is denied, the physician may decide to add a small charge for that as a separate procedure as well. In other words, the transforaminal injection and the dorsal root ganglion injection are billed as two separate procedures.
Peripheral Nerve PRF Procedures Peripheral nerve damage can occur in a number of ways, including previous surgery to the area, compression or crush injuries, and various chemical etiologies. Cohn and Griffith state, “following peripheral nerve injury, ion channel modulation occurs leading to nociceptor sensitization, expansion of receptive fields, diminished central inhibition, increased neuronal excitability in the spinal cord, and reorganization in the dorsal horn.”3 The results are hyperexcitability and increased spontaneous firing of damaged nerves (Figure 9.8). There are currently few options for patients suffering from peripheral nerve damage, and the available methods are complicated, risky, and/or expensive. PRF is a relatively simple, inexpensive procedure that can produce long-term pain relief without the risks that frequently accompany other methods. The general principles for the treatment of peripheral nerve injury using PRF are similar to those described in the techniques for the treatment of DRG. As in the DRG procedure, the electrode must be meticulously positioned directly perpendicular to, and very close to, the targeted nerve. However, when performing peripheral nerve lesioning, the nerve is usually located using surface landmarks and by palpation of the artery associated with the nerve. In most cases, fluoroscopy is not used. The anatomy and indications vary depending on the involved nerve. These are well described in Waldman’s Atlas of Interventional Pain Management.45 These procedures can also be more problematic than the DRG procedure because most peripheral nerves are not very deep in the tissue. This results in the needle wobbling within the tissue, which makes it difficult to hold the tip in position, especially because the head of the needle is heavy relative to the tip. It requires a great deal of attention and concentration to keep the tip pointed directly over the targeted nerve. The technique can be used on a number of peripheral nerves, including the occipital nerve, lateral femoral cutaneous nerve, ilioinguinal nerve, tibial nerve, and others. The sphenopalatine ganglion is another logical target for a pulsed radiofrequency procedure given the proximity of the maxillary nerve (a somatic nerve that, when heated, can cause a deafferentation pain). This technique is not recommended for intercostal nerves; the DRG procedure is a safer and more reliable
197
198
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
Figure 9.8. Process of peripheral nerve damage.
technique for this area of the spine as it avoids the possible complication of pneumothorax. In one study, the authors found that PRF of the dorsal root ganglia gave superior relief to medical management and PRF of intercostal nerves. They also found that in the patients who responded positively to PRF, pulsing the DRG lasted twice as long as PRF of intercostal nerves (2.87 months in the intercostals group vs. 4.74 months in the DRG group).44,46 Typical Peripheral PRF Procedures Suprascapular Nerve Treatment Using Radiofrequency Aided by Fluoroscopy Anatomy The suprascapular artery and vein travel with the nerve. The suprascapular nerve originates from the ventral ramus of C5 and C6. It travels inferiorly, posteriorly, and laterally to run under the transverse scapular ligament before entering the suprascapular notch. It is at this point the nerve is accessible to block and PRF. The nerve then travels to the muscles of the shoulder girdle and the posterior joint capsule. The nerve is a mixed motor and sensory nerve that also contains nociceptive fibers. Seventy percent of the sensory information to the shoulder girdle is provided by this nerve.45,47
Treatment of Post Surgical Pain after Inguinal Herniorrhaphy using Surface Landmarks
Indications The procedure is indicated for patients with shoulder pain who have been evaluated by a shoulder specialist and found not to be a candidate for surgery. Many of these patients have had a prior rotator cuff repair and still have pain. It is also indicated for patients with a frozen shoulder from any cause. Patients must have responded positively (70–100% reduction in pain) to blockade of the nerve with local anesthetic on two separate occasions are candidates for PRF. Technique Fluoroscopy is used to perform the procedure. The patient is placed in a prone position on the table. It can sometimes be difficult to visualize the suprascapular notch. The technique is to angle the c-arm in a cephalocaudal direction and then oblique towards the affected side. The suprascapular notch is identified just above the spine of the scapula and medial to the coricoid process.47 After sterile prep and drape, a 5 cm RFK needle with a 4 mm active tip is advanced down the angle of the X-ray beam in order to position the needle perpendicular to the nerve (Figure 9.9a, b). Next, the nerve is stimulated at 50 Hz with the intention of obtaining a paresthesia in the shoulder at less than 0.2 V. This should be repeated to test for reproducibility. It is not necessary to perform motor stimulation, as PRF does not affect motor fibers. Then the impedance is checked. It should register at less than 400 ohms. If not, saline is injected. It is important not to numb the targeted nerve in order to ensure that the patient feels continued pulsing throughout the entire procedure. Finally, the nerve is lesioned in the pulsed mode for 4 min at 2 Hz, and 45–60 V. It is not necessary to measure temperature as long as it stays below 42°C. The operator should hold the needle in place throughout the procedure and occasionally ask the patient if they continue to feel pulsing. If not, it implies that the tip has moved away from the targeted nerve and is no longer producing any effect. A visible pulsation of the supraspinatus and infraspinatus muscles will be seen.47 At the completion of the procedure, ½–1 ml of 3% phenol can be injected through the needle if the operator plans to bill insurance for the procedure. This concentration is adequate to cause neurolysis of the small pain fibers without damaging the motor or sensory fibers.
Treatment of Post Surgical Pain after Inguinal Herniorrhaphy using Surface Landmarks Introduction Inguinal hernia repairs are a commonly performed surgical procedure. Although rare, the procedure presents the possibility of ilioinguinal nerve damage. This can leave patients with a chronic debilitating neuropathic pain.44 Pain medicine specialists most often treat these patients. Many different treatment options have been tried, from chemical neurolysis to peripheral nerve stimulation using spinal cord stimulator leads over the damaged nerve. Use of PRF may provide these patients a long-term solution without risking further nerve damage.
199
Figure 9.9. (A) This is an image of a 5-cm needle directed at the suprascapular nerve located within the notch. (B) This provides a good view of the suprascapular notch with a RF needle located medial to the target point.
Treatment of Post Surgical Pain after Inguinal Herniorrhaphy using Surface Landmarks
Indications Patients usually present with a burning or dysesthetic pain in the lower abdomen, which radiates to the genital area. The patient usually has a history of some type of trauma to the area, most often from prior surgery.48 Diagnosis The condition is most often diagnosis on an anatomical level by positive response to an ilioinguinal nerve block after more serious conditions have been considered and ruled out. Anatomy The ilioinguinal nerve originates from the ventral roots of L1 and sometimes the T12 level. The nerve then travels inferiorly towards the ilium. The nerve perforates the transversus abdominis muscle at the level of the anterior superior iliac spine.48 It then pierces the fascia of the external oblique muscle. It is here that the nerve can be blocked. Technique PRF of the nerve is performed in the same manner as the block. The patient lies supine with a pillow under the knees to relieve traction on the nerve. A point 2 in. medial and 2 in. inferior to the anterior superior iliac spine is identified. A sterile prep is applied to the area. After anesthetizing the skin, needle entry occurs at this point.48 A 5-cm SMK needle with a 4-mm active tip is most often used. The operator must search for a paresthesia using either the needle tip or a nerve stimulator. When stimulated, the patient will report a sensation that travels from the lower abdomen toward the groin. The operator should use the needle to feel for the most superficial muscle layer, the external oblique muscle. Do not go any deeper as the abdominal cavity may be penetrated, resulting in an infection.48 Once a paresthesia is elicited, the pulsed mode is activated without anesthetizing the nerve. The nerve must not be anesthetized since the patient should feel the pulsations of electricity during the entire procedure. This assures the operator that contact with the nerve is not lost during the procedure. As mentioned earlier, this requires careful concentration, as the head of the needle tends to make the tip wobble. If the patient no longer feels pulsing, the needle should be relocated until they do. At the completion of the procedure, ½–1 ml of 3% phenol can be injected through the needle if the operator plans to bill insurance for the procedure. This concentration is adequate to cause neurolysis of the small pain fibers without damaging the motor or sensory fibers. Complications and Efficacy of Peripheral Nerve PRF Side effects and complications are minor and vary according to the location of the procedure being performed. The efficacy of most proce-
201
202
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice
dures has only been studied retrospectively, but with promising results. For example, the Ahadian study found fair to excellent relief in 56% of patients on whom various peripheral nerve PRF procedures were performed.43 Given that there is not much else to offer these patients (short of spinal cord or peripheral nerve stimulation), these results warrant attempting this simple and inexpensive procedure before deciding on more costly remedies.
Summary Although the clinical evidence showing its efficacy is still accumulating, there is enough evidence currently to recommend the use of PRF in the treatment of pain. For many practitioners, both in the US and abroad, PRF has become a standard modality of treatment for a variety of conditions. At this point, its best potential is in the treatment of radicular pain and peripheral nerve damage. The previously reported prospective study reported good results. A retrospective study showed that PRF is effective as a replacement for both epidural steroid injections and surgery in the treatment of radicular pain. PRF can be used in cases of peripheral nerve damage as a safe alternative to more aggressive treatments, such as chemical neurolysis, which can result in further injury. Additionally, PRF is considerably less expensive than spinal cord and peripheral nerve stimulators, which have been traditionally been options. There is still much unexplored potential for the use of PRF as a safe, effective, and affordable option in pain management. As new studies are conducted and the modality is more widely utilized in clinical practice, these new uses will be uncovered. References 1. Sluijter ME, Cousam E, Rittman W, van Kleef M. The effects of pulsed radiofrequency fields applied to the dorsal root ganglion-a preliminary report. Pain Clinic 1998;11:109–117. 2. Comsan E Jr., Cosman E Sr. Electrical and thermal field effects in tissue around radiofrequency electrodes. Pain Medicine 2005;6(6):405–424. 3. Cohn S, Griffith S. Dorsal root ganglia radiofrequency procedures. In Manchikanti L, Singh V (eds): Interventional Techniques in Chronic Spinal Pain. Paducah, KY: ASIPP Publishing, 2007:623–632. 4. Van Kleef, Sluijter M, van Zundert V. Radiofrequency treatment. In Benzon HT, Rathmell JP, Wu CL, Turk DC (eds): Raj’s Practical Management of Pain. 4th ed. Philadelphia: Mosby, 2008. 5. Bogduk N. Letter to the Editor. Pain Medicine 2007;8(4):390–391. 6. Lord S, Bansley L, Wallis B, McDonald G, Bogduk N. Percutaneous radiofrequency neurotomy for chronic cervical zygapophyseal-joint pain. N Engl J Med 1996;335:1721–1726. 7. Dreyfuss P, Halbrook B, Pauza K, Joshi A, McLarty J, Bogduk N. Efficacy and validity of radiofrequency neurotomy for chronic lumbar zygopophysial joint pain. Spine 2000;25:1270–1277. 8. Cahana A, Van Zundert J, Macrea L, van Kleef M, Sjuijter M. Pulsed radiofrequency: current clinical and biological literature available. Pain Med 2006;7(5):411–423.
Summary 9. Van Zundert JV, Patjin J, Kessels A, Lame I, van Suijlekon H, van Kleef M. Pulsed radiofrequency adjacent to the cervical dorsal root ganglion in chronic cervical radicular pain: a double-blind sham controlled randomized clinical trial. Pain 2007;127:173–182. 10. Gallagher R. Pulsed radiofrequency treatment: what is the evidence of its effectiveness and should it be used in clinical practice. Pain Medicine 2006;7(5):408–410. 11. Slipman C, Isaac Z, Lenrow D, Chou L, Gilchrist R, Vresilovic E. Clinical evidence of chemical radiculopathy. Pain Physician 2002;5(3):260–265. 12. Bogduk N, Govind J. Medical Management of Acute Lumbar Radicular Pain: An Evidence-Based Approach. Newcastle: Cambridge Press, 1999:21–32, 41–42. 13. Bogduk N, ed. Practice Guidelines for Spinal Diagnostic and Treatment Procedures. International Spine Intervention Society, 2004. 14. Weber H, Holme I, Amlie E. The natural course of acute sciatica with nerve root symptoms in a double-blind placebo-controlled trial evaluating the effect of piroxicam. Spine 1993;18:1433–1438. 15. Saal JA, Saal JS. Nonoperative treatment of herniated lumbar intervertebral disc with radiculopathy: an outcome study. Spine 1989;14:431–437. 16. Boswell MV, Jansen HC, Trescot AM, Hirsch JA. Epidural steroids in the management of chronic spinal pain and radiculopathy. Pain Physician 2003;6(3):319–334. 17. Riew D, Yin Y, Gilula L, Bridwell K, Lenke L, Lauryssen C, Goette K. The effect of nerve-root injection on the need for operative treatment of lumbar radicular pain. A prospective, randomized, controlled, double-blind study. J Bone Joint Surg 2000;82-A:1589–1593. 18. Vad VB, Bhat AL, Lutz GE, Cannisa F. Transforaminal epidural steroid injection in lumbosacral radiculopathy: a prospective randomized study. Spine 2002;27:11–16. 19. Lutz G, Vad V, Wisneski R. Fluoroscopic transforaminal lumbar epidural steroids: an outcome study. Arch Phys Med Rehabil 1998;79:1362–1366. 20. Rozin L, Rozin R, Koehler SA, Shakir A, Ladham S, Barmada M, Dominick J, Wecht CD. Death during transforaminal epidural steroid nerve root block (C7) due to perforation of the left vertebral artery. Am J Forensic Med Pathol 2003;24(4):351–5. 21. Tiso RL, Cutler T, Catania JA, Whalen K. Adverse central nervous system sequelae after selective transforaminal block: the role of corticosteroids. Spine J 2004;4(4):468–474. 22. Houten J, Errico T. Paraplegia after lumbosacral nerve root block: report of three cases. Spine J 2002;2(1):70–75. 23. Young WF. Transient blindness after lumbar epidural steroid injection: a case report and literature review. Spine 2002;27:E476–477. 24. Furman MB, O’Brien EM, Zgleszewski TM. Incidence of intravascular penetration in transforaminal lumbosacral epidural steroid injection. Spine 2000;25(20):2628–32. 25. Furman MB, Giovanniello MT, O’Brien EM. Incidence of intravascular penetration in transforaminal cervical epidural steroid injections. Spine 2003;28 (1):21–5. 26. Karasek M, Bogduk. Temporary neurologic deficit after cervical transforaminal injection of local anesthetic. Pain Med 2004;5(2):202–205. 27. Teixeira A, Grandinson M, Sluijter M. Pulsed radiofrequency for radicular pain due to a herniated intervertebral disc-an initial report. Pain Pract 2005;5(2):111–115. 28. Raj P, Lou L, Erdine S, Staats P, Waldman S, Racz G, Hammer M, Niv D, Ruiz-Lopez R, Heavner J. Thermal and pulsed radiofrequency. In Raj P,
203
204
Chapter 9 Pulsed Radiofrequency Procedures in Clinical Practice Lou L, Staats P (eds.): Interventional Pain Management, 2nd Edition. Philadelphia: WB Saunders, 2008. 29. Sluijter ME. Radiofrequency, Part 1. Meggen, Switzerland: Flivopress, 2001. 30. Van Zundert J, Lamé IE, de Louw A, Jansen J, Kessels F, Patijn J, van Kleef M. Percutaneous pulsed radiofrequency treatment of the cervical dorsal root ganglion in the treatment of chronic cervical pain syndromes: a clinical audit. Neuromodulation 2003;6(1):6–14. 31. Sluijter M. Radiofrequency ablation in the management of spinal pain. Controversies and Consensus in Imaging and Intervention (serial online). 2006;4(1). 32. Gaucci CA. Manual of RF Techniques. Meggen, Switzerland: Flivopress, 2004. 33. Higuchi Y, Nashold BS Jr, Sluijter M, Cossman E, Pearlstein RD. Exposure of the dorsal root ganglion in rats to pulsed radiofrequency currents activates dorsal horn lamina I and II neurons. Neurosurgery 2002;50(4):850–5, discussion 856. 34. Higuchi Y, Nashold BS Jr, Sluijter M, Cossman E, Pearlstein RD. Exposure of the dorsal root ganglion in rats to pulsed radiofrequency currents activates dorsal horn lamina I and II neurons. Neurosurgery 2002;50:850–856. 35. Van Zundert J, de Louw AJA, Joosten EAJ, Kessels AGH, Honing W, Dederen J, Veening JG, Vles JSH, van Kleef M. Pulsed and continuous radiofrequency current adjacent to the cervical dorsal root ganglion of the rat induces late cellular activity in the dorsal horn. Anesthesiology 2005;102:125–131. 36. Podhajshy RJ, Sekiguchi Y, Kikuchi S, Myers RR. The histologic effects of pulsed and continuous radiofrequency lesions at 42 degrees C to rat dorsal root ganglion and sciatic nerve. Spine 2005;30(9):1008–1013. 37. Edrine S, Yucel A, Cimen A, Aydin A, Sav A, Brlir A. Effects of pulsed versus conventional radiofrequency current on rabbit dorsal root ganglion morphology. Eur J Pain 2005;9(3):251–256. 38. Slappendel R, Crul BJ, Braak GJ, Geurts JW, Booij LH, Voerman VF, de Boo T. The efficacy of radiofrequency lesioning of the cervical spinal dorsal root ganglion in a double-blinded randomized study: no difference between 40 degrees C and 67 degrees C treatments. Pain 1997;73:159–163. 39. Raj P, Lou L, Erdine S, Staats P, Waldman S. Lumbar sleeve and dorsal root ganglion block. In: Radiographic Imaging for Regional Anesthesia and Pain Management. 1st ed. New York: Churchill Livingstone, 2003;26:153–157. 40. Cramer G, Darby S. Basic and Clinical Anatomy of the Spine, Spinal Cord and ANS. St. Louis, MO: Mosby, 1995:65. 41. Kikuchi S, Sato K, Konno S, Hasue M. Anatomic and radiographic study of dorsal root ganglion. Spine 1994;19(1):6–11. 42. Abejon D, Garcia-del-Valle S, Fuentes M, Gomez-Arnau J, Reig E, van Zundert J. Pulsed radiofrequency in lumbar radicular pain: clinical effects in various etiological groups. Pain Pract 2007;7(1):21–26. 43. Ahadian FM. Pulsed radiofrequency neurotomy: advances in pain medicine. Curr Pain Headache Rep 2004;8:34–40. 44. Rozen D, Parvez U. Pulsed radiofrequency of lumbar nerve roots for treatment of chronic inguinal herniorrhaphy pain. Pain Physician 2006;9:153–156. 45. Waldman S. Atlas of Interventional Pain Management, 2nd ed. Philadelphia: WB Saunders, 2004;42:163–165. 46. Cohen SP, Sireci A, Wu CL, Larkin TM, Williams KA, Hurley RW. Pulsed radiofrequency of the dorsal root ganglia is superior to pharmacotherapy or pulsed radiofrequency of the intercostal nerves in the treatment of chronic postsurgical thoracic pain. Pain Physician 2006;9:227–236.
Summary 47. Shah R, Racz G. Pulsed mode radiofrequency lesioning of the suprascapular nerve for the treatment of chronic shoulder pain; a case report. Pain Physician 2003;6:503–506. 48. Waldman S. Pain Management. Elsevier Publishing, 2007:742–745.205
205
10 Facet Joint Injections and Sacroiliac Joint Injections Louis J. Raso
Introduction Back pain and neck pain are the most common cause of chronic pain and disability. It is a complex, often multifactorial condition affecting millions of persons worldwide. Disability from spinal pain is associated with a nonspecific diagnosis and suboptimal outcomes. It continues to pose a peculiar diagnostic challenge because of overlapping clinical features and nonspecific radiological findings.1 Although many technologic, pharmaceutical, and surgical advances for the treatment of back pain have occurred in recent years, the search for the precise cause of back pain remains a difficult process. Although radicular pain secondary to herniated disc is most commonly suspected, pain originating from facet joints is likely to be the etiology of 15–40% of nonradicular low back pain and 40–60% of nonradicular neck pain.2 Pain originating from facet joints can coexist with other causes of multifactorial back and neck pain including radicular, myofascial, sacroiliac, and intradiscal pathology. Diagnostic injection techniques are utilized to isolate the source of pain. Facet joint injections and medial branch blocks are an example of such techniques. The methods of diagnostic procedures performed to localize the source of pain vary widely. The interpretation and the relevance of these studies continue to be controversial.3
Facet Joint Injections Background The facet joint allows the spine to flex, extend, and rotate.4 It is a true synovial joint. Many conditions can affect the facet joints; however, the major cause of facet disease is osteoarthritis, a degenerative process that results in reduction or loss of facet joint cartilage, erosions of the adjacent bone margins, bony overgrowth of the facets and From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_10, © Springer Science + Business Media, LLC 2010
207
208
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
articular processes, and ultimately, instability of the joint itself, which may result in vertebral subluxation. Within the facet joint are sensory nerves that become irritated by this degenerative process, resulting in pain. Facet joints are well innervated by the medial branches of the posterior rami. Neuroanatomic, neurophysiologic, and biomechanical studies have shown free and encapsulated nerve endings in facet joints.5 These nerves contain substance P, calcitonin gene-related peptide, low-threshold mechanoreceptors, and nociceptors. In neighboring structures, pain fibers are also present, and this effect may cloud the clinical picture. These structures include the multifidus muscle, the nerve root, the epidural space, and the dura. During the inflammatory process, venous stasis and hyperemia produce increasing pressure on these adjacent structures, accounting for the association of radicular symptoms with facet pathology. Imaging studies, including radiography, MR imaging, and CT, are unreliable indicators of disease involving facet joints. Some patients may be asymptomatic despite exhibiting severe disease on MRI or CT.6 Conversely, some patients may complain of severe pain with relatively minor abnormalities on imaging studies. Images may demonstrate joint pathology, including osteophytic spurring, fluid in the joint capsule, or a synovial cyst. MRI may reveal actual joint space narrowing or hypertrophy, and a bone scan may reveal abnormalities of bone turnover.7 Anatomy The facet joint is a true sinuarthrodial joint and is formed by the union of adjacent articular processes. It lies at the posterior aspect of the spinal column (Figure 10.1). Each lumbar facet joint is supplied by a branch of the posterior primary ramus from the nerve root at the corresponding level, and a branch from the nerve root one level above (Figure 10.2). As an example, the L4–5 facet receives nerve supply from the L4 medial branch and the L5 medial branch. The cervical facet joint
Figure 10.1. Sagittal diagram depicting the components of the lumbar facet joint. The joint is a true encapsulated synovial joint formed from the articulation of adjacent inferior and superior articular processes (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Facet Joint Injections
Figure 10.2. Axial anatomical diagram of the lumbar facet joint. The articular surfaces of the superior and inferior articular processes are capped with hyaline cartilage. The fibrous joint capsule contains a synovial membrane and a small amount of synovial fluid. The joint is bordered by the ligamentum flavum anteriorly and by the multifidus muscle posteriorly (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
receives its innervation from the same level, and from one level above and below.2 Lumbar Facet Joints The orientation of the facet joints varies considerably, depending on the level of the spinal column. At the cervical level, they are oriented in an oblique coronal plane, angled superior to inferior in a posterior direction. The thoracic joints are nearly vertical and coronal in orientation, rotating toward the sagittal plane at the thoracolumbar junction. The superior lumbar facets are oriented in a nearly sagittal plane, which rotates outward toward the coronal plane as one descends the lumbar spine. At the lumbosacral junction, the joints lie in a sagittal, coronal, and oblique plane.8 The lumbar facet joints have a “C” shaped configuration on various imaging views with the concavity of the “C” arc facing inward. It is important to access the posterior portion of the joint, and it is most easily entered with a needle using a shallow angle of 10–20° from straight anteroposterior (AP).9 The posterior portion of the facet joint is enveloped by a thick, tough, fibrous capsule. Hyaline cartilage lines the articular surfaces of the superior and inferior articular processes. The joint capsule is redundant at the superior and inferior margins of the joint (superior and inferior recesses). It is the inferior portion of this posterior recess that is accessible
209
210
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
Figure 10.3. Pathophysiologic aspects of facet joint pain (Netter Anatomy Illustration Collection, © Elsevier, Inc. All Rights Reserved).
to percutaneous needle puncture. The optimal target zone for needle placement is the inferior portion of the posterior joint recess just inferolateral to the inferior articular process (Figure 10.3).10 Cervical Facet Joints The superior and inferior articular facets are joined by articular pillars of bone. They correspond to the pars inter-articularis of the lumbar spine. The joint is oriented in a plane 45° from the coronal plane and
Facet Joint Injections
obliquely in the craniocaudal direction. Subsequently, only a portion of the joint is usually visualized.11 Joint recesses lie at the lateral margin of the superior and inferior articular facets and are lined with synovium. The lateral recesses of the cervical facet are the target for injection techniques.12 Innervation of the facet joints of the cervical spine varies from the upper to the lower cervical levels. The C4–7 levels are similar to the lumbar and thoracic regions, and the posterior primary rami arise from the dorsal root ganglia lateral to the intervertebral foramen. The medial branch then arises from ramus and passes around the ipsilateral articular pillar. Each medial branch then sends out an ascending and descending branch innervating the facet joint above and below, respectively.13 For example, the C5–6 facet receives innervation from the medial branches of C5 and C6. The numbering of the thoracic and lumbar facet innervation is different due to the C8 medial branch (Figures 10.4 and 10.5). The C3 medial branch differs from the other cervical levels. It contains a superficial and deep branch.14 The deep branch courses
Figure 10.4. The intersection of the diagonals of the articular pillar indicates the target point for medial branch blocks at typical cervical levels (Provided by Newcastle Pain Management and Research Group. Reprinted with permission from Bogduk N. Cervical Medial Branch Block. In Bogduk (ed): Practice Guidelines for Spinal Diagnostic and Treatment Procedures. San Francisco: International Spine Intervention Society, 2004).
211
212
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
Figure 10.5. A lateral view of the cervical spine. The dot shown by the arrow indicates the target point for blocks of the C7 medial branch (Provided by Newcastle Pain Management and Research Group. Reprinted with permission from Bogduk N. Cervical Medial Branch Blocks. In Bogduk N (ed): Practice Guidelines for Spinal Diagnostic and Treatment Procedures. San Francisco: International Spine Intervention Society, 2004).
around the articular pillar of C3 similar to the C4–7 levels. This branch then innervates the C3–4 facet joint but provides only negligible supply to the C2–3 facet joint. The C2–3 facet joint is primarily supplied from a large superficial medial branch of C3. This nerve is known as the “third occipital nerve.” It courses around the lateral aspect of the C2–3 facet and then supplies cutaneous innervation to the inferior portion of the occipital region. Small communicating branches of the third occipital nerve also supply innervation to the C2–3 facet joint. Radiofrequency heat ablation can be used to treat various sources of cervical pain, including headaches that originate from the Third Occipital Nerve (TON) or from the C2 branch. By using a 10-mm exposed RF needle, a temperature of 70° for 40 s is used to denervate the median branch nerve contributing to the headache. Each level of the cervical spine can be treated with heat radiofrequency lesioning. Correct placement of the needle is performed by placing the needle tip on the articular pillar. For medial branch RF lesioning of C3–6, the target point is the centroid of the articular pillar with the same segmental number as the target nerve. This centroid is found at the intersection of the two diagonals of the diamond-shaped pillar.
Facet Joint Injections
For medial branch RF lesioning at C7, the target point lies high on the apex of the superior articular process of C7. This is because the base of the C7 transverse process occupies most of the lateral aspect of the C7 articular pillar and thrusts the medial branch relatively higher than the typical cervical medial branches. Because of the consistent variability of the nerves path, one should do multiple lesioning at the site. Thoracic Facet Joints The thoracic facet joints are oriented in a coronal plane and support the spine from shearing forces. They are difficult to access from a pure posterior approach secondary to the lamina. When attempting to perform a percutaneous procedure, the lungs prevent a lateral approach to the facets. Utilizing a typical posterolateral approach, direct intra-articular access is difficult. The innervation of the facets and numbering of the joints are similar to the lumbar spine.3 Diagnostic Criteria Lumbar Facet Pain • Paravertebral low back pain, which is often aggravated by remaining in any posture • Pain worsened by twisting or rotation • Pain increased on extension may be relieved by flexion • Dull pain limited to the low back, buttock, and hip can extend to thigh and knee in a nondermatomal distribution • Pain rarely extends below the knee • Pain exacerbated by moving from a sitting to a standing position • Pain relieved by standing, walking, rest, or repeated activity • Morning stiffness • A normal neurologic examination • Tenderness to palpation of the affected joint • Absence of radicular pain and straight leg raising.15 Cervical Facet Pain • • • • • •
Paravertebral neck pain Local tenderness over the affected facet joints Pain referred to the shoulder, possibly extending to the elbow Pain following a nondermatomal pattern Decreased range of motion of the cervical spine The upper cervical levels can not only cause neck pain but also cervicogenic headaches.13
Injection Techniques Because imaging is not a reliable indicator of pain resulting from facet disease, clinical information and a careful history and physical examination are most helpful in arriving at a diagnosis. Once the diagnosis of facet syndrome is suspected, a diagnostic or therapeutic block should be performed.15 A facet block is considered positive when a patient experiences 80% pain reduction. It is of the utmost importance to discover whether the patient’s typical pain was relieved by the injection.
213
214
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
Medial branch blocks are most frequently requested as a diagnostic tool prior to proceeding with a medial branch rhizotomy. One can evaluate lumbar facets by blocking the nerves innervating them or by intra-articular injection. Three techniques for facet joint blocks exist and have been described in the literature. They include intra-articular injection, periarticular injection, and medial branch block.2 Periarticular injections are the most common technique used in pain management. Intra-articular injections are also used in chronic pain management and in specific indications such as drainage of a synovial cyst. As stated previously, medial branch blocks are frequently used as a diagnostic tool and are considered more accurate than intra-articular joint injection for prognosis and outcome (Figures 10.6 and 10.7). Contraindications for facet joint injections are similar to other minimally invasive targeted spinal injections and include patient refusal, coagulopathy, systemic infection, or a localized infection at the site of injection, and allergy to the medications to be used in the injection.16 Complications, risks, and side effects also need to be discussed with the patient. Documentation of the discussion and a procedure-specific consent needs to be signed. Potential risks include allergic reaction, early postprocedural increase in pain, bleeding, and a transient increase in blood glucose levels.
Figure 10.6. (A) Sacroiliac joint needle placement; most inferior; intraarticular.
Facet Joint Injections
(B) Second needle placed at the middle section of lesioning; periarticular. (C) Third needle placement at the most superior portion of the joint, median branch.
215
216
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
Figure 10.7. Oblique radiograph of the lumbar spine demonstrating a typical lumbar facet joint arthrogram. A small volume of contrast material can be seen between the superior and inferior articular processes, and extending into the capsular recesses (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Radiofrequency ablation can be applied to the affected median branch nerve of the thoracic spine with good results and outcomes. Placement of an RF needle will follow the placement of the median branch block. Some confusion does arise with the nomenclature of each nerve and how it innervates each joint. Two types of articular branches arise from each median branch. Ascending branches arise where the medial branch passes caudal and to the zygapophysial joint. Descending branches arise where the medial branch crosses the transverse process and passes to the next joint. When medial branches subsequently cross a transverse process, it will be the transverse process of the vertebra with the next segmental number. Lumbar Intra-articular Injection There are occasions when intra-articular injections are the preferred method. They include when patients cannot undergo rhizotomy because of an indwelling cardiac pacer, when patients have undergone
Facet Joint Injections
rhizotomy with inadequate pain relief, and for rupture and drainage of a synovial cyst. Levels to be injected are selected on the basis of physical symptoms and imaging studies. Referring back to the pain diagram prior to selecting the levels is essential. Method 1. The patient is placed in the prone position with increased padding under the abdomen, allowing flexion of the lumbar spine. 2. Position the C-arm in an ipsilateral oblique view to properly visualize the inferior aspect of the joint. 3. Insert the needle “down the barrel” of the x-ray beam. 4. Contact the inferior aspect of the facet joint. Adjust and advance until the needle either encounters bone or feels as if joint space has been entered. 5. Inject 0.1–0.3 mL of nonionic contrast and document an adequate arthrogram. Reposition the needle as necessary. 6. Once intra-articular position is confirmed, inject anesthetic and steroid agents as preferred. Limit intra-articular volume to 0.5–1.0 mL of local anesthetic and 0.5–1.0 mL of steroid. Larger volumes are injected if capsular rupture is desired. 7. Monitor for pain response for 30 min prior to discharge. 8. Discharge patient when stable. Have patient call the office for persistent or increased pain, numbness, or fever. Lumbar Periarticular Injection In periarticular injections, the initial approach is identical to intra-articular injections, but an arthrogram is not performed. The needle is advanced until bone is contacted at the level of the joint capsule. The injection is then performed after negative aspiration. A larger volume of injectate is used for this technique (2.0–2.5 mL). The needle may be repositioned multiple times during injection to different points along the joint capsule, and the injectate may be deposited at multiple sites. It is important to keep in mind that, although one is using a larger volume, the maximum dose of steroid is not to be exceeded (80–120 mg).17 Lumbar Medial Branch Block As a purely diagnostic technique, medial branch block may be employed as a method of confirming facetogenic pain. It is typically performed as a pre-op workup to medial branch neurotomy. A positive response to a diagnostic block is reassuring in that it indicates that the patient may respond to rhizotomy with prolonged pain relief.18 The medial branch courses over the base of the transverse process, where it joins the superior articulating process. The primary target for this nerve lies just caudad to the superior margin of the medial aspect of the transverse process.8 Method (Figure 10.8) 1. The patient is positioned in the prone position with pelvic padding to reduce the lordotic curve. 2. The lumbar area is prepped and draped. 3. The C-arm is rotated to 10–20° from AP to visualize the junction of the base of the transverse process and the superior articular process in the center of the field.
217
218
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
Figure 10.8. Median branch block.
4. With this view, the “Scotty Dog” configuration, the nose represents the transverse process, the eye the pedical, and the ear the superior articulating process. 5. With the use of intermittent fluoroscopy, the needle is advanced until the tip contacts the bone where the base of the transverse process, the superior articulating process, and the pedicle join. 6. The needle is then slowly advanced over the superior aspect of the transverse process. At this position, the nerve is totally exposed to the needle tip. 7. A lateral view is obtained to rule out that the needle tip is not encroaching on the neural foramen. 8. Aspiration is performed to prevent vascular injection. 9. 0.3–0.5 mL of anesthetic is injected for a diagnostic block (1% Xylocaine). 10. Monitor the patient for pain response, and record percentage pain relief at 30 min. 11. Release the patient when stable. Cervical Facet Block Cervical facet blocks are also used to confirm facetogenic pain.13 When therapeutic effects also are intended, a steroid compound may be added to the anesthetic mixture. The two main approaches to the cervical facet joint are a direct lateral approach and a posterolateral approach.14
Facet Joint Injections
Method
Lateral Approach 1. Position the patient in the lateral decubitus position, with the painful side up, with a pad under the head to keep the head parallel to the table. 2. The skin surface is prepped and draped. 3. A 25-gauge, 2.5-in. spinal needle is advanced toward the facet joint in a direct lateral approach. 4. AP and lateral views are continuously obtained to assess needle depth. Care must be taken to ensure that the needle has not passed through the joint and into the spinal canal. 5. The needle is advanced until it contacts either the superior or inferior articulating process. 6. The needle is then withdrawn a few millimeters and re-advanced into the joint. 7. The physician may choose to inject 0.1–0.3 mL of water-soluble contrast. 8. If a diagnostic block is to be performed, 0.5–1.5 mL of short or longacting anesthetic is injected (1% Xylocaine). 9. The patient is monitored for 30 min, and the percentage pain relief is recorded.
Posterolateral Approach 1. Position the patient in the prone position, and the posterior cervical spine is prepped and draped. 2. The puncture site is located along the posterolateral skin surface, 2–3 levels below the intended facet block. 3. A 22-gauge, 3.5 in. spinal needle is advanced in a caudocephalad direction under fluoroscopic guidance to enter the joint parallel to its oblique plane. 4. At this point, if desired, 0.1–0.3 mL of water-soluble contrast is injected for arthroscopic confirmation. 5. Aspiration is performed to rule out intravascular placement. 6. 0.5–1.5 mL of anesthetic plus or minus a steroidal agent is injected (1% Xylocaine). Thoracic Facet Block Thoracic facet injections are rarely performed and are usually indicated to rule out other causes of thoracic para-spinal pain. One clear indication is pain localized over a joint space. As with other levels in the spine, imaging studies rarely give a clear source of facet pathology. There may be increased bone turnover on a bone scan or abnormalities identified on a MRI scan.18 A thoracic facet block is positioned in the later part of the treatment algorithm for thoracic pain. It is a more difficult procedure to perform intra-articular than in the lumbar area due to the coronal orientation of the joints and the close proximity of the lungs. Many physicians who perform this procedure advocate using CT guidance rather than fluoroscopy to have clear visualization of the lungs, pleural space, and nervous tissue.19 Under ideal conditions, CT fluoroscopy provides the best visualization during performance of the procedure.
219
220
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
Method 1. Position the patient prone on the OR table, and a sterile prep is performed in the thoracic spine. 2. Utilizing the practitioner’s preferred imaging device, a posterolateral approach is taken. 3. The needle target site is below the inferior margin of the inferior articular facet and posterior to the superior articular facet. 4. This allows access to the inferior recess of the joint and the injectate to reach the joint space. 5. An alternative approach is to deposit the injectate extracapsular to the posterior to the inferior portion of the facet joint. Conclusion for Facet Joint Injections Treating back pain can be challenging yet rewarding for the interventional pain practitioner due to the multifactorial causes. The physician needs to rely on his/her clinical skills and assimilate the information presented. Imaging studies may provide clues as to the source but also can lead down the wrong path.20 Numerous studies have shown that facet joint pathology is a major contributor to axial back pain.21,22 With targeted injections, directed either at the joint itself or at the nerve supply to that joint, one is able to improve the quality of life of patients, allowing them to function at a diminished level of pain.23,24
Sacroiliac Joint Injections Background Pain emanating from the sacroiliac (SI) joint commonly occurs after the patient stands up while in an awkward position, thereby putting strain on the supporting ligaments, soft tissue, and the joint itself.25 It has been a controversial potential cause of low back pain. A study conducted by Bernard and Kirkaldy-Willis concluded that the SI joint was the source of pain in 22.5% of 1,293 patients.25,26 Anatomy The sacrum is wedge-shaped and is formed by the fusion of five sacral vertebrae. It transmits forces from the pelvis and lower extremities to the vertebral column, and also supports the lumbar spine. The sacrum articulates superiorly with the lumbar spine and inferiorly with the coccyx. It allows a few millimeters of glide and 2–3° of rotation, and it is a joint that functions as a structure of stability.27 The sacrum is a C-shaped, diarthrodial true synovial joint. It contains synovial fluid within its joint cavity. The adjacent structures are joined by ligaments. The joint has a fibrous capsule and an inner synovial lining. The sacroiliac joint is stabilized by multiple major ligaments. They include the iliolumbar, interosseous, anterior and superior sacroiliac, sacrospinous, and sacrotuberous. These ligaments connect the spine, sacrum, iliac bones, and pubic symphysis. They connect with multiple muscles located in the
Sacroiliac Joint Injections
Figure 10.9. (A, B) Sacrum and coccyx (Reprinted with permission from Moore KL, Dalley AF. Clinically Oriented Anatomy, 5th ed. Philadelphia: Lippincott Williams & Wilkins, 2006).
vicinity.27,28 Posteriorly, the interosseous ligament forms the corresponding border of the joint due to lack of a posterior capsule. There are two different types of cartilage present: a thin layer of hyaline on the sacral side and fibrocartilage on the iliac side (Figures 10.9–10.11). Innervation Posteriorly, the surrounding tissues are supplied by the lateral branches of the posterior primary rami from L4 to S3. Anteriorly, the joint is innervated by the lateral branches of the posterior primary rami from L2 to S2. It has been noted that there is a large variation in the patterns of referred pain from individual to individual. This may be due to the fact that the innervation originates from multiple sources, with a wide
221
222
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
Figure 10.10. An illustrative anatomy of the sacroiliac joint. (A) Posterior view of an opened right sacroiliac joint demonstrating some of the important bony and soft tissue components. (B) Horizontal sections through A, the upper, B, middle, and C, lower portions of the sacroiliac joint. Diagram also demonstrates the synovial versus fibrous portions of the sacroiliac joint (Reprinted with permission from Cramer G, Darby S. Basic and Clinical Anatomy of the Spine, Spinal Cord and ANS, 2nd ed. Philadelphia: Mosby, 2005. Copyright Elsevier).
individual variation in nerve supply. This fact affects the ability of the clinician to accurately diagnose the SI joint as the pain generator.29,30 Presentation The precise diagnosis is often difficult to make since it is often accompanied by herniated discs, spinal stenosis, facet arthropathy, and a multitude of other abnormalities.31,32 The pain is not limited to the low back and buttocks, and may be referred to the groin, thigh, foot, and even upper lumbar area. Different pathologic areas in the joint may refer pain to a variety of areas due to the variable innervation of the joint. It is therefore often a diagnosis of exclusion.33 Diagnosis Numerous physical examination maneuvers have been associated with a positive diagnosis of sacroiliac joint pathology and dysfunction.34 They include: 1. Patrick’s test 2. Gaenslen’s test 3. Lateral compression test
Sacroiliac Joint Injections
223
Figure 10.11 Ligaments of the SI joint. (a) Anterior view. (b) Posterior view. (c) SI joint in horizontal section (Reprinted with permission from Cramer G, Darby S. Basic and Clinical Anatomy of the Spine, Spinal Cord and ANS, 2nd ed. Philadelphia: Mosby, 2005. Copyright Elsevier).
4. Anteroposterior pelvic compression test 5. Passive straight leg raising 6. SI joint compression test. Computed tomography (CT) may demonstrate joint space widening or narrowing, articular degeneration, osteophytes, subchondral sclerosis, erosions of the cortical surfaces, and ankylosis. Unfortunately, pain may not correlate with these findings. These abnormalities do not correlate to the patient’s response to blockade.35,36 Elgafy and colleagues evaluated the CT scans of 62 patients with SI joint pain who responded to SI joint injection and compared these with the CT scans of 50 asymptomatic age-matched controls. At least one CT finding suggestive of SI joint pathology (osteophytes, joint space narrowing <2 mm, subchondral sclerosis, joint erosions, or ankylosis) was seen in 57.5% of symptomatic patients and 31% of controls. In contrast, CT findings were negative in 42.5% of symptomatic patients (Figures 10.12 and 10.13).37 Bone scan findings also have been determined by some authors to correlate poorly with SI joint symptoms. Slipman et al.38 demonstrated poor
224
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
Figure 10.12. (A) Coronal reconstructed CT image of the sacroiliac join. The upper portion of the joint is a synarthrosis, while the inferior third is a true synovial joint. (B) Axial CT image of the sacroiliac joint, demonstrating orientation of the joint along a posteromedial-to-anterolateral plane (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
sensitivity (12.9%) of positive bone scan findings in patients responding to SI joint injection. In comparison with conventional bone scans, SPECT bone scans permit better differentiation of radiotracer uptake in the ventral synovial portion of the joint, suggestive of inflammatory causes of sacroiliitis, from uptake in the dorsal syndesmotic portion of the joint, more typical for bony changes due to axial loading.39,40 Magnetic Resonance Imaging (MRI) allows for detailed evaluation of the SI joint and adjacent soft tissues, and is particularly valuable in detecting early changes in the joint in inflammatory and infectious sacroiliitis.36,39–41 Typically, MRI demonstrates focal hyperintensity
Sacroiliac Joint Injections
Figure 10.13. Radiofrequency ablation of the SI joint.
225
226
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections
in periarticular bone on T2-weighted and STIR sequences. Bollow et al.42,43 found evidence of early periarticular erosions and contrast enhancements of the joint capsule in MRI imaging of 72% of patients with seronegative spondyloarthropathy and early sacroiliitis, but essentially no enhancement in control patients with mechanical causes of low back pain. Injection Technique The injections are usually performed in a procedure room under fluoroscopic guidance. Alternatively, CT guidance can be utilized. The patient is placed in the prone position, and wide sterile preparation of the soft tissues over the sacrum and buttocks is performed. Utilizing C-arm fluoroscopy, the x-ray beam is angled medial to lateral and is rotated until the anterior and posterior projections of the inferior aspect of the joint are superimposed on each other. For fixed fluoroscopy, the patient is positioned in the prone oblique position. The inferior third of the joint is identified, and the skin is marked to identify the entry point for the planned needle trajectory. The overlying skin and soft tissues are infiltrated with lidocaine. A 22-gauge spinal needle is then directed along the axis of the x-ray tube and advanced into the joint. For fluoroscopic imaging, injection of 0.2–0.5 mL of contrast material can be used to confirm position. Alternatively, the C-arm or patient can be rotated to confirm position; if the needle tip is placed correctly, it should remain within the joint. In general, injection of contrast and medication will be difficult if good positioning within the joint has not been obtained. The SI joint can accommodate only a small volume, ~3 mL. A mixture of 1 mL of 0.5% bupivicaine plus 40 mg of methylprednisolone acetate (DepoMedrol®, Pharmacia and Upjohn, Kalamazoo, MI) is injected with a 3 mL syringe. Conclusion for Sacroiliac Joint Injections Diagnosis of SI joint dysfunction and pain is one of exclusion. Injections are performed initially as a diagnostic tool and may be followed by a therapeutic injection when combined with an anti-inflammatory compound. The injection requires imaging in order to properly perform the procedure. The interventional practitioner may elect to use fluoroscopy or CT depending on availability and personal preference. The therapeutic injections provide pain relief of a varying duration. If the diagnostic procedure yields a favorable outcome, then the practitioner can elect to continue the patient on conservative therapy or can elect to perform a radiofrequency ablation procedure. The technique can be performed along the medial border of the joint line (Figures 10.9–10.11) or from the emitting S 1-2-3 foramen. If the medial joint line technique is elected, then, using a 15-mm exposed SMK needle and starting at the most inferior border and using 70°C and for 40 s, the first of three lesions is performed. Next, “walk” the needle superiorly along the joint line and perform a second lesion. For the third and final lesion, move the needle superior again and repeat the lesion at the same temperature and time.
Sacroiliac Joint Injections
References 1. Hirsch D, Inglemark B, Miller M. The anatomical basis for low back pain. Acta Orthoscan 1963;33:1–17. 2. Fenton D, Czervionke L. Facet joint injection and medial branch block. In Fenton D, Czervioke L (eds): Image-Guided Spine Intervention. Philadelphia: WB Saunders, 2003:9–50. 3. Giles LGF. Zygapophysial (facet) joints. In Giles LGF, Singer KP (eds): Clinical Anatomy and Management of Low Back Pain. Oxford: ButterworthHeinemann, 1997. 4. Adams MA, Hutton WC. The mechanical function of the lumbar apophyseal joints. Spine 1983;8:327–330. 5. Ashton IK, Ashton BA, Gibson SJ, Polak JM, Jaffray DC, Eisenstein SM. Morphological basis for back pain: The demonstration of nerve fibers and neuropeptides in the lumbar facet joint but not in the ligamentum flavum. J Orthop Res 1992;10:72–78. 6. Schwarzer AC, Wang SC, O’Driscoll D, Harrington T, Bogduk N, Laurent R. The ability of computed tomography to identify a painful zygapophysial joint in patients with chronic low back pain. Spine 1995;20:907–912. 7. Apostolaki E, Davies AM, Evans N, Cassar-Pullicino VN. MR imaging of lumbar facet joint synovial cysts. Eur Radiol 2000;10:615–623. 8. Dory MA. Arthrography of the lumbar facet joints. Radiology 1981;140:23–27. 9. Giles LGF. Innervation of spinal structures. In Giles LGF, Singer KP (eds): Clinical Anatomy and Management of Low Back Pain. Oxford: Butterworth-Heinemann, 1997. 10. Bogduk N, Long E. The anatomy of the so-called “articular nerves” and their relationship to facet denervation in the treatment of low back pain. J Neurosurg 1979;51:172–177. 11. Aprill C, Dwyer A, Bogduk N. Cervical zygapophyseal joint pain patterns: II. A clinical evaluation. Spine 1990;15:458–461. 12. Bogduk N, Lord S. Cervical zygapophysial joint pain. Neurosurg Q 1998;8:107–117. 13. Bogduk N, Marsland A. The cervical zygapophysial joints as a source of neck pain. Spine 1988;13:610–617. 14. Barnsley L, Lord S, Bogduk N. Comparative local anesthetic blocks in the diagnosis of cervical zygapophysial joint pain. Pain 1993;55:99–106. 15. Mooney V, Robertson J. The facet syndrome. Clin Orthop Relat Res 1976;140:149–156. 16. Manchikanti L, Pampati B, Fellows B, Bakhit CE. The diagnostic validity and therapeutic value of lumbar facet joint nerve blocks with or without adjuvant agents. Curr Rev Pain 2000;4(5):337–344. 17. Kaplan M, Dreyfuss P, Halbrook B, Bogduk N. The ability of lumbar medial branch blocks to anesthetize the zygapophyseal joint: a physiologic challenge. Spine 1998;23:1847–1852. 18. Manchikanti L. Facet joint pain and the role of neural blockade in its management. Curr Rev Pain 1999;3(5):348–358. 19. Renfrew D. Facet Joint Procedures. In Renfrew D (ed): Atlas of Spine Injection. Philadelphia: WB Saunders, 2004:73–101. 20. Helbig T, Lee CK. The Lumbar facet syndrome. Spine 1988;13:686–689. 21. Raymond J, Dumas JM. Intraarticular facet block: diagnostic test or therapeutic procedure? Radiology 1984;151:333–336. 22. Bogduk N, Long DM. Percutaneous lumbar medial branch neurotomy: a modification of facet denervation. Spine 1980;5:193–200. 23. Eisenstein SM, Parry CR. The lumbar facet arthrosis syndrome: Clinical presentation and articular surface changes. J Bone Joint Surg Br 1987;69:3–7.
227
228
Chapter 10 Facet Joint Injections and Sacroiliac Joint Injections 24. Lynch MC, Taylor JF. Facet joint injection for low back pain: a clinical study. J Bone Joint Surg Br 1986;68:138–141. 25. Bernard TN Jr, Kirkaldy-Willis WH. Recognizing specific characteristics of nonspecific low back pain. Clin Orthop Relat Res 1987;217:266–280. 26. Schwarzer AC, Aprill CN, Bogduk N. The sacroiliac joint in chronic low back pain. Spine 1995;20:31–37. 27. Bernard Jr TN, Cassidy JD. The sacroiliac joint syndrome: pathophysiology, diagnosis, and management. In Frymoyer JW (ed): The Adult Spine: Principles and Practice, 2nd ed. New York: Raven Press, 1997:2343–2366. 28. Williams PL, Warwick R. The sacroiliac joint. In Williams PL, Warwick R (eds): Gray’s Anatomy, 36th ed. Edinburgh: Churchill-Livingstone, 1980:473–475. 29. Grob KR, Neuhuber WL, Kissling RO. Die innervation des Sacroiliacalgelenkes beim Menschen (Innervation of the sacroiliac joint of the human). Z Rheumatol 1995;54:117–122. 30. Sakamoto N, Yamashita T, Takebayashi T, Sekine M, Ishii S. An electrophysiologic study of mechanoreceptors in the sacroiliac joint and adjacent tissues. Spine 2001;26:E468–E471. 31. DonTigny RL. Anterior dysfunction of the sacroiliac joint as a major factor in the etiology of idiopathic low back pain syndrome. Phys Ther 1990;70:250–265. 32. Fortin JD, Dwyer AP, West S, Pier J. Sacroiliac joint: pain referral maps upon applying a new injection/arthography technique. I: asymptomatic volunteers. Spine 1994;19:1475–1482. 33. Fortin JD, Aprill CN, Ponthieux B, Pier J. Sacroiliac joint: pain referral maps upon applying a new injection/arthrography technique. II. Clinical evaluation. Spine 1994;19:1483–1489. 34. Dreyfuss P, Michaelsen M, Pauza K, McLarty J, Bogduk N. The value of medical history and physical examination in diagnosing sacroiliac joint pain. Spine 1996;21:2594–2602. 35. Slipman CW, Jackson HB, Lipetz JS, Chan KT, Lenrow D, Vresilovic EJ. Sacroiliac joint pain referral zones. Arch Phys Med Rehabil 2000;81:334–338. 36. Solonen KA. The sacroiliac joint in light of anatomical, roentgenological, and clinical studies. Acta Orthop Scand Suppl 1957;27:1–127. 37. Elgafy H, Semaan HB, Ebraheim NA, Coombs RJ. Computed tomography findings in patients with sacroiliac pain. Clin Orthop 2001;382:112–118. 38. Slipman CW, Sterenfeld EB, Chou LH, Herzog R, Vresilovic E. The value of radionuclide imaging in the diagnosis of sacroiliac joint syndrome. Spine 1996;21:2251–2254. 39. Hanly JG, Barnes DC, Mitchell MJ, MacMillan L, Docherty P. Single photon emission computed tomography in the diagnosis of inflammatory spondyloarthropathies. J Rheumatol 1993;20:2062–2068. 40. Hanly JG, Mitchell MJ, Barnes DC, MacMillan L. Early recognition of sacroiliitis by magnetic resonance imaging and single photon emission computed tomography. J Rheumatol 1994;21:2088–2095. 41. Docherty P, Mitchell MJ, MacMillan L, Mosher D, Barnes DC, Hanly JG. Magnetic resonance imaging in the detection of sacroiliitis. J Rheumatol 1992;19:393–401. 42. Bollow M, Braun J, Hamm B, Eggens U, Schilling A, Konig H, Wolf KJ. Early sacroiliitis in patients with spondyloarthropathy: evaluation with dynamic gadolinium-enhanced MR imaging. Radiology 1995;194:529–536. 43. Bollow M, Braun J, Taupitz M, Haberle J, Reibhauer BH, Paris S, Mutze S, Seyrekbasan F, Wolf KJ, Hamm B. CT-guided intraarticular corticosteroid injection into the sacroiliac joints in patients with spondyloarthropathy: indication and follow-up with contrast-enhanced MRI. J Comput Assist Tomogr 1996;20:512–521.
11 Autonomic Nerve Blockade Stanley Golovac and John M. Mathis
Introduction Sensory nerves from deep visceral and somatic organs travel with sympathetic nerves of the autonomic nervous system (also see Chap. 1 for more details about the autonomic nerve anatomy). The ability to block sympathetic nerves at key points can help to reduce pain of deep somatic and visceral origin. In addition, some of these sensory inputs along the sympathetic pathways may establish reflex arcs capable of sending impulses back to deep visceral and somatic organs. These reflex arcs can exacerbate pain on aggravation or activation by pain fiber input. Blocking certain key relay centers along the sympathetic nervous system can break such painful reflex arcs, resulting in relief from deep visceral and somatic pain cycles. Autonomic block can be achieved as a temporary, semi-permanent, or permanent event. Typical blocks use a local anesthetic (plus or minus steroid) to produce a temporary block. A semi-permanent or permanent block is achieved with materials such as phenol or absolute alcohol (see the Technique sections in this chapter). Permanent blockade must be preceded by appropriate injection testing and is considerably higher risk than the temporary block. Radiofrequency ablation is indicated for the denervation of small or unmyelinated nerves of the facets, Dorsal Root Ganglion (DRG), periosteum, joints, discs, and sympathetic nerves. It is critical to be precise with placement of the RF needle prior to ablating the tissue. Heat radiofrequency lesioning is contraindicated for large, myelinated nerves, since the coagulation of proteins can lead to anesthesia dolorosa and neuroma formation. Pulse radiofrequency, conversely, does not apparently disrupt or injure the myelin of the nerve. Appropriately selected techniques can be used to the patient’s advantage.
From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_11, © Springer Science + Business Media, LLC 2010
229
230
Chapter 11 Autonomic Nerve Blockade
Common Sympathetic Blockades The common sympathetic blockades are: • • • • • •
Stellate: pain from face, neck, upper extremities Thoracic/splanchnic: pain from deep mediastinum Celiac: pain from upper abdomen (especially pancreas) Lumbar: pain from lower extremity Hypogastric: pain from upper pelvis Impar: pain from lower pelvis, perineum.
Stellate Ganglion Blockade The stellate ganglion is composed of the fusion between the most inferior cervical ganglion and the most superior thoracic ganglion. It is located posterior to the junction of the subclavian and vertebral arteries at the C7–T1 level, anterior to the junction point of the C7 vertebral body and its transverse process (Figure 11.1a). The stellate ganglion represents a key relay station for sympathetic nerves from the head and neck as well as from the upper extremity. Indications The following are indications for stellate ganglion blockade: • Pain from upper face and neck (e.g., herpes zoster, Ménière’s disease) • Pain from upper extremities (e.g., chronic arterial embolic disease, Raynaud’s disease, reflex sympathetic dystrophy) • Hyperhydrosis and posttraumatic shock syndromes of the upper extremity. Injection Technique Anterior-to-posterior image guidance is used in placing the tip of a thin, 25-gauge, 3.5 in. spinal needle at the junction of the C7 vertebral body and the proximal transverse process.1,2 Confirmation that the needle tip is not in a vascular structure, such as the vertebral artery, can be obtained by aspirating and injecting under real-time fluoroscopy 3–4 mL of radiographic contrast (Omnipaque 240 or equivalent). The operator should see local pooling of contrast material, never any vascular runoff (Figure 11.1b). A slow injection of 5–10 mL of 0.25% bupivacaine is used for temporary relief. For permanent neurolysis, 5–10 mL of absolute alcohol is injected slowly under general anesthesia or heavy conscious sedation; 3–6% phenol can also be used in similar volumes. Permanent neurolysis should always follow a temporary test with anesthetic. Treatment with the smaller volumes should be tried, increasing as needed for effect. An effective stellate ganglion blockade will typically produce an ipsilateral Horner’s syndrome along with ipsilateral venous engorgement of the ipsilateral upper extremity. There may also be ipsilateral paresthesia of the face and upper extremity.
Stellate Ganglion Blockade
Figure 11.1. (A) The sympathetic chain in the region of the stellate ganglion, which lies behind the adjacent arteries and in front of the longus colli muscles at the C7–T1 level on each side. (B) Stellate ganglion blockade in a supine patient: anterior–posterior view of the lower neck, with a fluoroscopically guided 25-gauge needle at the junction point of the transverse process and vertebral body of C7. Radiographic contrast material spreads along the muscle plane, but there is no evidence of a vascular spread (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
231
232
Chapter 11 Autonomic Nerve Blockade
The risks of stellate ganglion blockade include intravascular injection, particularly into the vertebral artery. This could lead to vertebral dissection or occlusion, seizure, and stroke. In addition, the phrenic nerve and recurrent laryngeal nerve are in close proximity to the stellate ganglion, so that either could be temporarily or permanently paralyzed. Bilateral stellate ganglionic block is not advised because it can result in respiratory compromise and loss of laryngeal reflexes. Hypotension and brachycardia also may occur. Contraindications to stellate ganglion blockade include contralateral pneumothorax, recent myocardial infarction (as the accelerator nerves to the heart pass through the stellate ganglion and will be affected such that any compensatory increase in cardiac output will be prevented), untreated heart block, glaucoma, and uncorrected coagulopathy. Radiofrequency-Induced Blockade Begin by proper positioning of the patient; supine is most comfortable for the patient. Next, minimal sedation can be used to allay anxiety and control the situation. With the fluoroscope, identify the C7 vertebral notch. This will allow a safer approach to the stellate ganglion. Anesthetize the skin and deeper tissue with the “no sting solution,” via a 30-gauge needle. Some clinicians will even go a step further to anesthetize down to the affected nerve area with a 1½ 25-gauge needle. This will create a track to offer less discomfort to the patient. One must avoid placing the needle inferior to the clavicle or a pneumothorax may occur.3 Position the 22-gauge RF 5-mm exposed tip needle at the stellate ganglion. Upon coming in contact, stop immediately and begin stimulation at 0.1 mA. Administer 1 cc of 2% Xylocaine to blunt any electrode discomfort resulting from heat. Test the area with sensory stimulation at 0.1 mA; impedance should be between 250 and 400 W. Begin with a “pulse” setting at 42°C for 120 s, then change over to heat mode at 70°C for an additional 20 s. This will allow the nerves to be initially stunned with the pulse mode and then treated with heat ablation. Pulse radiofrequency utilizes brief episodes of high-voltage, highfrequency (~300 kHz) electrical current to provide the same amount of voltage without tissue heating to the degree of tissue coagulation. One needs to remember that, with the pulse mode, energy only is produced at the “tip” of the RF needle. Therefore, the needle will need to be perpendicular to the target or imbedded in the nerve. With “Heat,” the lesion is formed as an ellipse around the tip. After completion of the denervation, some pain may be experienced at the site for 24–48 h. Side Effects and Complications Local anesthetic toxicity, bleeding, pneumothorax, neural trauma, and cardiovascular compromise can occur if strict attention is not adhered to. Destruction of adjacent structures, adjacent organs, and vital structures, hypotension, neuralgia, dysesthesia, spinal cord injury, pneumothorax, and hematoma may result.
Thoracic and Splanchnic Sympathetic Blockades
Key Points l
l
l l
l
Radiofrequency lesioning provides a safe, reliable, and controllable neurolysis Small volume diagnostic injections are the key to accurate prediction of the radiofrequency response Radiofrequency can be used practically anywhere in the body Radiofrequency is also suited to neurolysis of small unmyelinated nerves such as facet joints, sympathetic ganglia, or periosteal attachments Although pulse radiofrequency is useful, its applicability is presently limited by available reimbursement.
Thoracic and Splanchnic Sympathetic Blockades The thoracic sympathetics run vertically along the anterior lateral aspect of the vertebral bodies from T2 to T8 and supply the middle and upper deep mediastinal structures (Figure 11.2a). The splanchnic sympathetics arise from T11 to T12 and give sympathetic supply to the lower mediastinum.4,5 Indications The indications for thoracic or splanchnic sympathetic blockade include pain from deep mediastinal structures (e.g., locally invasive esophageal cancer, lung cancer). Injection Technique The technique for thoracic or splanchnic sympathetic blockade involves placing a needle (22- or 25-gauge) adjacent to the thoracic vertebral bodies just deep enough to the pleural surface so that the tip will lie along the lateral aspect of the vertebrae at the level to be treated.2,6,7 The actual location of the thoracic ganglion may vary from the anterolateral vertebral margin to 15–20 mm behind the anterior vertebral margin.7 Usually, needle positioning is accomplished from a posterior oblique approach by means of computed tomographic (CT) guidance. Injecting small amounts of saline while passing the needle along an extrapleural course may help to avoid pneumothorax by expanding the extrapleural space (Figure 11.2b). An injection of 7–10 mL of 0.25% bupivacaine can be administered for temporary relief. After appropriate temporary testing, permanent neurolysis can be achieved by using 5–10 mL of absolute alcohol. Again, the dose should be the minimum one that will produce the desired effect. The risks of thoracic sympathetic blockade include pneumothorax, bleeding, and intravascular injection. The contraindications to thoracic sympathetic blockade are uncorrected coagulopathy and contralateral pneumothorax, and a relative contraindication is allergy to any of the medications that might be administered.
233
234
Chapter 11 Autonomic Nerve Blockade
Figure 11.2. (A) The upper thoracic spine region, showing the sympathetic ganglia along the lateral aspect of the vertebral bodies. (B) Thoracic sympathetic blockade in a prone patient. Under computed tomography, the needle is guided from posterior to anterior obliquely (small arrows) along the lateral aspect of the vertebral body. The needle tip (large arrow) should lie along the anterior-lateral aspect of the vertebral body: thoracic sympathetic block, T2–T3; splanchnic sympathetic block, T11–T12 (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Celiac Plexus Blockade
Radiofrequency Block Technique: Splanchnic Nerve Denervation (Retrocrural Approach) Begin by proper positioning of the patient; prone is the most comfortable for the patient. Next, minimal sedation can be used to allay anxiety and control the situation. With the fluoroscope or equivalent, identify the T12–L1 vertebral body on an AP view. Begin approximately two vertebral bodies below the targeted area of T11. Anesthetize the skin and deeper tissue with the “no sting solution,” via a 30-gauge needle. Some clinicians will even go a step further to anesthetize down to the affected nerve area with a 1½ 25-gauge needle. This will create a track to offer less discomfort to the patient. Position the 22-gauge RF 10-mm exposed tip needle adjacent to the T11 vertebral body, allowing the needle to contact the body at the middle portion of the body. Be careful not to advance any further because of the great vessels located just anterior to the needle tip and the pleura also. Test the area with sensory stimulation at 0.1 mA; impedance should be between 250 and 400 W. Administer 1 cc of 2% Xylocaine to blunt the heat portion of the radiofrequency ablation. Begin with a “pulse” setting at 42°C for 120 s, then change over to heat mode at 70°C for an additional 20 s. This will allow the nerves to be initially stunned with the pulse mode and then treated with heat ablation. Pulse radiofrequency utilizes brief episodes of high-voltage, highfrequency (~300 kHz) electrical current to provide the same amount of voltage without tissue heating to the degree of tissue coagulation. One needs to remember that with the pulse mode, energy only is produced at the “tip” of the RF needle. Therefore, the needle will need to be perpendicular to the target or imbedded in the nerve. With “Heat,” the lesion is formed as an ellipse around the tip. After completion of the denervation, some pain may be experienced at the site for 24–48 h. Side Effects and Complications Hypotension and increased bowel gastrointestinal motility occur to a greater and lesser extent in most patients following a splanchnic nerve denervation. Common complications that need to be noted include: bleeding, infection, nerve injury, pneumothorax, and spinal cord trauma. Hypotension occurs as a result of regional vasodilatation and pooling of blood within the splanchnic vessels. This side effect is more likely to occur in elderly and debilitated patients. Prophylaxis should be performed by administering 500–1,000 mL of a normal saline solution in order to prevent the occurrence. Gastrointestinal hypermotility may occur as a result of unopposed parasympathetic activity. It occasionally manifests as diarrhea, except in cancer patients, who tend to be chronically constipated from high doses of opioids.
Celiac Plexus Blockade The celiac sympathetic ganglia are located on both sides of the celiac artery anterior to the aorta and anterior to the cura of the diaphragms
235
236
Chapter 11 Autonomic Nerve Blockade
(Figure 11.3a). Celiac sympathetic nerves receive and send out impulses to upper abdominal viscera including the pancreas, spleen, liver, gallbladder, mesentery, transverse colon, and stomach. Indications Indications for celiac plexus blockade include the following: • Intractable pain from terminal pancreatic cancer • Intractable pain from chronic pancreatitis • Intractable pain from other sources of the upper abdomen including visceral arterial insufficiency. Injection Technique Celiac plexus blockade should always be performed with image guidance; typically CT is used.2,6,8 However, some operators prefer ultrasound for needle guidance. For CT guidance, one starts at approximately the T12 level to locate the celiac artery. Caudal-to-cranial tube angulation may be quite helpful to keep the needle out of the posterior inferior lung. Needles should be directed from posterior to anterior such that the tips pass very close to the adjacent T12 vertebral body and terminate on either side of the aorta while passing through the cura of the diaphragms (Figure 11.3b). In some situations, it may be necessary to pass the needle through the aorta. (A 22- or 25-gauge needle should not pose a problem as long as the patient is not coagulopathic (Figure 11.3c).) An alternative to a posterior-to-anterior approach is an anterior-toposterior approach through the left lobe of the liver (Figure 11.3d). This can be done by ultrasound or CT guidance. The needle tip should lie just anterior to the celiac artery. Often, an anterior approach requires only a single needle for adequate distribution of medication along both sides of the celiac plexus. Once the needle tip has reached the target, confirmation is achieved by injecting 3–4 mL of iodine contrast medium (Omnipaque 240 or equivalent) to confirm that the needle tips are anterior to the cura of the diaphragms and are not in a vascular structure. For therapy, 10–20 mL of 0.25% bupivacaine can be injected for temporary relief. For permanent relief, 5–10 mL of absolute alcohol (or 6% phenol) can be administered for a neurolysis (under general anesthesia). Following celiac plexus blockade, it is important to hydrate the patient generously with intravenous fluids for 24 h since vascular pooling of blood in the visceral circulation due to splanchnic vasodilation may render the patient quite hypotensive.
Figure 11.3. (A) The sympathetic chain and distribution in the lower thoracic, upper abdominal region. (B) Cross-sectional drawing at the level of T12, depicting bilateral needle placement for a celiac block via a posterolateral approach. The needle tip should be anterior to the aorta and diaphragmatic crura and at or above the celiac artery origin.
238
Chapter 11 Autonomic Nerve Blockade
Figure 11.3. Continued. (C) Celiac plexus blockade in a prone patient. Under CT guidance, the needles enter posterior to anterior, obliquely. The needle tips should lie on each side of the celiac artery (approximately T12 level). (D) Celiac plexus blockade in supine patient. Under CT guidance, the needle passes through the left lobe of the liver. The needle tip should be positioned immediately anterior to the celiac artery (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Lumbar Sympathetic Blockade
Contraindications to celiac plexus blockades include uncorrected coagulopathy, bowel obstruction, and allergy to any of the medications that might be used. Celiac plexus blockades should be avoided when there is an underlying bowel obstruction, since unopposed parasympathetic activity might lead to increased bowel motility. A common complication to celiac plexus block is backache. Vascular damage or embolization can occur with intravascular injections.
Lumbar Sympathetic Blockade The lumbar sympathetic plexus lies along the anterolateral aspect of the lumbar vertebral bodies from L2 to L5 (Figure 11.4a). To block this sympathetic chain ipsilaterally, the needle tip is placed along the anterior lateral aspect of the L2 vertebral body. Indications Indications for lumbar sympathetic plexus blockade include the following: • Reflex sympathetic dystrophy of the lower extremities • Phantom limb pain (lower extremity) • Lower extremity pain from vascular insufficiency (e.g., chronic arterial emboli, Raynaud’s disease) • Lower extremity pain from gangrene, frostbite • Lower extremity hyperhydrosis and posttraumatic syndromes leading to pain and venous engorgement. Injection Technique Needle placement is accomplished with image guidance from either CT or fluoroscopy.2,9,10 A long (6–8 in.) 22- or 25-gauge needle is passed via an oblique route from posterior to anterior. The needle tip is positioned along the anterior lateral aspect of the L2 vertebra (Figure 11.4b). Injection of radiographic contrast (3 mL of Omnipaque 240 or equivalent) is used to confirm needle tip position and to ensure the absence of any vascular communication (Figure 11.4c–e). Injection of 10–20 mL of bupivacaine 0.25% will provide temporary relief (This procedure may need to be repeated weekly for several weeks for reflex sympathetic dystrophy.) Administration of 10 mL of absolute alcohol (or 6% phenol) will provide permanent neurolysis, again with general anesthesia. The risks of lumbar sympathetic blockades include intravascular injection into the aorta or inferior vena cava (which may lead to neurological or cardiac toxicity), ureteral injury, and bleeding. Psoas necrosis and visceral perforation have also occurred. Radiofrequency Block Technique The lumbar sympathetic approach to denervate the lumbar sympathetic chain is performed essentially the same way a sympathetic injection is approached.11
239
240
Chapter 11 Autonomic Nerve Blockade
Figure 11.4. (A) The lumbar sympathetic chain, which lies along the anterior-lateral border of the lumbar vertebra. The intended target for a lumbar sympathetic blockade is ipsilateral L2. (B) Slightly oblique image of the lumbar spine with a clamp marking the site for local anesthesia and needle insertion for a right lumbar sympathetic block. Note that the trajectory is slightly over the transverse process.
The patient is placed in a prone position, well padded and cushioned with pillows. Reversal of the lordosis always helps at aligning the spine on an AP fluoroscopic view. By angling the fluoroscope towards the affected target, one then can visualize the transverse process superimposed on the vertebral body. This will allow needle placement on the sympathetic ganglion above the transverse process.
Lumbar Sympathetic Blockade
241
(C) Fluoroscopic image (slightly oblique) showing the 22-gauge needle inserted for the lumbar block. Note the bend at the tip of the needle (arrow), which facilitates steering during insertion. (D) Lateral radiograph shows the tip of the needle along the anterolateral margin of the L2 vertebra.
242
Chapter 11 Autonomic Nerve Blockade
(E) Lateral radiograph after injection of 3 mL of radiographic contrast medium (arrowheads). The contrast material spreads along the margin of the vertebra, and there is no sign of vascular filling. It is now safe to inject the local anesthetic for the sympathetic block (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Next, anesthetize the skin and deeper fascial plane with the “no sting” solution. Then, by using a 22-gauge 3.5 in. spinal needle, approach the target and deposit 3–5 cc of the “no sting” solution to create a track for the radiofrequency needle tip follow. By using a 22-gauge silicone-coated needle with a 10 mm exposed tip, this will allow the use of pulse and heat mode to treat the sympathetic chain. After needle placement has been confirmed in a PA and lateral view, the needle tip should lie against the vertebral body and not away from the lateral wall. After injection of 3–5 mL of either Omnipaque or Isovue contrast, the spread should appear contained within the fascia surrounding the ganglia. If the contrast appears to spread away and caudad, then it may be in the psoas muscle and not the fascia. Next, begin with 0.3-mA sensory testing; a reproducible paresthesia should occur along the lower leg area. Inject 2 mL of 2% Xylocaine prior to beginning to anesthetize the ganglia. Next, use the “pulse” mode at 42° for 120 s; as soon as that portion has been completed, switch the RF generator to “heat” at 70°, and heat-treat the ganglia for 20 s. Once both levels have been treated, transport the patient to the recovery area, monitor for 30 min, and look for bleeding, hypotension, and leg weakness. If recovery is uneventful, the patient can be discharged and followed as an out-patient.
Impar Ganglion Blockade
Hypogastric Plexus Blockade The hypogastric sympathetic plexus is situated at the inferior end of the sympathetic chain and is located just anterior and slightly lateral to the L5–S1 intervertebral disc space (Figure 11.5). It is in close proximity to the iliac artery and vein. Indications Indications for hypogastric plexus blockade include the following: • Upper pelvic malignant pain • Endometriosis to the upper pelvis. Injection Technique The technique for hypogastric plexus blockade involves placement of needles from posterior to anterior by means of fluoroscopic or CT guidance.2,9 The needles pass in an anterior fashion and slightly superior to inferior over the iliac crest in a lateral to medial angulation. The needle tips will lie just anterior to the L5–S1 disc space. Aspiration followed by injection of 3–5 mL of radiographic contrast material ensures that the needle tips are not in a vascular structure (Figure 11.6). Following confirmation of optimal needle tip location, treatment can be with 10–15 mL of bupivacaine 0.25% for temporary relief. For permanent neurolysis, 10 mL of absolute alcohol (or 6% phenol) is injected (with the patient under general anesthesia). Complications result from intravascular injection of alcohol or phenol or injury to the bowel from injection of these substances.
Impar Ganglion Blockade The most caudal ganglion of the sympathetic chain, the impar ganglion, is located anterior to the sacrum and posterior to the rectum (Figure 11.5). It marks the end of the sympathetic chain. It receives innervation from the low pelvis and perineum. Indications Indications for impar ganglion blockade include the following: Intractable low pelvic pain and perineal pain as a result of rectal cancer, uterine cancer, or prostate cancer Endometriosis causing lower pelvic and perineal pain. Injection Technique The technique for impar ganglion blockade involves placement of the needle such that the tip is located just anterior to the surface of the sacrum.2,9 This may require a double curved needle to be angled superiorly and posteriorly such that the needle tip will lie along the anterior face of the sacrum (Figure 11.7). Alternatively, the needle may be passed through the sacrococcygeal junction. Radiographic contrast
243
244
Chapter 11 Autonomic Nerve Blockade
Figure 11.5. The hypogastric sympathetic plexus and the impar ganglion. Variability in location occurs, and the intended block is anterior to the L5–S1 disc and anterior to the sacrococcygeal junction, respectively (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. ImageGuided Spine Interventions. New York: Springer Science + Business Media, 2004).
should be injected to confirm optimal needle tip location and to exclude a position within the rectum or a vascular structure. For temporary relief, 8–10 mL of bupivacaine 0.25% is administered. For permanent relief, 6–10 mL of absolute alcohol or 6% phenol can be administered (with the patient under general anesthesia).
Impar Ganglion Blockade
245
Figure 11.6. (a) Hypogastric plexus blockade in a prone patient. In this posteroanterior view, the needle is directed fluoroscopically from a starting point slightly superior to the iliac crest and lateral to the spine in an inferior–medial direction (arrow). The tip of the needle is situated anterior to L5–S1. Radiographic contrast material (arrowheads) should spread along the prespinus area, but should not be in vessels or the bowel. (b) The lateral view confirms the trajectory of the needle (arrows). The needle tip lies immediately anterior to the L5–S1 disc. Radiographic contrast material (arrowheads) spreads along the anterior aspect of the L5–S1 disc without evidence of spread into the bowel or adjacent vessels (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
246
Chapter 11 Autonomic Nerve Blockade
Figure 11.7. Lateral view of the sacrococcygeal region. A needle, bent to produce a back-looking curve (arrows), is introduced fluoroscopically inferior to the coccyx. It is directed superiorly and posteriorly to position the tip at the anterior face of the lower sacrum near the sacrococcygeal junction. Because the rectum lies immediately anterior to the sacrum locally, radiographic contrast material is introduced to ensure that the needle tip is not inside the bowel (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Complications include puncture or injury of the rectum and nerve root injury during neurolysis.
Conclusion Early intervention in neuropathic and neuralgic conditions will allow patients to regain function of the injured body part with the aid of sympathetic blockade. By carefully performing these various blockades, physicians help patients participate in physical therapy and desensitization of the injured body part, and normality in daily life is reestablished. References 1. Erickson SJ, Hogan QH. CT guided injection of the stellate ganglion: description of the technique and efficacy of sympathetic blockade. Radiology 1993;188:707–709. 2. Waldman SD. Atlas of Interventional Pain Management. Philadelphia: WB Saunders, 1996:269–271. 3. Forouzanfar T, van Kleef M, Weber WE. Radiofrequency lesions of the stellate ganglion in chronic pain syndromes: retrospective analysis of clinical efficacy in 86 patients. Clin J Pain 2000;16(2):164–168. 4. Massad M, LoCiceo III J, Matano J, Oba J, Greene R, Gilbert J, Hartz R. Endoscopic thoracic sympathectomy: evaluation of pulsatile laser, non-
Conclusion pulsatile laser, and radiofrequency-generated thermocoagulation. Lasers Surg Med 1991;11(1):18–25. 5. Garcea G, Thomasset S, Berry DP, Tordoff S. Percutaneous splanchnic nerve radiofrequency ablation for chronic abdominal pain. ANZ J Surg 2005;75(8):640–644. 6. Gangi A, Dietemann JL, Schultz A, Mortazavi R, Jeung MY, Roy C. Interventional procedures with CT guidance: cancer pain management. Radiographics 1996;16:1289–1304. 7. Yarzebski JL, Wilkinson H. T2 and T3 sympathetic ganglia in the adult human: a cadaver and clinical-radiographic study and its clinical application. Neurosurgery 1987;21:339–342. 8. Gimenez A, Martinez-Noguera A, Donoso L, Catala E, Serra R. Percutaneous neurolysis of the celiac plexus via anterior approach with sonographic guidance. Am J Roentgenol 1993;161:1061–1063. 9. Wong W. Management of back pain using image guidance. J Women’s Imaging 2000;2:88–97. 10. Gangi A, Dietemann JL, Mortazavi R, Pfleger D, Kauff C, Roy C. CT-guided interventional procedures for pain management in the lumbosacral spine. Radiographics 1999;18:621–633. 11. Racz GB, Stanton-Hicks M. Lumbar and thoracic sympathetic radiofrequency lesioning in complex regional pain syndrome. Pain Pract 2002;2(3):250–256.
247
12 Percutaneous Vertebroplasty John M. Mathis and Charles Cho
Introduction Vertebroplasty is a term that describes a surgical therapy that has been performed as an open operative procedure for decades, using bone graft, cement, or metal implants to modify or reconstruct damaged or destroyed vertebra.1–12 In these procedures, polymethylmethacrylate (PMMA) has been the cement most often used for reconstruction and augmentation of bone damaged by trauma or tumor invasion.1,3,11,12 Shortly after Galibert et al.13 performed the first Percutaneous Vertebroplasty (PV) in 1984 (by injecting PMMA into a C2 vertebra that had been destroyed by an aggressive hemangioma), Dusquenel adapted the procedure to treat the pain resulting from the compression fractures associated with osteoporosis and malignancy; this was reported by Lapras and colleagues in 1989.14 A small series followed in 1991 by Debussche-Depriester and coworkers, who reported good pain relief in five osteoporotic compression fractures treated with PV.15 Even though the procedure was known to be useful in osteoporotic compression fractures, its early use in Europe focused on the treatment for pain resulting from tumor invasion of the spine. In 1993, PV was introduced into the United States at the University of Virginia by Dion and colleagues. These investigators focused their work primarily on osteoporotic compression fractures and subsequently provided the first clinical series from the United States in which PV was used.16 Their report noted significant pain relief in 85–90% of patients treated for painful osteoporotic compression fractures. This was similar to the early reports about PV from Europe. Since that time, the procedure has grown in popularity and is now becoming the standard of care for pain produced by osteoporotic compression fractures of the spine.17 The osteoporotic population at risk of fracture is huge, with between 700,000 and 1,200,000 vertebral compression fractures a year in the United States resulting from osteoporosis alone.18 The incidence of compression fracture exceeds that of hip fracture, and the direct From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_12, © Springer Science + Business Media, LLC 2010
249
250
Chapter 12 Percutaneous Vertebroplasty
costs of fractures yearly in the United States due to osteoporosis is in excess of $15 billion.18–20 Osteoporosis is greatest in elderly Caucasian females, and the number of affected individuals is growing yearly.20 Additionally, significant numbers of fractures occur in males and in patients receiving steroids for conditions such as cancer, collagen vascular disease, transplant therapy, and severe allergy or asthma. Percutaneous vertebroplasty is indicated in patients who exhibit pain resulting from Vertebral Compression Fractures (VCFs) due to the weakening associated with bone mineral loss secondary to osteoporosis and who are not effectively treated by medical or conservative therapy (i.e., analgesics, bed rest, external bracing, etc.).16,17,21–33 Without PV, chronic pain in these individuals typically lasts from 2 weeks to 3 months.34 The chronic debilitation, limitation of activity, and decline in quality of life resulting from these fractures have been shown to result in depression, loss of self-esteem, and physical impairment. Recent data reveal that vertebral compression fractures are associated with an increased mortality of 25–30% compared with age-matched controls.35 Though less common than osteoporosis, neoplastic disease is well known as a cause of painful VCFs. These fractures can be produced by primary malignant or metastatic lesions, myeloma, and aggressive benign tumors such as hemangiomas. Painful compression fractures may have a clinical picture similar to that of the osteoporotic variety. If the etiology is in question, biopsy should precede or accompany the PV, which will not alter or impair other therapeutic measures such as chemotherapy or radiotherapy. The risk of cement leak is higher with a tumor etiology for VCF than with osteoporosis, generally because the vertebra is less intact. The risk of significant cement leak (or tumor extrusion by the cement) is increased with destruction of the posterior wall of the vertebra. With tumor extension into the spinal canal (even without symptoms), PV will have a high risk of creating or exacerbating neural compression and should generally be avoided.
Patient Selection and Workup Some osteoporotic fractures may generate only mild pain, or there may be a rapid decrease in the initially severe pain after VCF. In either of these situations, PV is usually not indicated. However, persistent pain that limits the activities of daily living or requires narcotic analgesics (with or without hospitalization), may be rapidly diminished with the use of PV. The time between fracture and PV may be prolonged by failed attempts at conservative management or delayed referral. Patients with severe disability requiring hospitalization and parenteral analgesics should be treated immediately. There is no definite medical requirement for delay of therapy with PV if significant benefit to the patient is to be gained by its use. Some patients may present late with chronic, persistent pain, and limitation of normal activity. There are no absolute exclusion criteria based on the time between fracture and PV. However, old fractures (>3 months) are less likely to have beneficial results from PV unless one can show signs of nonunion or signs of
Patient Selection and Workup
251
Figure 12.1. (A) Extreme vertebral compression with the patient in expiration. The vertebral height at the location measured is 8 mm. (B) In inspiration, the vertebral height increases to 11 mm. This motion, though small, is consistent with nonunion and usually associated with chronic, severe pain. This pain will not subside without treatment such as PV (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
recurrent fracture (Figure 12.1). Nonunion is indicated by persistent motion noted on fluoroscopy and can signify osteonecrosis (Kummell’s disease). Also, the finding of persistent marrow edema on Magnetic Resonance Imaging (MRI) scans (which may indicate new or recurrent fracture) is a good indication for PV. Preoperative augmentation of vertebra prior to instrumentation and routine prophylactic use of PV are not validated for benefit or safety at this time, and these measures should be used with extreme caution and only under investigational protocols. On physical examination, the patient’s pain location should be consistent with the anatomical location of the fracture considered for treatment with PV. The patient’s pain should not be radicular, since this suggests nerve root compression. However, it is not uncommon
252
Chapter 12 Percutaneous Vertebroplasty
to have referred pain, and this should not be considered a contraindication to treatment (i.e., referred intercostal pain associated with a thoracic vertebral fracture or referred hip pain associated with a lower lumbar fracture). It is often helpful to place a metallic marker at the site of maximal pain and to correlate fluoroscopically the anatomical location of the pain and the compression fracture. It should be remembered that pain localization is limited to no better than plus or minus one vertebral level in most patients. Simple clinical situations in which physical findings are well correlated with recent radiographic exams may be treated without the addition of complex studies, such as MRI, Computed Tomography (CT), or nuclear medicine (Figure 12.2). However, one may miss bone injury (minimal fracture) that contributes to pain but that can not be recognized by simple radiographs or CT. Routine screenings, therefore, require recent MRI imaging. An MRI should be obtained on all patients
Figure 12.2. Lateral radiograph showing a typical osteoporotic compression fracture (arrow). Compression is typically more in the anterior two thirds of the vertebra, with sparing of posterior wall height (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Patient Selection and Workup
when possible. This MRI should be as recent as possible (never older than 30 days). Patients with multiple fractures or nonfocal pain often pose diagnostic dilemmas and require a more complex imaging evaluation. These patients should have magnetic resonance imaging in addition to a recent, standard radiographic evaluation. Acute fractures will be easily demonstrated on T1-weighted sagittal images as having loss of signal in the affected vertebral marrow space (Figure 12.3). Short-Tau Inversion Recovery (STIR) images with fat suppression offer high sensitivity for recent fracture and marrow edema (represented by an abnormal bright signal in the involved region) . Images made with T2 weighting occasionally give additional information, as these sequences can show fluid-filled clefts that can result after fracture. These findings are important because the clefts or spaces should be filled with cement for dependable pain relief. On T1-weighted MRI sequences, normal marrow will exhibit high (bright) signal, including any vertebra that were previously compressed and have undergone healing. One should be reluctant to perform PV for pain based on MRI unless an acute fracture or persistent marrow abnormality can be demonstrated. If MRI cannot be performed or leaves doubt with respect to the need for therapy, a Nuclear Medicine (NM) bone scan may be utilized. However, NM may not be as useful as MRI for primary screening because the former has poorer anatomical resolution (even when Single-Photon-Emission Computed Tomography (SPECT) is used) and does not give information about conditions such as spinal stenosis, disc herniation, or tumor extension into the epidural space. Also, abnormal activity on a bone scan may persist long after healing has been demonstrated on MRI. A low-level positive NM scan may indicate only normal, progressive healing, which in turn might mislead a physician about the possible benefit of PV.36 However, there is a definite place for NM in patient evaluation. Some patients cannot tolerate MRI, and NM becomes the next best alternative. Rarely, information from the MRI will be insufficient to accurately localize an acute fracture. This usually happens in very heterogeneous marrow (which may be found as a normal variation in the elderly or with conditions such as myeloma). Then, NM will usually add sufficient information to identify an acute fracture or determine the need for treatment (Figure 12.4). Computed tomography offers anatomical information (as do standard radiographs), but it is unable to distinguish acute from chronic fractures under most circumstances. Therefore, CT is not a part of the routine initial patient workup. It may be very helpful to evaluate the cause of complications that are possible after PV, such as a cement leak outside the vertebral body. This mode of diagnosis should be used immediately if symptoms worsen or new symptoms present after PV. The degree of compression does not correlate with the quantity of local pain. Minimal compressions, as measured radiographically, may cause incapacitating pain to some individuals. Even with minimal deformity, acute fractures are easily identified on MRI because they demonstrate local marrow edema. MRI may also show more than one acute compression injury (Figure 12.5). This finding will indicate
253
Patient Selection and Workup
255
Figure 12.4. Nuclear medicine bone scan showing increased uptake at T12 (arrow) resulting from an osteoporotic compression fracture (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
a need for therapy at each of the involved and painful levels. As the amount of compression increases, the degree of technical difficulty of performing the PV may increase as well. This is particularly true when the compression exceeds 70%. With complete or nearly complete vertebral collapse, the likelihood of successful PV is reduced but not eliminated.37,38 Before one attempts PV in a nearly complete collapse,
Figure 12.3. Three sagittal views. (A) The T1-weighted MRI shows an acute vertebral compression (arrow) with low signal in the marrow space. Chronic (healed) compressions have normal (bright) marrow signal (stars). (B) The STIR MRI reveals high signal in the marrow space of the acutely fractured vertebra (arrow). (C) The T2-weighted MRI demonstrates a high signal zone below the superior endplate in a recently fractured vertebra (arrow). This is believed to represent a fluid-filled cleft. Filling of the cleft with cement is essential for pain relief (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
256
Chapter 12 Percutaneous Vertebroplasty
Figure 12.5. Sagittal T1-weighted MRI revealing two acute fractures (arrows) at different locations in the spine (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
one should obtain an MRI that indicates no additional cause of pain. The same MRI allows one to evaluate the quantity of residual marrow space in the vertebra to be treated. Often, severe collapse is greatest centrally, while sparing residual marrow space laterally that can be successfully treated with PV (Figure 12.6). Patients with these lesions should be made aware that there may be a reduced chance of pain relief (in comparison to a modestly compressed vertebral fracture) and higher risk of complication.
Cement Selection and Preparation
257
Figure 12.6. (A) Sagittal T1-weighted MRI (midline) reveals extreme compression of the center of the L1 vertebral body (arrow). (B) Images along the lateral edge of L1 reveal less compression and more residual marrow space, which can accept bone cement (arrow) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Although PV has been shown to be very durable, on rare occasions, one may see a refracture with progressive height loss after PV. This usually occurs when the patient has had a less than optimum fill during an initial treatment (even with good initial pain relief) or in the situation of an extremely fragile vertebra. In either case, the amount of cement introduced probably was not sufficient to restore adequate strength to resist recurrent compression. Pain relief and cement filling are poorly correlated. Recurrence of pain, marrow edema, and additional vertebral collapse may indicate the rare need for repeat treatment.
Cement Selection and Preparation The first bone cement used for PV was the PMMA Simplex™ P (Stryker Howmedica Osteonics, Mahwah, NJ).13 This was the only cement approved by the U.S. Food and Drug Administration (FDA)
258
Chapter 12 Percutaneous Vertebroplasty
for use in the treatment of pathological fractures in the spine. It was not specifically approved for PV. There were no cements approved for PV in the early days of the procedure. Multiple non-approved PMMA cements were used for PV and seem to have had similar clinical results.16,17,33 It is important to note that bone cement is not treated as a pharmaceutical by the FDA but rather as a device. Alterations in the composition are therefore equivalent to making a new (non-approved) material. It has been suggested by other authors that such alterations constitute “off-label” use.39 Off-label use would be correctly applied if an unaltered cement were used in a non-indicated application or location. Alteration in the ratio of monomer to copolymer (liquid to powder) or addition of other materials (opacification agents or antibiotics) results in the creation of a new material, and the FDA approval no longer exists. Patients should be informed that such alterations in the cement are to be used, and the reasons and consequences behind these changes should be discussed. Fortunately, there are now multiple FDA-approved bone cements for both PV and balloon-assisted PV. Under all but the most unusual circumstances, cement should be used as supplied and not modified. Inherent in performing PV safely is the need to accurately monitor the injection of cement in real time.33 This is usually accomplished with fluoroscopy and requires that the cement be opacified so that it may be adequately seen in small quantities during introduction. It has been determined that barium sulfate, in quantities of 30% by weight mixed with the PMMA, will provide an adequate level of opacification.33,40,41 All FDA-approved cements have appropriate opacification for fluoroscopic monitoring.42 Some investigators add antibiotics routinely to PMMA prior to injection, the most common antibiotic being tobramycin.16,33 However, the infection rate with PV is very low, and the efficacy of adding antibiotics to the cement has not been scientifically substantiated in normal, uninfected patients. One report in the orthopedic literature did show reduced infection rates in hip replacements in which cement-containing antibiotics were used for immunosuppressed patients.43 As there is no scientific indication to add antibiotics in otherwise normal patients and because this addition alters the cement and negates its FDA-approved composition, the authors do not recommend the addition of antibiotics to cements except in the situation of immunocompromise. Adequate precaution should be used during cement mixing to maintain sterility. Cement manufacturers should provide closed, vacuummixing devices that aid in maintaining a sterile environment. Open mixing, which increases the risk of cement contamination and reduces the cement strength by the inclusion of air bubbles, should be avoided. Thick PMMA seems to limit leaks and these leaks, when large, are associated with complications. However, small leaks (not clinically significant) are simply technical events like small, non-significant blood loss at surgery. Though traditional percutaneous vertebroplasty and Kyphoplasty (KP) have been performed with PMMA, new cements (non-PMMA) are being developed in an attempt to improve on short comings know
Image Guidance
to exist with PMMA. These problems include: 1.) no intrinsic radioopacity, requiring the addition of non-structural materials like barium, 2.) toxic monomer leaching and a very high exotherm during polymerization (both of these are capable of killing adjacent cells within the bone), 3.) bone treats PMMA as a foreign body (actually forming a scar adjacent to it) with no local bonding. The first of these biologic materials has now been approved by the FDA for use in both PV and KP. This material is named Cortoss (Orthovita; Malvern, Pa). A multi-year, randomized trial was performed against PMMA and demonstrated that Cortoss was equivalent to PMMA in relieving pain and safety. It had better outcome profiles in long term improvement in patient function as well as in lower subsequent and adjacent fractures. Cortoss is a modern, improved cement for bone augmentation. It contains a bioactive ceramic which bone bonds to forming a tight connection and eliminating loosing. It has no monomer leaching and low exotherm eliminating the toxic reactions seen with PMMA. Because its’ biomechanical properties are more similar to native bone, it requires a smaller amount to restore strength and stiffness. As less needs to be injected, leaks may be reduced. Finally, its mixing system is totally different from PMMA. It uses a dual cartridge device that only combines the two active elements of the cement when needed (mix on demand) eliminating the short work time issues faced with PMMA. These changes should markedly improve PV and KP.
Informed Consent Written permission for the procedure is recommended, following a complete discussion of the procedure, including the risks and complications, with the patient and/or the patient’s representative. This must include a discussion of any intent (or need) to modify the cement from its FDA approved mixture. These type of modifications (or the performance of non-standard or investigational procedures) should be used only under the supervision of an IRB and approved investigational protocol.
Image Guidance Since the first PV procedure,13 fluoroscopy has been the preferred method of image guidance for performing PV, although CT has been used infrequently as a primary or adjunctive tool.44,45 Because this procedure was initiated and popularized by interventional neuroradiologists, biplane fluoroscopic equipment was commonly available and often used. This equipment allows multiplanar, real-time visualization for cannula introduction and cement injection, and permits rapid alternation between imaging planes without complex equipment moves or projection realignment (Figure 12.7). However, this type of radiographic equipment is expensive and is not commonly available in interventional suites or operative rooms unless it is used for neurointerventional procedures. It takes longer to acquire two-plane guidance and monitoring information with a single-plane than with a biplane system. However, it is feasible and safe to use a single-plane fluoroscopic system as long as the
259
260
Chapter 12 Percutaneous Vertebroplasty
Figure 12.7. Biplane fluoroscopy/angiography room. The ability to perform fluoroscopy in two projections without having to move equipment greatly speeds and simplifies vertebroplasty (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
operating physician recognizes the necessity of orthogonal projection visualization during the PV to ensure safety. With a single-plane system for PV, these C-arm moves will mean a slower procedure than that offered by a biplane system. Gangi et al.45 introduced the concept of using a combination of CT and fluoroscopy for PV. This method gained a brief period of popularity in the United States with the study published by Barr et al.44 Barr and colleagues subsequently abandoned CT for routine PV. Although the contrast resolution with CT is superior to that of fluoroscopy, with CT, one gives up the ability to monitor needle placement and cement injection in real time. Even so, CT may be acceptable for needle placement, particularly if a small-gauge guide needle is first placed to ensure accurate and safe location before a large-bore bone biopsy system is introduced. However, CT certainly is not optimal for monitoring the injection of cement. For this reason, Gangi et al.45 and Barr et al.44 used fluoroscopy in the CT suite during cement introduction. CT does not afford one the opportunity to watch the cement as it is being injected or to alter the injection volume in real time if a leak occurs. Also, unless a large section is scanned with each observation, leaks can occur outside the scan plane, and they may be missed if one is looking only locally in the middle of the injected body. Barr et al.44 used general
Anesthesia
anesthesia with CT guided surgery because of the need to minimize patient motion. This was successful but added a small additional risk to the procedure and considerable complexity and cost. For all these reasons, CT has not found a primary role in image guidance for PV; it is reserved for extremely difficult cases.
Laboratory Evaluations Coagulation test results should be normal, and the patient should not be taking Coumadin® (Dupont-Merck, Wilmington, DE). Coumadin® may be discontinued and replaced with enoxaparin sodium (Lovenox®, Rhône-Poulenc Rorer Pharmaceuticals, Inc., Collegeville, PA), given once or twice a day on an outpatient basis. Coumadin® may also be stopped and replaced with heparin, but this medication must be administered intravenously, requiring hospital admission. Both enoxaparin sodium and heparin can be reversed with protamine sulfate before PV and restarted postoperatively. Aspirin use is not a contraindication to the procedure, but most interventional groups now take a much more conservative approach to these drugs as well. (The Society of Interventional Radiology recommends stopping anti-platelet drugs 5 days prior to interventional procedures). PV is not recommended for patients with signs of active infection, but elevated white blood cell counts clearly associated with medical conditions such as myeloma or secondary to steroid use are not contraindications.
Antibiotics For PV, as for other surgical procedures that implant devices into the body, intravenous antibiotics are routinely given before (usually 30 min) the procedure is begun. The most common antibiotic used in this application is cefazolin (1 g).46 If an alternative must be used because of allergy, ciprofloxacin (500 mg orally, two times daily) may be substituted and continued for 24 h after the completion of the procedure. Optimally, an oral antibiotic should be started 12 h before the PV procedure. As mentioned earlier, antibiotics are added to the cement itself only in the situation of immunocompromise (and this renders the cement no longer FDA approved).
Anesthesia During PV, it is common to use both local anesthetics and conscious sedation to make the patient comfortable and relaxed. Patients who request not to receive intravenous sedation or cannot have it for safety reasons still can be treated with only mild discomfort if appropriate attention is given to local anesthetic placement. To reduce the sting and discomfort associated with locally administered anesthetics (lidocaine, etc.), one may buffer the anesthetic by the addition of a mixture of 1 mL of bicarbonate and 9 mL
261
262
Chapter 12 Percutaneous Vertebroplasty
Table 12.1. Modified local anesthetic solutions Amount (mL) Lidocaine (4%)
Lactated Ringer’s
Bicarbonate
1
4
24
2
0
2
4
24
2
0.15 (1:1,000)
Solutiona
Epinephrine
Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004. a Solution 1 makes a “sting-free” local anesthetic with 0.5% lidocaine. Solution 2 is “sting free” with 0.5% lidocaine and 1:200,000 epinephrine. These preparations should be mixed daily and discarded at the end of the day. The total volume of each mix is 30 mL.
of lidocaine. This mixture reduces, but does not eliminate, the anesthetic sting. The authors commonly use a mixture that includes both bicarbonate and Ringer’s lactate (Table 12.1), which essentially eliminates the sting of the local anesthetic. At the authors’ institution, this mixture is prepared daily for all procedures requiring local anesthetics. The excess is discarded at the end of each day, as it contains no preservative. This preparation has a low concentration of lidocaine (0.5%) and allows the use of a more generous volume locally with less risk of toxicity. Whatever the chosen local anesthetic preparation, the skin, subcutaneous tissues along the expected needle tract, and periosteum of the bone at the bone entry site must be thoroughly infiltrated. When this has been accomplished, the patient will experience only mild discomfort while the bone needle is being placed, regardless of whether conscious sedation is used. Conscious sedation has become a common adjunctive method of pain and anxiety control in awake patients who undergo minimally invasive procedures. The authors use a combination of intravenous midazolam (Versed®, Roche, Manati, PR) and fentanyl (Sublimase®, Abbott Labs, Chicago, IL). To decrease anxiety and diminish the discomfort associated with positioning, it may be helpful to begin these medications before the patient is placed on the operating table. Dosages are chosen according to patient size and medical condition. The final amount is determined with titration while observing the patient’s response. General anesthesia is rarely needed for PV, but it is used occasionally for patients in extreme pain who cannot tolerate the prone position used in PV or for patients with psychological restrictions that preclude a conscious procedure. It is not needed for routine PV and should be avoided when possible because it adds a mild risk and considerable cost to the procedure. As described earlier, Barr et al.44 used general anesthesia routinely with CT-guided procedures to ensure minimum patient motion.
Needle Introduction and Placement The original choice of a device for percutaneous cement introduction was based on device availability. The size of the needles was empirically chosen to allow the viscous PMMA cement to be injected.
Needle Introduction and Placement
Originally, 10- to 11-gauge trocar-cannula systems were used. It is becoming progressively common to see smaller gauge (13–15) needles used routinely. All will work with the least resistance during injection found with the larger bore, while the smaller needles are useful in small pedicles or in the cervical spine. From the thoracic through lumbar spine, a 13-gauge cannula can be placed through the adult pedicle without fear of it being too large. Several introductory routes for needle delivery are possible, including: (1) transpedicular, (2) parapedicular (transcostovertebral), (3) posterolateral (lumbar only), and (4) anterolateral (cervical only). The classic route for most PV is transpedicular. It offers the following advantages: 1. It usually provides the operating physician with a definite anatomical landmark for needle targeting (Figure 12.8). 2. It is very effective for PV and for biopsy of lesions inside the vertebral body. 3. It is inherently safe and does not carry the risk of needle damage to other adjacent anatomical structures (nerve root, lung, etc.) as long as an intrapedicular location is maintained. In the upper thoracic region and in small patients, the pedicle may be too narrow for a 11-gauge needle. In this situation, a 13-gauge needle should be used. The parapedicular or transcostovertebral approach (Figure 12.9) was devised to allow access when the transpedicular route is not desirable or possible (e.g., small pedicle). Since the needle passes along the lateral aspect of the pedicle rather than through it, a small pedicle does not preclude using a 11-gauge needle for cement introduction. Also, this approach angles the needle tip more toward the center of the vertebral body than does the transpedicular approach. At least in theory, this angle may allow easier filling of the vertebra with a single injection. There is a higher chance of pneumothorax with a parapedicular approach than with the transpedicular route. A second potential problem with the parapedicular route is that the needle enters the body only through its lateral wall. This approach may increase the risk of paraspinous hematoma after needle removal. Because with a parapedicular approach the osteotomy site occurs laterally along the side of the vertebra, one cannot apply local pressure after needle removal as can be done with the transpedicular route. In the cervical spine, a transpedicular route is very difficult, so an anterolateral approach may be used as an alternative. Needle introduction must avoid the carotid–jugular complex. To accomplish this goal, the operating physician (as in cervical discography) can manually push the carotid out of the path of the needle. Alternatively, CT can be used to visualize the carotid, and a trajectory that will miss the vascular structures can then be chosen. A small guide needle can be inserted to ensure accurate placement outside the carotid complex. The authors prefer the guide needle alternative because it gives positive guidance and confirmation without excessive fluoroscopy to the physician’s hands during needle introduction. However, because osteoporotic fractures in this area are rare, the cervical spine only occasionally undergoes PV. Neoplastic disease may produce the occasional need for PV intervention in this region.
263
Figure 12.8. (A) Typical transpedicular route for needle placement into the vertebral body. (B) Anterior–posterior radiograph demonstrates the placement of the needle through the pedicle, which is seen as a well-circumscribed oval (arrow). In this projection, the needle is initially positioned during fluoroscopy while being held with a clamp (arrowhead) to avoid x-ray exposure to the operator’s hands. (C) Lateral fluoroscopic image demonstrates the final needle position beyond the midline of the vertebra (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Needle Introduction and Placement
265
Figure 12.9. (A) Needle location for parapedicular (extrapedicular) placement. (B) Lateral projection demonstrating that the needle must enter above the transverse process on the parapedicular approach (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Once the needle route is chosen, local anesthesia is administered, and a small dermatotomy incision is made with a no. 11 scalpel blade. The trocar–cannula system is introduced through the skin incision and subcutaneous tissue to the periosteum of the bone. This introduction can be facilitated with a sterile clamp to guide the needle during fluoroscopy, thus avoiding radiation to the operating physician’s hands (Figure 12.8b). In osteoporotic bone, penetrating the bone cortex and advancing the needle into the vertebral body is usually very easy. In a patient with neoplastic disease, the bone still may be very dense and strong (except where it has been destroyed by a tumor). The use of a mallet to advance the needle through very dense bone is a technique clearly superior to manual advancement. Regardless of whether a transpedicular or parapedicular route has been chosen, the tip of the needle should lie beyond the vertebral midpoint as viewed from the lateral projection. The authors usually try to obtain an even more anterior position by placing the needle tip at the junction of the anterior and middle thirds of the vertebra. Two needles are routinely placed, usually via the transpedicular approach. This takes minimally longer than a single needle placement and affords a large margin of safety for being able to dependably complete a vertebral fill with a single mix of cement. There is no question that a single needle placement can give an adequate fill in a large number
266
Chapter 12 Percutaneous Vertebroplasty
of cases. However, the single-needle method fails to produce uniform fills more often than the double-needle technique and may oblige the operator to accept a larger cement leak during filling (if the second needle is not in place as an alternate injection route).
Venography Venography was never used much in Europe and was introduced in the United States in an attempt to discover potential leak sites prior to injecting cement. However, this technique worked poorly because the contrast material and the bone cement are very different in viscosity. The authors discontinued using venography in 1996 and have found no disadvantage or added risk without its use.47 Other long-term proponents have belatedly stopped its use in routine PV as they found no safety benefit.48
Cement Injection Cement is prepared only after all needles are placed, as described in the earlier section of this chapter on “Cement Selection and Preparation.” Cement with an appropriate opacification is prepared and injected using small syringes (typically 1 mL) or devices made specifically for injection (Figure 12.10) (no actual advantage to a specific PMMA
Figure 12.10. Cement injection with a 1 mL syringe. Note bipedicular needle placement prior to beginning cement injection (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Cement Injection
versus another has been shown. The newly FDA approved biologic cement Cortoss (Orthovita; Malvern, Pa) does give great promise to improve many of the short coming seen with PMMA). Either the cement injection should be monitored in real time, or small quantities (i.e., 0.1–0.2 mL) should be injected and the result visualized before additional cement is introduced. The latter approach, which allows one to step back from the fluoroscopy beam during visualization, minimizes radiographic exposure to the operator. Any cement leak outside the vertebral body is an indication to stop the injection. When using a rapidly polymerizing cement, halting the injection may be necessary only for a minute or two while the injected cement hardens. Restarting the injection may then redirect flow into other areas of the vertebra. If leakage is still seen, it is advisable to terminate the cement injection through this needle and move to the second needle. This will usually allow completion of the vertebral fill without further leakage, since the original leak now will be occluded by the initial cement, which will have hardened. One should work through a single needle at a time. This avoids contamination of both needles at once and preserves a route for subsequent injection if a leak is encountered. Cement can still be introduced beyond the point at which most injection devices begin to fail. The trocar is useful to push additional thick cement from the cannula into the vertebra. Bone fillers are special cannula-plunger systems that fit through the existing bone cannula. They allow the injection of very thick cement (beyond what can be injected with a syringe or injector). They also prevent cement from touching the introductory cannula and precluding cannula closure by hardening cement. This allows an additional margin of safety as the cannula remains open allowing multiple injections as needed to complete the fill and minimize leaks. The on demand mix of the new cement Cortoss makes this option much more viable. The 5 in., 13-gauge cannula holds 0.5 mL, and the 5 in., 11-gauge cannula holds 0.9 mL. Reintroducing the trocar will push the residual cement in the cannula into the vertebra. This is done only if the additional amount of cement is desired. The cannula can be removed safely without reintroduction of the trocar when the cement has hardened beyond the point at which it can be injected. Simply twisting the needle through several revolutions will break the cement at the tip of the cannula and will prevent leaving a trail of cement in the soft tissues. However, removing the cannula before the cement has hardened sufficiently can allow cement to track backward from the bone into the soft tissues and may create local pain. The amount of cement needed to produce pain relief has not been accurately documented in available clinical reports. The authors believe that pain relief is related to fracture stabilization (not to a chemical or thermal effect), and thus the amount of cement needed to restore the initial vertebral body’s mechanical integrity should also give an approximation of the quantity needed to relieve pain clinically. In an in vitro study, the authors showed that the initial pre-fracture strength and stiffness of a vertebra could be restored by injecting 2.5–4 mL of PMMA in the thoracic vertebra, while 6–8 mL provided
267
268
Chapter 12 Percutaneous Vertebroplasty
similar augmentation in the lumbar region. (a smaller quantity of Cortoss will be required to cover the same region of the vertebra compared to PMMA).49 A reasonable guideline for the quantity of any cement to be injected is the amount that is needed to fill 50–70% of the residual volume of the compressed vertebra (determined by visual estimate during cement introduction). These amounts should not be taken as absolute but rather as a guide. The previously described study suggests that relatively small amounts of cement are needed to restore initial biomechanical strength and that these amounts vary with the relative vertebral level in the spine, as well as with individual vertebral body size and the degree of vertebral collapse. The authors also have demonstrated that significant strength restoration is provided to the vertebral body with a unipedicular injection, where cement filling crosses the midline of the vertebral body.50 This would imply that unipedicular fills that achieve adequate cement injection volumes are likely to be successful at achieving pain relief. This fact notwithstanding, there is a higher likelihood of achieving more uniform fills, with fewer leaks, when two needles are used rather than one (Figure 12.11).
Postoperative Care After adequate vertebral filling has been accomplished, the needle is removed. Occasionally, venous bleeding is experienced at the needle entry site. Hemostasis is easily achieved with local pressure for 5 min. The entry site is dressed with Betadine ointment and a sterile bandage. The patient is maintained recumbent for 1–2 h after the procedure and monitored for changes in neurological function or for signs of any other clinical change or side effects. Table 12.2 lists typical postoperative orders. Any sign of adverse events should trigger the use of appropriate imaging modalities (usually CT) in the search for an explanatory cause. It is well known that 1–2% of patients will have a transient period of benign increase in local pain following PV. However, this is a diagnosis of exclusion and should prompt extended monitoring (or hospitalization if the pain is severe and requires aggressive therapy) and imaging evaluation to exclude other causes for the pain (such as cement extravasation). Pain alone will usually be adequately treated with analgesics, nonsteroidal anti-inflammatory drugs (such as Toradol®, Roche Pharmaceuticals, Nutley NJ), or local steroid injections adjacent to affected nerve roots or in the epidural space. Large cement leaks (Figure 12.12) or neurological dysfunction should prompt an immediate surgical consultation. PV is easily performed on an outpatient basis, with the patient discharged after 1–2 h of uneventful recovery. Table 12.2 gives typical discharge instructions. Follow-up is indicated to monitor the results of therapy and should be incorporated into a quality management program. Reports of complications and results should be maintained by the facility as well as for each individual provider. Additional informa-
Postoperative Care
269
Figure 12.11. (A) Anterior–posterior radiograph showing a good bipedicular vertebral fill of bone cement. (B) Lateral radiographs show the same vertebra. Note that the entire central volume of the vertebra is not filled (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
270
Chapter 12 Percutaneous Vertebroplasty
Table 12.2. Sample postoperative orders and discharge instructions Postoperative orders Bed rest 1 h (may roll side to side) May sit up after 1 h with assistance Vital signs and neurological examinations (focused on the lower extremities) every 15 min for the first hour, then every 30 min for the second hour Record pain level (Visual Analog Scale, 1–10) at end of procedure and at 2 h postoperatively (before discharge). Compare with baseline values and notify physician if pain increases above baseline May have liquids by mouth if no nausea Discontinue oxygen (if used) after procedure (if saturation is normal) Discontinue intravenous drips after 1 h if recovery is otherwise uneventful Discharge patient home with adult companion after 2 h if recovery is uneventful Discharge instructions Return home; bed rest or minimal activity for next 24 h May resume regular diet and medications Keep operative site covered for 24 h. Bandages may then be removed and site washed with a damp cloth. Do not soak Notify physician or facility if there is increasing pain, redness, swelling, or drainage from the operative site Notify physician or facility if there is difficulty with walking, changes in sensation in hips or legs, new pain, or problems with bowel or bladder function The area of the procedure will be tender to the touch for 24–48 h This is to be expected If there is pain similar to that before the procedure, prescribed pain medications may be continued as needed Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004.
tion and recommendations about the credentialing and quality management for PV can be found in the American College of Radiology manual on standards of practice (ACR.org/StandardsofPractice).
Results To date there are no substantial prospective, randomized trials evaluating PV published in the literature. However, Zoarski and colleagues presented a small prospective (nonrandomized) evaluation of the effectiveness of PV for relieving pain.51 This report utilized the MODEMS method to establish that 22 of 23 patients improved after PV and remained satisfied during the 15- to 18-month follow-up. Additionally, numerous retrospective series are available and uniformly report good pain relief and reduced requirements for analgesics following PV.16,17,22,27,44 This is especially true of pain related to compression fractures produced by osteoporosis, where significant pain relief of between 80 and 90% has been observed. This pain relief is persistent with no reports of additional compression of vertebra previously treated with PV. Additional fractures at other levels remain a possibility and source of morbidity. If osteoporotic compression fracture occurs, every effort to minimize future bone loss
Results
Figure 12.12. CT scan of a patient who experienced paraplegia following vertebroplasty as a result of a large cement leak. The cement (stars) occupies a large amount of the spinal canal at the level of the CT scan and creates cord compression (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
medically should be made. Also, modifications in lifestyle should be attempted to minimize mechanical stress on the spine and thereby lessen the risk of additional fractures (NEJM articles discussed below.) Much discussion has been given as to whether there is an advantage of balloon-assisted vertebroplasty (kyphoplasty) compared to traditional vertebroplasty. Again, there are no good randomized trials comparing the two techniques. Indeed, the mechanical stabilization used in both is the same (cementation). Safety is probably similar, though a higher permanent complication and death rate is reported with kyphoplasty. A consensus statement by multiple societies using both techniques (American Society of Neurological Surgeons/Congress of Neurological Surgeons, American Society of Spine Radiology, and American Society of Interventional and Therapeutic Neuroradiologist) was issued in 2007 and states: “After reviewing the published literature on kyphoplasty, the Societies have determined that the clinical response rate in individuals treated with kyphoplasty is equivalent to that seen in patients treated with vertebroplasty. There is no proved advantage of
271
272
Chapter 12 Percutaneous Vertebroplasty
kyphoplasty relative to vertebroplasty with regard to pain relief, vertebral height restoration, or complication rate.”52 There has been growing concern over the risk of secondary vertebral fracture after treatment, especially at adjacent levels. However, the natural history of vertebral fractures treated conservatively is approximately 20% at 1 year. With the incidence of two fractures, incidence increases to 24%. Fifty percent of these new fractures will be in adjacent levels.53 This compares very well with reported incidence of new and adjacent levels of fractures occurring following treatment with PV or KP.54
Complications Complications, though initially considered to be uniformly low, are unfortunately higher for inexperienced physicians or those who attempt the procedure without adequate image guidance or cement opacification. Appropriate training needs to be completed before the procedure is attempted. Recommendations can be obtained from the American College of Radiology Standards of Practice on Percutaneous Vertebroplasty (ACR.org/StandardsofPractice). In osteoporosis-induced vertebral fractures, clinical reports of complications are around 1%.16,17,22,27 Many of these are transient and include increase in local pain after cement introduction (nonradicular and not associated with neurological deficit). This is usually easily treated with nonsteroidal anti-inflammatory drugs and resolves within 24–48 h. Uncommonly, cement leaking from the vertebra adjacent to a nerve root will produce radicular pain. Analgesics combined with local steroid and anesthetic injections usually provide adequate relief. A trial of this type of therapy is warranted as long as there are no associated motor deficits. The discovery of a motor deficit (or bowel or bladder dysfunction) should initiate an immediate surgical consultation. This type of severe complication will almost always be associated with large-volume leaks that have resulted in neurological compression. Correction of the complication surgically should be considered a medical emergency. Cement leaks also have been implicated in producing pulmonary embolus.16 These are usually not symptomatic but rarely have produced the clinical symptoms accompanying pulmonary infarct. With a right-to-left shunt, this can result in cerebral infarct.55 Likewise, infection has been reported but is rare with PV. The complication rate found when treating compression fractures resulting from malignant tumors is considerably higher.22,26,29,30,56 This occurs because there are frequently lytic areas involving the vertebral cortex and a greater propensity for cement to leak into the surrounding tissues or vessels. Cement leaks causing symptoms in this setting occur in 5–10% of patients; again, most are transient. Much is made of cement leaks, but often without distinguishing the type of leaks. A small leak that has no clinical consequence should be thought of as a “technical event,” much like a small and insignificant blood loss at surgery. Only the rare (and usually large) leak
Pain relief after Vertebroplasty: Real or Sham
causes patient injury and is then termed as a “clinical complication.” Unfortunately, these distinctions have rarely been made in the vertebroplasty/kyphoplasty literature. Indeed, marketing has selectively used these “technical events” to try to gain advantage routinely. Death is a rare complication associated with PV and KP. Though the exact details usually are not known, the likely cause seems to be pulmonary compromise, which is suspected to be due to fat (from the vertebral marrow) or cement emboli. A safe number of vertebrae to treat at one time has yet to be definitely established. Mathis and colleagues reported treating seven vertebrae in a 35-year-old patient with multiple fractures associated with steroid use for lupus.46 This patient’s therapy occurred in three treatment sessions. Because the introduction of cement is a hydraulic event with as much marrow pushed out of the inter-trabecular space as cement injected, there is concern about fat emboli in large volume cement injections. For reasons described earlier, the authors recommend treating no more than three vertebrae in any one session in healthy individuals and less if there is known cardiopulmonary disease. Those with increased risk should be appropriately warned as part of their consent process. There are no data that support the prophylactic use of PV to treat vertebra that are believed to be at risk of fracture. Except for prophylactic use, there is little conceivable reason to perform PV on large numbers of vertebrae at one time. No safety is gained with kyphoplasty, as the balloon also displaces marrow products that go to the lungs. Any deviation from an expected good result (such as increased pain or neurological compromise) should initiate an immediate imaging search with CT to look for a cause of the clinical change. Unremitting or progressive symptoms may require surgical or aggressive medical intervention, and outpatients should be hospitalized and monitored.
Pain relief after Vertebroplasty: Real or Sham With over 1000 positive peer reviewed papers about Percutaneous Vertebroplasty (PV) or balloon assisted vertebroplasty or (Kyphoplasty) and with hundreds of thousands cases performed, it seems obvious to those of us that have performed these procedures that pain relief should be considered very real. However, 2009 articles in the New England Journal of Medicine described sham therapy that was equivalent to PV for pain relief in compression fractures of the spine.57,58 The NEJM articles would not have gotten nearly as much interest if good, randomized studies of PV against conservative therapy (the traditional, historic therapy used for compression fractures prior to PV) had been initially performed. However, these were not accomplished for numerous reasons. The questions raised by the NEJM articles therefore had a magnified effect. Both NEJM articles had a limited number of patients enrolled despite a prolonged acquisition period (over 4 years for the Kallmes study, the largest by a factor of 2). The Kallmes study intended to enroll 250 patients but cut its study off after 4.5 years with 131 total
273
274
Chapter 12 Percutaneous Vertebroplasty
patients.59 It found (but did not emphasize) a trend toward a “clinically meaningful improvement in pain for PV compared to sham therapy” (64% of PV patients vs. 48% of sham treated patients). This fell short of statistical significance with a P=.06 ( P=.05 was needed to be statistically significant). Obviously, the P value lacked only .01 reaching significance for PV to be better than sham. If one simply continues the proportions of improved patients in each group, an additional 19 patients were needed to reach this clinical significance of PV over sham. This would have been a total number of patients of 150, still far below the intended number of 250 which were initially planned for. This simple change in the study would have nullified most of the importance as it would have shown PV better than sham in this category. The 131 number did not provide the sensitivity to predict this outcome. Patients in the sham group did get initial pain relief equal to PV. However, it must be remembered that this sham therapy is not conservative therapy. The sham procedure consisted of injecting anesthetic into the periosteum of the fractured bone (along its posterior cortex where the pedicle entry point would be if PV had been performed). One may think this therapy should be insignificant and very transient in effect. However, similar injections have been used for years in pain management for spine pain such as facet blocks and for other pain control like occipital nerve blocks for migraine headaches. For those of us that perform such procedures, we commonly see weeks and sometimes months of relief from problems like migraine headaches with these simple anesthetic injections. Therefore, it is not surprising that pain relief was seen with this “sham” procedure. However, very telling is the fact that crossover from one group to another found 43% of the sham patients crossing over to the PV group after 1 month (regardless of the amount of pain relief these patients claimed initially, it obviously did not last). Only 12 % of PV patients crossed over (this is exactly expected as we find 85–90% of patients getting good pain relief traditionally after PV.60 The crossover numbers were highly significant with a P=.001. So sham therapy was associated with initial pain relief but failed to be persistent in many treated in this group. Is this sham a real therapy or a placebo? Probably we will never know for sure and from a clinical point of view, it may matter little. New evidence published in Science (2009) found functional MRI evidence of changes in uptake in the dorsal columns (sensory nerve region) when a placebo was applied to the skin in regions of pain.61 This paper shows the strong and real effect of placebo, long known to be active in most pain management treatments. The NEJM articles found real or placebo effects of pain reduction with “sham” (not conservative) therapy. They found that the sham therapy was not long lasting with a statistically significant number of patients crossing over to the PV group after one month. Finally, if these papers had completed there initially advertised patient accrual (250 patients for Kallmes), a clinically significant improvement in PV over Sham would have been established as well.
Conclusion
Conclusion Percutaneous vertebroplasty has been shown to be very effective at relieving the pain associated with compression fractures of vertebra caused by both primary (age-related) and secondary (steroid-induced) osteoporosis. It also has substantial benefit in neoplastic-induced vertebral compression fracture pain but with a higher chance of associated complication. PV is rapidly becoming the standard of care for compression fracture pain that does not respond to conservative medical therapy. However, this simple procedure must be treated with respect, for its application without appropriate judgment and physician training can quickly result in increased pain, permanent neurological injury, and even death. Consensus has now been reported to show no benefit of kyphoplasty over vertebroplasty with regard to pain relief, vertebral height restoration, or complication rate.52 Due to the high cost of kyphoplasty compared to vertebroplasty, kyphoplasty therefore seems rarely warranted for general medical use. References 1. Alleyne CH Jr, Rodts GE, Jr, Haid RW. Corpectomy and stabilization with methylmethacrylate in patients with metastatic disease of the spine: a technical note. J Spinal Disord 1995;8:439–443. 2. Cortet B, Cotten A, Deprez X, Deramond H, Lejeune JP, Leclerc X, Chastanet P, Duquesnoy B, Delcambre, B. [Value of vertebroplasty combined with surgical decompression in the treatment of aggressive spinal angioma. Apropos of 3 cases.] Rev Rheum Ed Fr 1994;61:16–22. 3. Cybulski GR. Methods of surgical stabilization for metastatic disease of the spine. Neurosurgery 1989;25:240–252. 4. Harrington KD. Anterior decompression and stabilization of the spine as a treatment for vertebral collapse and spinal cord compression from metastatic malignancy. Clin Orthop 1988;233:177–197. 5. Harrington KD, Sim FH, Enis JE, Johnston JO, Diok HM, Gristina AG. Methylmethacrylate as an adjunct in internal fixation of pathological fractures. Experience with three hundred and seventy-five cases. J Bone Joint Surg (Am) 1976;58:1047–1055. 6. Knight G. Paraspinal acrylic inlays in the treatment of cervical and lumbar spondylosis and other conditions. Lancet 1959;2:147–149. 7. Kostuik JP, Errico TJ, Gleason TF. Techniques of internal fixation for degenerative conditions of the lumbar spine. Clin Orthop 1986;203:219–231. 8. Mavian GZ, Okulski CJ. Double fixation of metastatic lesions of the lumbar and cervical vertebral bodies utilizing methylmethacrylate compound: report of a case and review of a series of cases. J Am Osteopath Assoc 1986;86:153–157. 9. O’Donnell RJ, Springfield DS, Motwani HK, Ready JE, Gebhardt MC, Mankin HJ. Recurrence of giant-cell tumors of the long bones after curettage and packing with cement. J Bone Joint Surg (Am) 1994;76:1827–1833. 10. Persson BM, Ekelund L, Lovdahl R, Gunterberg B. Favourable results of acrylic cementation for giant cell tumors. Acta Orthop Scand 1984;55:209–214. 11. Scoville WB, Palmer AH, Samra K, Chong G. The use of acrylic plastic for vertebral replacement or fixation in metastatic disease of the spine. Technical note. J Neurosurg 1967;27:274–279.
275
276
Chapter 12 Percutaneous Vertebroplasty 12. Sundaresan N, Galicich JH, Lane JM, Bains MS, McCormack P. Treatment of neoplastic epidural cord compression by vertebral body resection and stabilization. J Neurosurg 1985;63:676–684. 13. Galibert P, Deramond H, Rosat P, Le Gars D. [Preliminary note on the treatment of vertebral angioma by percutaneous acrylic vertebroplasty.] Neurochirurgie 1987;33:166–168. 14. Lapras C, Mottolese C, Deruty R, Lapras C Jr, Remond J, Duquesnel J. [Percutaneous injection of methyl-methacrylate in osteoporosis and severe vertebral osteolysis (Galibert’s technic).] Ann Chir 1989;43:371–376. 15. Debussche-Depriester C, Deramond H, Fardellone P, Heleg A,. Sebert JL, Cartz L, Galibert P. Percutaneous vertebroplasty with acrylic cement in the treatment of osteoporotic vertebral crush fracture syndrome. Neuroradiology 1991;33(Suppl):149–152. 16. Jensen ME, Evans AJ, Mathis JM, Kallmes DF, Cloft HJ, Dion JE. Percutaneous polymethylmethacrylate vertebroplasty in the treatment of osteoporotic vertebral body compression fractures: technical aspects. Am J Neuroradiol 1997;18:1897–1904. 17. Mathis JM, Barr JD, Belkoff SM, Barr MS, Jensen ME, Deramond H. Percutaneous vertebroplasty: a developing standard of care for vertebral compression fractures. Am J Neuroradiol 2001;22:373–381. 18. Riggs BL, Melton LJ. The worldwide problem of osteoporosis: insights afforded by epidemiology. Bone 1999;17:505S–511S. 19. Ray NF, Chan JK, Thamer M, Melton III LJ. Medical expenditures for the treatment of osteoporotic fractures in the United States in 1995: report from the National Osteoporosis Foundation. J Bone Miner Res 1997;12(1):24–35. 20. Cooper C, Atkinson EJ, O’Fallon WM, Melton LJ. Incidence of clinically diagnosed fractures: a population based study in Rochester, Minn. J Bone Miner Res 1992;7:221–227. 21. Bostrom MP, Lane JM. Future directions. Augmentation of osteoporotic vertebral bodies. Spine 1997;22:38S–42S. 22. Chiras J, Depriester C, Weill A, Sola-Martinez MT, Deramond H. [Percutaneous vertebral surgery. Technics and indications.] J Neuroradiol 1997;24:45–59. 23. Cortet B, Cotten A, Boutry N, Flipo RM, Duquesnoy B, Chastanet P, Delcambre B. Percutaneous vertebroplasty in the treatment of osteoporotic vertebral compression fractures: an open prospective study. J Rheumatol 1999;26:2222–2228. 24. Cotten A, Boutry N, Cortet B, Assaker R, Demondion X, Leblond D, Chastanet P, Duquesnoy B, Deramond H. Percutaneous vertebroplasty: state of the art. Radiographics 1998;18:311–323. 25. Cotten A, Deramond H, Cortet B, Lejeune JP, Leclerc X, Chastanet P, Clarisse J. Preoperative percutaneous injection of methyl methacrylate and N-butyl cyanoacrylate in vertebral hemangiomas. AJNR Am J Neuroradiol 1996;17:137–142. 26. 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 percentage of lesion filling and the leakage of methyl methacrylate at clinical follow-up. Radiology 1996;200:525–530. 27. 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 surgical cement in 20 patients. AJR Am J Roentgenol 1999;173:1685–1690. 28. Deramond H, Depriester C, Galibert P, Le Gars D. Percutaneous vertebroplasty with polymethylmethacrylate. Technique, indications, and results. Radiol Clin North Am 1998;36:533–546. 29. Deramond H, Depriester C, Toussaint P. [Vertebroplasty and percutaneous interventional radiology in bone metastases: techniques, indications, con-tra-indications.] Bull Cancer Radiother 1996;83:277–282. 30. Deramond H, Depriester C, Toussaint P, Galibert P. Percutaneous vertebroplasty. Semin Musculoskelet Radiol 1997;1:285–295.
Conclusion 31. Deramond H, Galibert P, Debussche C, Pruvo J, Heleg A, Hodes J. Percutaneous vertebroplasty with methylmethacrylate: technique, method, results [abstract]. Radiology 1990;177P:352–352. 32. Dousset V, Mousselard H, de Monck d’User L, Bouvet R, Bernard P, Vital JM, Senegas J, Caille JM. Asymptomatic cervical haemangioma treated by percutaneous vertebroplasty. Neuroradiology 1996;38:392–394. 33. Mathis JM, Eckel TS, Belkoff SM, Deramond H. Percutaneous vertebroplasty: a therapeutic option for pain associated with vertebral compression fracture. J Back Musculoskelet Rehabil 1999;13:11–17. 34. Silverman SL. The clinical consequences of vertebral compression fracture. Bone 1992;13(Suppl):27–31. 35. Kado DM, Browner WS, Palermo L, Nevitt MC, Genant HK, Cummings SR. Vertebral fractures and mortality in older women: a prospective study. Study of Osteoporotic Fractures Research Group. Arch Intern Med 1999;159:1215–1220. 36. Maynard AS, Jensen ME, Schweickert PA, Marx WF, Short JG, Kallmes DF. Value of bone scan imaging in predicting pain relief from percutaneous vertebroplasty in osteoporotic vertebral fractures. Am J Neuroradiol 2000;21(10):1807–1812. 37. O’Brien JP, Sims JT, Evans AJ. Vertebroplasty in patients with severe vertebral compression fractures: a technical report. Am J Neuroradiol 2000;21(8):1555–1558. 38. Peh WC, Gilula LA, Peck DD. Percutaneous vertebroplasty for severe osteoporotic vertebral body compression fractures. Radiology 2002;223:121–126. 39. Jensen ME, Dion JE. Percutaneous vertebroplasty in the treatment of osteoporotic compression fractures. Imaging Clin North Am 2000;10(3):547–568. 40. Belkoff SM, Maroney M, Fenton DC, Mathis JM. An in vitro biomechanical evaluation of bone cements used in percutaneous vertebroplasty. Bone 1999;25:23S–26S. 41. Jasper L, Deramond H, Mathis JM, Belkoff SM. Material properties of various cements for use with vertebroplasty. J Mater Sci Mater Med 2002;13:1–5. 42. Jasper LE, Deramond H, Mathis JM, Belkoff SM. The effect of monomer-topowder ratio on the material properties of cranioplastic. Bone 1999;25:27S–29S. 43. Norden CW. Antibiotic prophylaxis in orthopedic surgery. Rev Infect Dis 1991;10:S842–S846. 44. Barr JD, Barr MS, Lemley TJ, McCann RM. Percutaneous vertebroplasty for pain relief and spine stabilization. Spine 2000;25:923–928. 45. Gangi A, Kastler BA, Dietemann JL. Percutaneous vertebroplasty guided by a combination of CT and fluoroscopy. Am J Neuroradiol 1994;15:83–86. 46. Mathis JM, Petri M, Naff N. Percutaneous vertebroplasty treatment of steroidinduced osteoporotic compression fractures. Arthritis Rheum 1998;41:171–175. 47. Wong W, Mathis JM. Commentary: is intraosseous venography a significant safety measure in performance of vertebroplasty? J Vasc Interv Radiol 2002;13:137–138. 48. Gaughen JR, Jensen ME, Schweickert PA, Kaufmann TJ, Marx WF, Kallmes DF. Relevance of antecedent venography in percutaneous vertebroplasty for the treatment of osteoporotic compression fractures. Am J Neuroradiol 2002;23:594–600. 49. Belkoff SM, Mathis JM, Jasper LE, Deramond H. The biomechanics of vertebroplasty: the effect of cement volume on mechanical behavior. Spine 2001;26:1537–1541. 50. Tohmeh AG, Mathis JM, Fenton DC, Levine AM, Belkoff SM. Biomechanical efficacy of unipedicular versus bipedicular vertebroplasty for the management of osteoporotic compression fractures. Spine 1999;24:1772–1776. 51. Zoarski GH, Snow P, Olan WJ, Stallmeyer M, Dick B, Hebel J, De Deyne M. Percutaneous vertebroplasty for osteoporotic compression fracture: quantitative prospective evaluation of long-term outcomes. J Vasc Interv Radiol 2002;13:139–148. 52. Jensen ME, McGraw JK, Cardella JF, Hirsch JA. Position Statement on percutaneous vertebral augmentation: A consensus statement developed by the American Society of Interventional and Therapeutic Neuroradiology, Society of Interventional Radiology, American Association of Neurological Surgeons/
277
278
Chapter 12 Percutaneous Vertebroplasty Congress of Neurological Surgeons, and American Society of Spine Radiology. J Vasc Interv Radiol 2007;18:325–330. 53. Lindsay R, Silverman SL, Cooper C, Hanley DA,Barton I, Broy SB,Licata A, Benhamou L, Geusens P, Flowers K, Stracke H, Seeman E. Risk of new vertebral fracture in the year following a fracture. JAMA 2001;285:320–323. 54. Voormolen MHJ, Lohle PNM, Juttmann JR, van der Graaf Y, Fransen H, Lampmann L. The risk of new osteoporotic vertebral compression fractures in the year after percutaneous vertebroplasty. J Vasc Interv Radiol 2006;17:71–76. 55. Scroop R, Eskridge J, Britz GW. Paradoxical cerebral arterial embolization of cement during intraoperative vertebroplasty: case report. Am J Neuroradiol 2002;23:868–870. 56. Weill A, Chiras J, Simon J, Rose M, Sola-Martinez T, Enkaoua E. Spinal metastases: indications for and results of percutaneous injection of acrylic surgical cement. Radiology 1996;199:241–247. 57. Kallmes DF, Comstock BA, Heagerty PJ, et al. A randomized trial of vertebroplaseoporotic spinal fractures. NEJM 2009,361:569–579. 58. Buchbinder R, Osborne RH, Ebeling PR, et al. A randomized trial of vertebroplasty for painful osteoporotic vertebral fractures. NEJM 2009, 361:557–568. 59. Gray LA, Jarvik JG, Heagerty PJ, et al. Investigational vertebroplasty efficacy and safety trial (INVEST): a randomized controlled trial of percutaneous vertebroplasty. BMC Musculoskel Disorders 2007, 8:126–134. 60. Mathis JM, Barr JB, Belkoff SM, et al. Percutaneous vertebroplasty: a developing standard of care for vertebral compression fractures. AJNR 2001, 22:373–381. 61. Eippert F, Finsterbusch J, Bingel U, Buchel C. Direct evidence for spinal cord involvement in placebo analgesia. Science 2009, 326:404.
13 Implanted Drug Delivery Systems Lisa Jo Stearns
Introduction Implanted Drug Delivery Systems are recommended to patients whose pain is poorly controlled with oral or parenteral medication or to those who experience drug related toxicities. Administration of intrathecal opioids and adjuvant medications allows reduction of up to 200% of oral or parenteral pain medication.1–3 In 1979, Wang and colleagues demonstrated the efficacy of intrathecal morphine in cancer patients.4 Since then, the adoption and use of intrathecal therapies have rapidly developed. Many medications have been trialed intrathecally and accepted into clinical practice with and without scientific support. In 2000, Hassenbusch and Portenoy gathered a team of clinical experts to establish best practice guidelines based on scientific evidence of safety, efficacy, and/or broad clinical usage. The panel surveyed implanting physicians to determine methods of drug selection and drug combination as well as clinical decision making with ongoing intrathecal therapy. The panel examined evidence for safety and efficacy of individual and combinations of medications utilized in intrathecal therapy. The panel established intrathecal medication utilization guidelines based on literature reviews, physician surveys, and evidence of widespread clinical usage. These guidelines were first published in 2000,5 and updated in 20036 and 2007.7 Axial delivery systems have not changed significantly over the last 10 years. The first models, described by Coombs et al.8 and Poletti et al.,9 utilized an implantable reservoir that delivered a bolus into the epidural space with compression. Other early models utilized an external catheter or an access port to deliver medication to the epidural or intrathecal space. Externalized systems are complicated with discomfort and inconvenience for the patient and a higher risk of infection,10 and have not been recommended for patients with a life expectancy greater than 3 months.11
From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_13, © Springer Science + Business Media, LLC 2010
279
280 Charter 13 Implanted Drug Delivery Systems
Current constant flow rate internal systems produced by Arrow International Inc. (Reading, PA) and Medtronic Inc. (Minneapolis, MN) were developed for the delivery of intravascular chemotherapeutic agents and intrathecal pain medications. These systems are utilized with a fixed or stable medication regimen. The constant rate systems deliver a predetermined rate of infusate. Several products are currently on the market and are all based on a similar mechanism of delivery. Figure 13.1 depicts a constant flow pump with a titanium shell and divided into two chambers: a compressed gas chamber separated from the medication chamber by a compression bellows. The compression bellows has a fluorocarbon propellant gas that expands as it becomes warmed and expels medication at a constant rate. The drug delivery of these devices is affected by altitude and change in ambient temperature. Medtronic produced the first programmable pump in 1988. It was developed to allow for the capability of changes in rate and volume of medication delivery for both primary and secondary infusion modes. These infusion modes may be programmed to be constant, variable, or have periodic bolusing features. The programmable pump allows the clinician to change the dose of medication without changing the prescription of the compound. While several versions of this pump have been released, the basic principles of drug delivery have not changed. The pump is filled through a self-sealing septum into a collapsible reservoir (Figure 13.2). The current models hold 18, 20, and 40 mL of medication. The models consist of a lithium battery module, an electronic module, and a peristaltic pump motor that pulls medication from the pump reservoir by compressing the internal tubing and moving the flow forward out of the catheter (Figure 13.2). The rate of delivery of medication is determined by the speed of the turning of the rotors set by the programming of the microprocessor. A radiofrequency transmitter allows the clinician to program the pump to the desired settings. The pump can be programmed to deliver a constant rate, a variable rate, or intermittent boluses given throughout a 24-h
Figure 13.1. Isomed factory preset constant flow rate infusion pump (Used with permission of Medtronic, Inc., © 2009, Minneapolis, MN. Note that the IsoMed Pump is no longer in production or distribution.)
Introduction
A
B
Figure 13.2. (A). SynchroMed® EL programmable pump demonstrating the self sealing septum, collapsible reservoir, and peristaltic pump motor. (Note that SynchrodMed® EL Pump has been superceded by the SynchroMed® II Drug Pump.); (B). SynchromedMed® II Drug Pump. (A, B: Used with permission of Medtronic, Inc., © 2009, Minneapolis, MN.)
period. The life of the lithium battery varies with the electrical activity of the device. The SynchroMed® II (Medtronic Inc., Minneapolis, MN) models report predicted battery life at programming. Medtronic also has developed a Patient Therapy Manager that allows the physician to pre-program boluses that the patient may administer for periods of pain exacerbations. The pump is filled by accessing the reservoir fill port. The old medication is discarded, and new medication is infused. A 22 mm filter protects the pump’s internal tubing and patient from bacteria and other contaminants. It does not protect against pyrogens, viral particles, or other small contaminants. After programming, the medication passes from the reservoir to the internal pump tubing, through the external catheter, and into the intravascular, epidural, or intrathecal space. The volume of the internal pump tubing is approximately 0.2 mL and should be added to the catheter volume to calculate the total catheter volume. This value is needed to accurately calculate bridge bolus for prescription changes. The current programmers are almost pocket-sized and have a variety of programming options. One to five medications may be entered into the data, but dose is only based off of the first listed or primary medication. The remaining medication concentrations should be calculated keeping in mind what the therapeutic range will be of the primary medication and should have a margin of safety for the others in case dose escalations or reductions are necessary. Once the dose of the primary medication is set, the others are calculated on a ratio basis by the software package. The newer programmers have software designed to accept changes in the primary medications to allow accurate bridge bolusing between the old and the new prescriptions.
281
282 Charter 13 Implanted Drug Delivery Systems
Personal Therapy Manager The Personal Therapy Manager (PTM) is a handheld device utilized to deliver physician programmed boluses with the SynchroMed II® (Medtronic Inc., Minneapolis, MN) infusion system. It is helpful for controlling breakthrough pain without additional oral medications. This device is similar to the demand dose with patient controlled analgesia. The clinician sets the parameters for the bolus amount, the lockout period, and the number of allowed boluses per day. This device allows the patient to become more actively involved in their therapy, in which unpredictable increases in his or her pain can be treated with additional doses of medication through the pump. Both malignant and non-malignant pain patients may benefit from the Personal Therapy Manager. Opiate dosing is similar to oral rescue dosing at 10–25% of the daily dose. Rate of bolus delivery is dependent on the physician’s desire to increase spread or control spread of the medication in the intrathecal space. Bolus delivery rate is also dependent on the infusate and the catheter tip position. Bolusing bupivacaine or clonidine too rapidly can have devastating effects. In the author’s clinic, clonidine boluses are limited to 10–25 mcg with a 5 min lockout, and bupivacaine boluses are limited to a 0.5–1 mg bolus over a 5 min period. The first therapeutic bolus is given in the office prior to instituting the PTM. The patient is monitored for 30 min following the bolus. The frequency of boluses is determined by the type of pain and patient dysfunction. End-stage cancer patients may be allowed to bolus every hour to limit need for frequent reprogramming at end of life, while chronic nonmalignant pain patients are typically limited to four boluses in a 24-h period. Patients with ziconotide should not utilize the PTM because of neurological sequelae of bolus dosing with this medication. The utilization of the PTM improves patient satisfaction of therapy by allowing the freedom to control pain without side effects of oral medications.
Intraspinal Drug Delivery Clinic The physician who implants and manages patients with indwelling intrathecal drug delivery devices should have a strong network of supporting individuals to assist in patient care. Accurate scheduling is needed to manage trialing and implant events as well as clinic follow up. Time should be available to provide pre-implant education and answer questions. Post-operatively, the patients will need to be scheduled for pump refills and reprogramming. Vigilance is needed to monitor alarm dates, assess patient function and response to therapy, and alert the physician to possible therapy complications. A refill clinic should have appropriate support staff and providers to fill and program the pumps. The clinic may be a hospital-based outpatient infusion center, an outpatient treatment center, or at the physician’s office. Managing physicians should be readily available in case of difficulty refilling the pump or in case of therapy issues.
Patient Selection
Medications should be prescribed on an individual basis with the latest recommendations of the Polyanalgesic Consensus Conference in mind.7 The physician or allied health care staff should be available after hours to trouble shoot problems that may or may not be associated with the pump. Record access should be available to the patient in case of emergency. It is recommended that patients leave the clinic with a copy of the telemetry from the latest pump readout. The medications utilized, dosages, and refill date should be double checked against the prescription and discussed and reviewed with the telemetry before the patient leaves the clinic.
Patient Selection Implanted drug delivery systems are indicated for patients with either chronic malignant or non-malignant pain. Chronic pain is defined as pain having lasted longer than 3 months. A candidate for intrathecal therapy has chronic intractable pain unresolved with conservative measures, has failed other treatments, and/or has intractable side effects to medications. Another interventional pain therapy, spinal cord stimulation, is considered less invasive and has fewer long terms risks than implanted drug delivery systems. In appropriate patients, this therapy should be trialed first. Goals of intrathecal therapy are established prior to the trial and should include a reduction in subjective pain greater than 50%, improved function, and diminished systemic effects from pain medication. The patient should have psychological clearance before trialing for pump implantation. The patient should be assessed for any contraindications to implantation, undergo preimplant education, understand short and long term risks, and have all questions answered prior to trialing and implantation of the device. Contraindications to implantation are categorized as absolute or relative (Table 13.1). In absolute conditions, correction of the underlying condition may allow for therapy consideration in the future. Malnourished and emaciated patients may be implanted if the site is carefully selected prior to implantation and monitored for breakdown. Emaciated cancer patients in particular benefit from intrathecal therapy and frequently gain weight following the procedure (Figure 13.3). Patients in whom normalizing coagulation is not an option because of cardiac arrhythmias or clotting disorders may be managed with low molecular weight heparin until prothrombin times normalize. Anticoagulation may safely be restarted 12 h post surgery.12 In addition to the above criteria, the candidate with malignant pain should have a life expectancy of 3 months or longer. If performance scores are diminished and prognosis is guarded, a long-term intrathecal catheter should be placed in lieu of an implanted drug delivery device. If the patient’s functional status improves, the catheter may be converted to an implanted system at a later date. Financial counseling regarding the cost of the trial, the procedure, and maintenance should occur prior to implantation. The patient should be made aware of additional medication co-pays and facility
283
284 Charter 13 Implanted Drug Delivery Systems
Table 13.1. Exclusion and inclusion criteria for intraspinal opioids Exclusion criteria Absolute Aplastic anemia Systemic infection Known allergies to implant materials (pump and catheter) Active drug abusers Psychosis Relative Malnourished, emaciated patients Occult infection Recovering drug addicts Dementia Unable to disrupt anticoagulation therapy Socioeconomic problems Lack of access to pump managers Inclusion criteria Pain type and generator appropriate Psychologically Stable and compliant with treatment plan Demonstrated responsiveness to spinal medication trial Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004.
Figure 13.3. Emaciated cancer patient with implanted SynchroMed® II (Medtronic Inc., Minneapolis, MN) pump.
Continuous Epidural or Intrathecal Infusion Trials
fees, which may occur at reprogramming and refill appointments. The implanting physician should be aware of Medicare guidelines regarding approved ICD-9 codes for intrathecal therapy reimbursement.
Patient Screening Once a patient has been identified as a suitable candidate for intrathecal therapy, a screening trial is scheduled. No specific screening trial method has proven superior to others.13 The clinician should decide which method is best for his particular practice. The goal of screening is to determine adequate response to intra-axial opioids or adjuvants with 50% diminished pain, increased function, and lower toxicities of administered medications.
Single Intrathecal Bolus Dosing A single dose of medication, typically an opioid (morphine, hydromorphone, or fentanyl), is injected into the intrathecal space. This may be done with or without fluoroscopy. Some clinicians trial ziconotide in this fashion as well. Following intrathecal bolusing, the patient is monitored for pain relief and side effects. This technique increases the incidence of urinary retention and other medication toxicities. In the opiate-naïve patient, fentanyl should be selected for trialing secondary to its lipid solubility and short half life. If morphine and hydromorphone are trialed, the patient should be observed for respiratory depression.7,14 Pain typically diminishes within a few hours and may have lasting effects for 24 h. Many physicians believe that this approach cannot exclude placebo effect, yet few physicians believe in blinded placebo trials.15
Continuous Epidural or Intrathecal Infusion Trials Many physicians prefer the continuous intrathecal infusion to epidural infusion trial as it best predicts the pain relief and side effect profiles the patient will experience with pump implantation. The trial may last days to months depending on the clinician and functional goals established for the patient. If the system is to remain in place for longer than a few days or if the patient will be discharged home or to a rehabilitation facility, the catheter should be tunneled to prevent easy dislodgement and to lessen the risk of infection. Intrathecal infusion best predicts actual pump medication delivery rates and toxicities of medications. If the epidural infusion is preferred, the clinician should reduce the infusion daily rate by 10% for intrathecal conversion on pump implantation. Continuous catheters should be placed under fluoroscopic guidance to demonstrate appropriate placement and to document catheter tip location. In a debilitated cancer patient or in a non-malignant pain patient for whom an extended trial is desired, the clinician may decide to implant the trial catheter surgically. An implanted long-term catheter reduces
285
286 Charter 13 Implanted Drug Delivery Systems
the risk of skin infection, and the position of the exteriorized catheter can be placed for comfort and ease of care. The patient is brought to the procedure area and placed in position. The back and flank are prepped to allow for tunneling and exteriorization of the catheter. A small incision is made in the back. The needle is placed into the intrathecal space below L1 or the level of the conus medullaris. The catheter is threaded through the needle to the appropriate level under fluoroscopy. This level is determined by the level of the pain generator and what medications will be utilized in the infusion. The needle is withdrawn, and the catheter tunneling device is utilized to tunnel the catheter laterally. The author makes a second incision in the lateral abdomen and tunnels the catheter to this wound. The catheter is anchored to the fascia in both wounds. The catheter is then externalized two inches from the anterior wound and connected to the infusion port. Alternatively, the catheter may be connected at this incision to a subcutaneously placed port to reduce the risk of catheter fracture and infection. Management of maintenance opiates during the trial period is controversial. Some clinicians believe in complete detoxification from opiates prior to the trial to maximize response and limit dose.16 Krames believes that medication should be converted to an equianalgesic intrathecal equivalent of 50% of the pre-trial oral dose and that the remaining oral dose be withdrawn by 20% per day.17 The author prefers to stop all extended release opiates for 72 h during the continuous infusion trial. The patient has access to immediate release and parenteral opiates during this time. If a prolonged trial is desired, the extended release opioids are restarted at 50% of their pretrial dose and titrated off over the following week. At the conclusion of the trial, the maintenance medications are restarted at 50–100% of the patient’s usual dose.
System Implantation Prior to the procedure, all questions regarding the procedure, the device, and maintenance should be answered. The desired location for the pump should be determined by the implanting physician and the patient. This location should be communicated with the circulating nurse and the anesthesiologist for proper positioning. The pump is most frequently placed in the left or right lateral abdomen with anatomical constraints of the ribcage, the iliac crest, and pubic symphysis. The implanter should also be aware of the patient’s preferred beltline to avoid future dissatisfaction of pump location site. Some patients or clinicians prefer to implant the pump over the iliac crest. This position is appropriate in patients with sufficient subcutaneous fat and supporting muscle. In the debilitated and malnourished patient, however, this site is prone to breakdown. Intravenous antibiotics should be given within 1 h of incision.18 The method of anesthesia for the procedure is determined among the surgeon, the anesthesiologist, and the patient. Provided that the dura entry site is planned below the level of the conus medullaris, there is little added risk to a general anesthetic. The use of muscle relaxation is
System Implantation
not recommended with general anesthesia, as paresthesias are elicited in motor response when floating the catheter in the spinal canal. Position of the patient is determined by pump location site. If the patient and the surgeon desire buttock implantation, the patient may be prone or in the lateral decubitus position. If the abdomen is selected, then the patient is placed in the lateral decubitus position with the pump pocket side upward. Once the patient is positioned, pressure points are padded, and an axillary roll is placed, if appropriate. Following a sterile prep and drape, the area of the catheter insertion is addressed. A one to one and a half inch incision is placed midline between L2 and S1. Electrocautery is used for hemostasis, and the tissues are dissected to the supraspinous ligament. A suture is placed just lateral to the spinous process to anchor the catheter. The author prefers to do this prior to the intrathecal needle placement to avoid possible catheter damage. The 15 g Tuohy needle is used to access the intrathecal space using a flat paramedian technique. The dura should be entered below the level of the conus medullaris. Once the intrathecal space is entered with needle, the catheter is advanced under fluoroscopy to the appropriate level. If the catheter begins to coil in the intrathecal space, both the needle and the catheter should be withdrawn and the procedure is repeated. One should not attempt to pull the catheter back through the needle if any resistance is encountered. Occasionally, turning the stylette may help to drive the catheter around the area of resistance. If desired, a purse string suture is placed around the needle entrance site to minimize the risk of spinal leak. Then the needle and stylette are removed, and Isovue M® (Bracco Diagnostics, Princeton, NJ) or Omnipaque™ (GE Healthcare, Chalfont St. Giles, United Kingdom) is injected. The flow and dispersion of dye are observed under fluoroscopy. If pooling of the dye is noted, the catheter should be moved and the dye study repeated. Areas of diminished cerebrospinal fluid circulation may result in drug pooling, intermittent overdosing, or underdosing or increase in the risk of inflammatory mass formation. After catheter placement, the pump pocket is addressed. Using a scapel, a three inch incision (four finger-widths) is made, and electrocautery is used for hemostasis. The tissues are dissected to the fascial plane. The pocket should be four inches long by three inches wide. Four anchoring sutures to secure the pump to the fascial plane are placed. An antibiotic-soaked lap sponge is placed into the pocket while the posterior incision is addressed. Using a malleable tunneling device, the catheter is tunneled from the posterior to the anterior incision. The catheter is anchored to the posterior fascia utilizing the previously placed suture. Excess catheter is trimmed and connected to the pump utilizing the catheter connection device provided with the catheter. A suture-less connector is now available. The pump is placed into the pocket, and the catheter tip is oriented for comfort away from bony processes or the natural waistline. Excess catheter is placed behind the pump, and the pump is secured to the posterior fascia utilizing the previously placed anchoring sutures and the suturing loops on the pump or by suturing to the
287
288 Charter 13 Implanted Drug Delivery Systems
silastic sleeve, if desired. Anchoring is necessary to prevent the pump inverting in the pocket. Both wounds are checked for bleeding and closed with absorbable sutures or staples. This author has noticed a decreased infection incidence when topical skin adhesive Dermabond® (Ethicon Inc, Somerville, NJ) was substituted for Steri-Strips™ (3M, St. Paul, MN). Completion of an implant checklist is recommended to document the procedure for dictation and for accurate records of the procedure since they are needed at a later date (Table 13.2).
Table 13.2. Implant check list Implant date: Patient name: Date of birth: Pain diagnosis: Indication for device use:
Malignant pain
Spasticity
Non-malignant pain
Right/left abdomen
Right/left buttock
Other
ICD-9 codes: Pump information: Model number: Serial number: Implant location: Direction of incision:
Transverse
Oblique
Depth of incision:
Subcutaneous
Subfascial
Catheter connection orientation (clock orientation):
3
6
Suture
Staples
Midline
Paramedian
Catheter tip position:
Dorsal
Dye dispersion in CSF:
9
12
Ventral
Right
Left
Slow
Moderate
Rapid
Purse string:
Yes
No
Pump connector:
90°
40°
Pump anchoring: Skin closure: Catheter information: Model number: Lot number: Catheter implanted length: Catheter implanted volume: Needle approach: Skin entry site: Dura entry site: Catheter tip location:
Primary catheter anchor: Secondary catheter anchor: Sutureless
Other
Complications
Outcomes Successful outcomes with intrathecal therapy occur when a patient reports a subjective decrease in pain and demonstrates objective increase in function while decreasing oral opioid usage. Patients with non-malignant pain should continue with physical and behavioral modalities to enhance their pain relief. Many types of pain may be treated with intrathecal therapy, and the clinician should utilize the Polyanalgesic Consensus Conference recommendations to guide therapy.7 Paice and colleagues reviewed data from a retrospective multicenter study utilizing intrathecal morphine infusion for chronic pain. Eighty-two percent of patients reported decreased pain and increased function as reported in activities of daily living. Patients with visceral pain demonstrated the greatest improvement.13 All types of cancer pain respond to intrathecal therapy. This therapy is recommended for cancer patients whose pain is uncontrolled with conservative therapy or for whom analgesic related toxicities have developed.19 In a multicenter, prospective randomized trial, Smith and colleagues demonstrated that patients receiving intrathecal therapy along with comprehensive medical management had reduced pain, fewer common drug toxicities, and improved survival compared to patients managed with comprehensive medical management only.20
Complications Intrathecal drug delivery complications can be divided into three types: surgery related, device-related, and drug-related. Surgical Complications There are five predominate complications from surgical implantation of intrathecal drug delivery devices: neurological injury, bleeding, seroma formation, infection, and cerebrospinal leak. Neurological Injury Neurological injury can occur with spinal needle placement or with catheter advancement. Needle induced injury occurs when the needle comes into contact with nerve roots or with the spinal cord. The needle should always be introduced below the level of the cauda equina. When performed under local anesthesia, the patient will complain of a paresthesia when a nerve is brushed with the needle. The needle should be removed and a new approach is selected. Nerve roots are less likely to be injured with a flat paramedian needle approach. If general anesthesia is selected by the patient, the risks of neurological injury, while small, should be reviewed. The spinal cord may be damaged with catheter advancement. The catheter should never be advanced against resistance and should be observed under fluoroscopic guidance. Neurological signs of cord injury may not be immediately apparent and will develop progressively with drug delivery. An acute
289
290 Charter 13 Implanted Drug Delivery Systems
decrease in neurological function post-operatively should be immediately evaluated with a CT myelogram or an MRI of the spinal cord. The pump should be emergently stopped until the cause of neurological compromise has been identified. Neurological injury can be identified immediately post-operatively or over several hours. The patient will complain of increased back pain, progressive changes in sensation and/or motor strength, and loss of bowel or bladder control. Careful post-operative monitoring with frequent neurological checks will help to identify changes early. Patients with post-operative changes in neurological function should have an immediate CT myelogram or MRI of the spine. Neurosurgery should be notified for possible emergent neurosurgical intervention if a hemorrhage is noted. Bleeding Bleeding is an early complication of surgery. The development of a pump pocket hematoma typically occurs within the first few postoperative days. The patient will complain of increased pain at the pump pocket site, and the swelling will be greater than normal. Intraoperative attention to hemostasis and a tight binder over the pump site post-operatively reduce the risk of hematoma formation. Intrathecal or epidural hemorrhage typically occurs at the time of catheter placement. Seroma Formation Seromas can form within the spine or pump pocket for a variety of reasons. Care should be taken during implant to minimize incisions and dissections to make tight pump and catheter pockets. Also, the pump should be firmly anchored to the fascia to minimize movement in the pocket, which increases the incidence of seroma formation. Post-operative vomiting, whether from opioid intolerance or spinal headache, is often the cause of pump dislodgement from the suture loop ties in the author’s center. When it occurs, the patient will present to the clinic with a swollen pump pocket and increased complaints of pain. If the seroma is not treated and the seroma expands, the seroma will migrate along the catheter to the posterior incision when pressures exceed capacity in the abdominal pocket, creating a spine seroma as well. This phenomenon occurs most frequently in patients who are non-compliant with wearing the abdominal binder. When a patient complains of pain or swelling at the pump incision, the clinician should examine the area to rule out infection. If there is no sign of infection on the skin, the seroma should be drained and sent for culture. The fluid is observed to determine if it is bloody, serosanguinous, or colorless. If there is blood and it is difficult to extract, it is likely that the hematoma will resolve on its own with continued pressure from an abdominal binder. If there is fresh blood, the patient should be taken back to surgery to stop the bleeding, which is most likely from an abdominal perforating vein or artery. Serosanguinous fluid can be problematic long-term. After the pocket is drained, a pressure dressing and abdominal binder should be placed to help the wound sclerose. If the seroma is persistent and cultures are negative, chemical sclerosis
Complications
is recommended. This author uses 1–2 g of doxycycline in 25 mL of 0.5% bupivacaine. The seroma is drained, and the above solution is injected, and the abdominal binder is placed tightly over the site. If a seroma is present in both the pump pocket and spine incisions, the dose is split to 5 mL in the spine site and 20 mL in the pump site. The patient will experience increased pain secondary to burning from the solution for 24 h, and the patient will need supplemental pain medication. The binder should continue to be worn for an additional 30 days. If the fluid is colorless, it is likely to be cerebrospinal fluid leaking from the catheter entry site of the spine, from a fractured catheter, or from a dislodged catheter connector from the pump. A catheter dye study will help to identify the problem. If the patient has a spine incision seroma and minimal pump pocket fluid, one must assume that this is either an infection or spinal fluid. The seroma should be carefully prepped and drained using an 18-gauge or larger needle. If the fluid is colorless, it is cerebrospinal fluid. The cerebrospinal fluid is either leaking around the catheter or has a leak in the catheter near the spinal entry site. The patient is given three choices: do nothing and monitor the site, proceed with a blood patch, or undergo surgical revision. A blood patch under fluoroscopy one level below or above the catheter entrance may help to decrease spinal fluid leaking retrograde. If it does not and the seroma is painful, surgical evaluation is recommended. Some people have recurrent spinal seromas which come and go. As long as the skin overlying the area is healthy and the patient has low risk for infection, no immediate treatment needs to be taken. The seroma may or may not resolve over time. In the author’s experience, however, the presence of a posterior seroma increases the incidence of future catheter migration out of the spinal space. Infection Surgical infections are a devastating complication of intrathecal therapy. Not only can the infection be life threatening, but the coinciding loss of hope that occurs with the removal of the device is psychologically challenging. Conditions that may increase infection risk include: malnutrition, diabetes, obesity, poor hygiene, fecal incontinence, and autoimmune disorders. Care should be taken perioperatively to minimize the risks of infection by addressing these conditions when appropriate. Surgically, the physician is able to minimize post-operative infection risk by paying strict attention to sterile technique, decreasing surgical time, and performing proper wound closure. Chronically infected patients with positive nasal cultures for Staphylococcus aureus should be treated perioperatively with mupirocin in addition to perioperative antibiotics. Guidelines for the prevention and management of intrathecal infections have been published.18 The rate of devicerelated infections is low and usually involves the pump pocket site.18 Most common organism is Staphylococcus epidermidis. Infected wounds should be assessed for severity. Superficial infections may be treated conservatively with antibiotics, wound care, and careful monitoring. If a seroma is present in the pocket and the wound is erythematous, or if other frank signs of infection are noted, the seroma should be drained and cultured to determine the cause of the infection prior to initiation
291
292 Charter 13 Implanted Drug Delivery Systems
of empiric antibiotics and explant. It is virtually impossible to save the implant with a wound that has an exudate. Severe infections need to be treated with surgical removal of the device, wound debridement, and intravenous antibiotics until cultures determine antibiotic sensitivity. A history of infection with meningitis will likely cause a delay of reimplantation of 6–8 weeks. Reimplantation should be in conjunction with the recommendations of an infectious disease specialist. Cerebrospinal Fluid Leak Patients who develop a post-dural puncture headache following the procedure are managed conservatively with bed rest, fluids, and caffeine. Fluids and caffeine may be administered orally or intravenously. Hydration should be encouraged to urination every 2 h. Patients should be encouraged to lay flat for 48 h to minimize further spinal fluid leakage and reduce the risk of nausea and vomiting secondary to decreased cerebrospinal fluid. If nausea and vomiting are problematic, promethazine 25 mg, administered intravenously or as suppositories, should be utilized every 6 h until the patient is able to hold down liquids and take oral medication. Promethazine is preferred at the author’s clinic for its sedating properties and synergism with opiates to allow the patient to tolerate laying flat. If the patient is not able to drink fluids for 6 h, IV hydration is recommended with concurrent caffeine infusions. If the patient has no previous history of hypertension or palpitations, 500 mg of caffeine in 250 mL of normal saline is given every 4 h times 24 h, then as needed. At that time, the author suggests to continue with the following oral regimen: the patient should consume one to two caffeinated beverages or two caffeine pills every 4 h while awake. If the headache is persistent and severe (lasting longer than 7 days), the patient is unable to stop vomiting, or the patient is unable to drink fluids and take caffeine, epidural blood patch is recommended. If blood patch fails to resolve the problem, a second blood patch may be performed, but catheter evaluation with fluoroscopy should be performed to rule out catheter dislodgement from the pump and/or accidental catheter tear during surgery. If the patient fails more conservative efforts and the problem is debilitating, then surgical exploration of the spine wound is recommended to look for catheter tears and/or dislodgement from the pump. If the catheter has dislodged from the pump, a seroma is usually present in the pump wound. Severe nausea and vomiting associated with post-dural puncture headaches or severe cerebrospinal fluid loss can result in meningeal bleeding causing life-threatening neurological complications if left unrecognized. All patients with post-dural puncture headache complaints should be closely monitored. Device-Related Complications Device-related complications may involve problems associated with the pump or the catheter. Pump-related problems may be related to the self-sealing port or the reservoir, electronic failure, battery failure, rotor failure, or movement of the pump. The self-sealing septum of the port may become compromised
Complications
if a coring needle is used to access the reservoir. Overfilling or overpressurization of the reservoir may result in changes in the accuracy of medication delivery. Non-programmable and programmable models prior to the SynchroMed® II (Medtronic Inc., Minneapolis, MN) decrease drug delivery rates as the refill interval approaches. This is secondary to maximum expansion of the freon gas against the bellows. Patients with these older pumps should have anticipated shortened refill intervals. Electronic failure may result from exposure to a high energy field and may result in a pump that is unable to receive reprogramming instructions from the telemetry unit. Following exposures to high energy fields, the pump should be interrogated and status logs are reviewed. Battery failure in programmable pumps varies from 5 to 7 years. In models prior to the SynchroMed® II (Medtronic Inc., Minneapolis, MN), the pump has approximately 30 days after the low battery alarm begins. During this period, however, the patient may experience fluctuations in drug delivery with underdosing of medications. It is recommended that these models be replaced every 5 years if the patient is at risk for withdrawal syndromes (baclofen, higher dose opioids). The SynchroMed® II (Medtronic Inc., Minneapolis, MN) models report battery life with each programming session. Battery life is diminished more rapidly with high flow states or exposure to a high energy field. Radiation therapy near the pump can drain the battery unless the pump is appropriately shielded with lead. Exposure to a magnetic field greater than 2 Tesla may reset the magnetic axis of the pump. Therefore, following any exposure to high energy fields or the initiation of any direct or indirect radiation therapy, the pump should be checked for electronic failure or battery drain. Pump rotor stalls are more frequent in the systems older than the SynchroMed® II (Medtronic Inc., Minneapolis, MN). If the patient is experiencing therapy failure or increased pain, the pump should be interrogated and the event logs examined. If the logs do not point to an electronic or battery problem, a rotor study should be conducted. Under fluoroscopy, a small bolus is given to the patient. The pump is monitored for 45 s. The rotors should change configuration during the bolus. In the EL model, a bar is seen to rotate 90°. In the SynchroMed® II (Medtronic Inc., Minneapolis, MN), apertures will appear to open and close with the movement of the rotors. If the patient has a normal pump rotor study, then the pump reservoir should be drained to compare actual versus calculated volumes remaining. Finally, the catheter should be examined for problems. Movement of the pump within the pocket may result in catheter occlusion or inability to fill the reservoir. The patient may report an episode of severe pain followed by swelling and seroma development in the pump pocket. The pump may freely flip in the pocket and result in stretching, kinking, or dislodgement of the catheter. If the pump is inverted in the pocket, the clinician will hear a hollow sound (similar to tapping on an empty can) as the needle hits the reservoir. If the pump is upright, the sound is more dull and full sounding.
293
294 Charter 13 Implanted Drug Delivery Systems
Revision of the pump pocket is often necessary. If the patient’s pump has recently flipped or the reservoir volumes are accurate on refill, conservative treatment with an abdominal binder may help to hold the pump in place while a new scar tissue capsule forms. The patient should be instructed not to touch or manipulate the pump in the pocket, as this may result in catheter occlusion from twisting. Catheter-related problems may be secondary to obstruction, kinking, fracture, separation from the pump, or development of an inflammatory mass at the catheter tip. Evaluation of a catheter-related problem begins with an accurate history and physical. If the patient has increased complaints of pain, neurological dysfunction, or seeming lack of medication efficacy when previously stable, an MRI of the spine both at the dermatomal distribution of the new pain and of the catheter tip should be ordered. If no physiologic changes are noted, then the clinician should proceed with a catheter dye study. Catheter dye studies may be performed under fluoroscopy or with nuclear medicine. Fluoroscopy is readily available and may identify the problem early. Nuclear medicine studies take several days but are more specific for microfractures of the catheter. Ultimately, treatment of catheter-related problems results in removal and replacement of the existing catheter. With the use of intrathecal polyanalgesic therapy, this author recommends replacing the catheter with the pump when the battery life has expired. If an inflammatory mass is noted on MRI, the clinician has two options: stop the therapy and infuse saline, or move the catheter out of the mass. Surgical excision of the mass is not appropriate, as this will resolve on its own. The clinician may either pull the catheter down to a lower level and dilute the infusate by 50% or replace the catheter. If there is any resistance while attempting to remove the catheter, it should be ligated and left in place to be removed at a later date. The clinician may float a second catheter into the spinal space, providing that the tip is two levels from the mass. Typically, a patient’s pain will rapidly diminish following discontinuation of the infusion into the mass. Whenever a patient complains of increased pain over subsequent clinic visits, inflammatory mass should be suspected. Allergic Reaction to the Pump and Catheter Persistent erythema, itching, and swelling around the pump or catheter sites may be noted. After infection has been ruled out, consultation with an allergist may confirm an actual problem. Low grade infections may also mimic allergies. Technical services of the device manufacturer should be called for recommendations. Infusate-Related Complications Infusate-related complications involve programming errors, drug compounding errors, and technician errors. Programming errors include incorrect data in regards to medication concentration, dosage, and bridge bolusing. Vigilance must be taken to ensure that the correct medication is infused into the pump and entered into the programmer with the correct concentration. Following reprogramming, it is recommended that a second individual review
Complications
the telemetry against the prescription and drug invoice before the patient leaves the clinic. Failure to do so may result in patient overdose and/or death. With utilization of the current programmers and the SynchroMed® II (Medtronic Inc., Minneapolis, MN), bridge bolusing is calculated by the programmer software. In older models, when switching drugs with variable potencies, the programming software was unable to distinguish between the different drugs. The clinician should know the implanted catheter volume and add this volume to the pump tubing volume of 0.2 mL to estimate the bridge bolus. This is the formula to calculate the bridge bolus by hand: Newdrug concentration × total catheter volume = bolusamount Olddrug concentration × total catheter volume × 24 = bolus timein hours Olddrug dailyrate (mg /day ) The United States Pharmacopoeia and the American Society of Health System Pharmacists have issued statements on the preparation of sterile products that have clinical, legal, and practical significance. Physicians who are utilizing compounded products for intrathecal therapy should only utilize pharmacies that follow these guidelines. The physician should have a strong working relationship with the compounding pharmacist, as compounding error could result in withdrawal symptoms, overdose, and death. When the therapy seems to produce adverse effects or lacks efficacy, the physician needs to have a systematic approach to problem-solve and know the therapeutic options. If the patient no longer seems to be getting therapeutic benefit from a previously stable treatment dose of medication, evaluation needs to consider progression of disease, catheter problems, and pump malfunctions. CT scans, CT myelograms, and MRIs of the spine may help to diagnose changes in the patient’s underlying condition. If these are negative for changes in disease state or if positive and additional modalities such as pain blocks are ineffective, the clinician should progress to system evaluation. The technician errors occur with pump refills. The technician may mistake the scar tissue capsule for the port and inject the infusate into the pump pocket. Some pockets have a small chronic seroma that is slightly yellow in color, similar to compounded drug on withdrawal from the pump reservoir. This mistake can result in vascular instability, respiratory depression, and death with complex pump formulas that include clonidine or in the opiate-naïve patient. If a 22-gauge non-coring needle is not used to access the port, the side-port may be accidently accessed. This results in direct cerebrospinal fluid injection of the infusate and will likely result in death if not recognized within minutes of the injection. If the side-port is accessed to clear the drug from the catheter tubing, at least 1 mL of cerebrospinal fluid should be withdrawn before pump bolusing under slow, constant pressure. Accurate calculations are mandatory in pump tubing volume and catheter volume if the technician desires to clear the pump system of old medication. As many as three separate catheter aspirations and
295
296 Charter 13 Implanted Drug Delivery Systems
subsequent catheter blousing events may be needed to clear the pump tubing of old medication. Patients who are suspected in receiving a drug overdose from a programming error, pump pocket infusion, or side-port injection should be admitted to the hospital and possibly the intensive care unit for observation for a minimum of 24 h.
Conclusion Intraspinal drug delivery systems are an alternative option for patients with severe chronic malignant and non-malignant pain. Patient selection, implantation techniques, and drug selection are variables that affect therapy success. The improvement in programmability and patient control features adds to the therapy’s utility. Education, stressing functional outcomes rather than focusing on pain relief will result in reasonable expectations of the therapy by patients. Because of the possibility of long-term complications such as inflammatory mass, this option should be considered after conservative options have been exhausted in the non-malignant pain patient. For the cancer patient, intrathecal drug delivery systems improve pain while reducing drug toxicities and improving function and quality of life. The addition of the Personal Therapy Manager to the system may allow the patient to have pain control without the use of systemic medication.
Acknowledgment We would like to posthumously acknowledge Dr. John C. Oakley as the author of this chapter in the first edition of this book and his influence on the current version of this chapter. References 1. Krames ES. Intraspinal opioid therapy for chronic nonmalignant pain: current practice and clinical guidelines. J Pain Symptom Manage 1996;11:333–352. 2. Smith TJ, Coyne P. How to use implantable drug delivery systems for refractory cancer pain. J Support Oncol 2003;1:73–76. 3. Burton AW, Rajagopal A, Shah HN, Mendoza T, Cleeland CS, Hassenbusch SJ, Arens JF. Epidural and intrathecal analgesia is effective in treating refractory cancer pain. Pain Med 2004;5(3):239–247. 4. Wang JK, Nauss LA, Thomas TE. Pain relief by intrathecally applied morphine in man. Anesthesiology 1979;50:149–151. 5. Hassenbusch SJ, Portenoy RK. Polyanalgesic Consensus Conference 2000. J Pain Symptom Manage 2000;20(2):S1–S50. 6. Hassenbusch SJ, Portenoy RK, Cousins M, Buchser E, Deer T, Du Pen S, Eisenach J, Follett K, Hildebrand K, Krames E. Polyanalgesic consensus conference 2003: an update on the management of pain by intraspinal drug deliveryreport of an expert panel. J Pain Symptom Manage 2004;27(6):540–563. 7. Deer T, Krames ES, Hassenbusch SJ, Burton A, Caraway D, Dupen S, Eisenach J, Erdek M, Grigsby E, et al. Polyanalgesic consensus conference 2007: recommendations for the management of pain by intrathecal (intraspinal) drug delivery: report of an interdisciplinary expert panel. Neuromodulation 2007;10(4):386–389.
Acknowledgment 8. Coombs DW, Saunders RL, Baylor MD, Block AR, Colton T, Harbaugh R, Pageau MG, Mroz W. Relief of continuous chronic pain by intraspinal narcotics infusion via an implanted reservoir. JAMA 1983;250:2336–2339. 9. Poletti CE, Cohen AM, Todd DP, Ojemann RG, Sweet WH, Zervas NT. Cancer pain relieved by long-term epidural morphine with permanent indwelling system for self administration. J Neurosurg 1981;55:581–584. 10. Akahosshi MP, Furuike-McLaughlin T, Enriquez NC. Patient controlled analgesia via intrathecal catheter in outpatient oncology patients. J Intraven Nurs 1988;11:289–292. 11. Bedder MD, Burchiel K, Larson A. Cost analysis of two implantable narcotic delivery systems. J Pain Symptom Manage 1991;6:368–373. 12. Horlocker TT, Wedel DJ, Benzon H, Brown DL, Enneking KF, Heit JA, Mulroy MF, Rosenquist RW, Rowlingson J, Tryba M, Yuan C-S. Regional anesthesia in the anticoagulated patient: defining the risks. Reg Anesth Pain Med 2004;29(Suppl):1–11. 13. Paice JA, Penn RD, Shortt S. Intraspinal morphine for chronic pain: a retrospective multicenter study. J Pain Symptom Manage 1996;11:71–80. 14. Gourlay GK, Cherry VA, Cousins MD. Cephalad migration of morphine in CSF following lumbar epidural administration in patients with cancer pain. Pain 1985;23:317–326. 15. Turner DB. The importance of placebo effects in pain treatment and research. JAMA 1994;221:1609–1614. 16. Witt W. Medtronic Lecture Series. Personal communication. University of Kentucky. 17. Krames ES. Intrathecal infusion therapies for intractable pain: patient management guidelines. J Pain Symptom Manage 1993;8:36–46. 18. Follett KA, Boortz-Marx RL, Drake JM, Du Pen S, Schneider SJ, Turner MS, Coffey RJ. Prevention and management of intrathecal drug delivery and spinal cord stimulation system infections. Anesthesiology 2004;100: 1582–1594. 19. Stearns LJ, Boortz-Marx R, Du Pen S, Friehs G, Gordon M, Halyard M, Herbst L, Kiser J. Intrathecal drug delivery for the management of cancer pain: a multidisciplinary consensus of best clinical practices. J Support Oncol 2005;3:399–408. 20. Smith TJ, Staats PS, Deer T, Stearns LJ, Rauck RL, Boortz-Marx RL, Buchser E, Català E, Bryce DA, Coyne PJ, Pool GE Pool, for the Implantable Drug Delivery Systems Study Group. Randomized clinical trial of an implantable drug delivery system compared with comprehensive medical management for refractory cancer pain: impact on pain, drug related toxicity and survival. J Clin Oncol 2002;20:4040–4049.
297
14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine Shahram Rahimi, Ali Shaibani, and Ajay K. Wakhloo
Spinal Vascular Disease Overview In general, the following entities have been listed under spinal vascular disease: • • • • •
Hemangioblastomas Cavernous malformation/angioma Spinal aneurysms Arteriovenous fistulae Arteriovenous malformations.
With regard to vascular lesions of the vertebral bodies, aneurysmal bone cysts and vertebral hemangiomas also can be mentioned. Many different classification schemes have been suggested over the last three decades. The newest proposed classification for spinal vascular lesions is by Spetzler et al.1 The authors list the most prevalent classification scheme for arteriovenous fistulae and malformations along with the new classification and the angiographic/anatomic classification (Table 14.1). Extradural AV Shunts Spinal AV shunts involving the extradural structures can involve the epidural space alone, or the bone and epidural space. The majority of purely epidural lesions encountered are AV fistulas, but AVMs affecting the epidural space also can occur. Those AV shunts involving the bone and epidural space are usually AVMs or complex AVFs with multiple arterial feeders. Epidural AVFs are fistulae to the ventral epidural venous plexus and are usually slow-flow lesions. Those fistulae, which drain only into the epidural venous system usually and present with compressive myelopathy or radiculopathy due to enlarged epidural veins. Lesions have been reported that drain primarily into the ventral epidural venous plexus and then secondarily into the From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_14, © Springer Science + Business Media, LLC 2010
299
Hemangioblastoma
Cavernous malformation
Spinal aneurysms
Arteriovenous fistulas:
Extradural
Hemangioblastoma
Cavernous angioma
Spinal aneurysms
Vascular malformations:
Epidural arteriovenous fistula
Classic avm, glomus avm
Venous hypertension, compression, hemorrhage
Hemorrhage, compression, vascular steal
MRI
MRI MRI
Diagnostic tests
Progressive myelopathy
Progressive myelopathy, radiculopathy
CTA, Mri/MRA DSA
Acute myelopathy, CTA, Mri/MRA pain, progressive myelopathy
CTA, Mri/MRA, DSA
CTA, Mri/MRA, DSA
CTA, Mri/MRA, DSA
Progressive CTA, Mri/MRA, myelopathy, DSA Epidural Hematoma
SAH
Parenchymal Hemorrhage
Masseffect
Presentation
Compression (enlarged Progressive veins/venous varix), myelopathy hemorrhage, vascular steal, venous hypertension
Venous hypertension/ compression; hemorrhage is rare
Cord compression (enlarged veins), vascular steal, venous congestion
Pathophysiology
Juvenile avm, Cord compression, Pain, progressive metameric avm hemorrhage, vascular myelopathy steal
Peri-medullary avf
Dorsal extramedullary avm
Epidural avf
Cavernous angioma
Other nomenclature
Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004. Note: CTA = CT Angiography; MRI = Magnetic Resonance Tomography; MRA = Magnetic Resonance Angiography; DSA = Digital Subtraction Angiography.
Paraspinous vascular malformation: vertebro-vertebral av fistula
Compact diffuse conus medullaris
Type ii
Intramedullary avm Intradural intramedullary
Type IV (subtypes a, b, c)
Type i (subtypes a, b)
Type iii
Ventral
Dorsal
Prevalent classification
Extradural/intradural Arteriovenous avm malformations: Extradural/intradural
Intradural (pial) arteriovenous fistula
Dural arteriovenous fistulas
Neoplastic vascular lesions
Neoplastic vascular lesions:
Intradural
New classification1
Angiographic/anatomic classification
Table 14.1. Classification of spinal vascular lesions and clinical presentation
Spinal Vascular Disease
intradural/medullary venous system. These lesions can cause venous hypertension or SAH. In general, extradural AV shunts in isolation (not as part of a juvenile type III malformation) are rare. Epidural Arteriovenous Fistulas Zhang et al.2 in 2006 performed a literature review, revealing 28 reported cases. In the review, males were affected more than twice as frequently as females (20:8), and the onset of symptoms occurred over a very wide age range (7 months to 79 years), but half were less than 28 years of age at presentation. Lesions were seen in all spinal segments, but most frequently in the cervical (13/28) and thoracic spine (11/28). Epidural hematoma occurred in ten of 28 patients, spontaneously or associated with trauma. Myelopathy, whether transient, acute, or progressive, was seen in ten patients. Paraparesis, hemiparesis, or quadriparesis (acute, transient, or progressive) was seen in nine patients. Other symptoms included local pain and radiculopathy. Myelopathy, paresis, and radiculopathy are usually a consequence of spinal cord or nerve root compression, either by epidural hematoma or dilated epidural veins. Spontaneous epidural AVFs have also been reported in patients with NF-1, apparently almost always in the cervical region2 (For more details, see section “Systemic Syndromes Associated with Spinal Vascular Disease.”). Vertebral (Bony) and Epidural AV Shunts AV shunts involving the bony vertebra and the epidural space do occur.3 Symptomatology is again due to compression of nerve roots by dilated epidural veins. If myelopathy occurs, it is either due to external compression by dilated epidural veins or secondary reflux of the epidural veins into perimedullary veins. Therapy Extradural AVF Endovascular therapy is recommended, with the objective of occluding the draining vein at and just beyond the fistula point. This can be done via both transarterial and transvenous approaches, using liquid embolic agents (Trufill NBCA, Codman Neurovascular, J&J, Raynham, MA or Onyx®, ev3 Neurovascular, Irvine, CA), coils, or a combination. Surgery can also be performed with the same objective. Extradural AVMs Endovascular treatment with liquid embolics and/or coils is usually necessary, whether as a primary curative/palliative procedure or as a presurgical adjunct. Dural Arteriovenous Fistulas (Dorsal Intradural AVF or Type I) Spinal dural fistula represents the most common type of spinal vascular disease and should be in the differential in an adult presenting with gradually worsening myelopathy. It should be noted that most authors have classified this lesion as a dural fistula. It is subdivided into Type A (single arterial feeder) and Type B (multiple arterial feeders). The most common location for these lesions is
301
302
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
between T4 and L3, with the peak incidence between T7 and T12.4 Spinal dural fistulas very uncommonly occur above the level of the heart, possibly due to the helpful effect of gravity on venous drainage above the level of the right atrium. This lesion is composed of a direct fistula between the dural branch of a radicular artery (only rarely of a radiculomedullary artery) at the level of the proximal nerve root and a radiculomedullary vein (Type A, Figure 14.1), or several abnormal connections between branches of adjacent radicular arteries and a radiculomedullary vein (Type B). The arterialized radiculomedullary vein then transmits the increased flow and pressure to the valveless coronal venous plexus and longitudinal spinal veins. The radiculomedullary vein is usually enlarged and tortuous as a consequence. The mean intraluminal venous pressure is increased to 74% of the systemic arterial pressure.5 The normal venous pressure in the coronal venous plexus is approximately 23 mmHg, which is almost twice that of the epidural venous plexus; this gradient is necessary for venous drainage. In one series, the mean venous pressure in the coronal venous plexus was measured at 40 mmHg.1 The consequent venous hypertension causes progressive myelopathy, often leading to paraplegia, bowel, bladder, and sexual dysfunction, with gradual worsening over months to a few years. The majority of patients become severely disabled within 3½ years.1 The overwhelming majority of patients (79–85%) are men, and 86 % of patients are 41 years of age or older at presentation.4,6 The mean age at presentation is 55, with patients as young as 26 reported as presenting with this kind of malformation. The most common presentation is progressive paraparesis of the lower extremities with sensory changes also. Complaints of back and leg pain are common. Although the progression is usually continuous, it can also present in a step-wise fashion or a waxing–waning course with gradual progression. Ten to twenty percent of patients can present with an acute exacerbation. The symptoms can be exacerbated by any physical activity that increases intra-abdominal pressure and, thus, central venous pressure, as well as by an upright posture (venous drainage hindered by gravity). Van Dijk et al.7 reported a series of 49 consecutive patients treated between 1986 and 2001. The mean age of the population was 63, with 80% being men. Ninety-eight percent of patients exhibited myelopathy, with 96% displaying leg weakness and/or paraparesis. Ninety percent had sensory numbness or paresthesias, and 55% had pain either in the lower back or lower extremities. Eighty-two percent had urinary incontinence/retention, and 65% complained of bowel dysfunction. Atkinson et al.8 reported a second series of 94 patients treated between June 1985 and December 1999. All of the patients had lower extremity weakness with or without perineal or bowel/bladder dysfunction. Five patients also had upper extremity symptoms, all of whom had high T2 signal within the cervical cord. Eighty-eight patients reported sensory loss, and 61 patients had bowel/bladder dysfunction. A very interesting finding in this series was an essentially 50/50 split among patients with symmetric vs. asymmetric lower extremity symptoms; in addition, approximately 50% of patients demonstrated worsening of symptoms with erect posture/valsalva and improvement with recumbent position. This effect was not as prominent in the group of patients
Spinal Vascular Disease
with the most severe symptoms. Eight of the patients included in this series had posterior fossa dural arteriovenous shunts with drainage into the medullary venous system, which is a well-described phenomenon and necessitates the injection of the posterior fossa and external carotid arteries in completion of a total spinal angiogram. The most common misdiagnosis for these lesions was “transverse myelitis.” Therapy The surgical treatment for these malformations has been well described and essentially consists of performing a laminectomy(ies), opening of the dura, and surgical disconnection of the draining vein, just distal to the fistulous site. In experienced hands, this is a very effective technique. Atkinson et al.8 reported a 97.9% success rate for obliteration of the fistula, with morbidity equal to that of patients undergoing decompressive laminectomy (one superficial wound infection and two DVTs). The endovascular treatment of these lesions also has been well described. Before the availability of acrylate products (“glue”), treatment consisted of selective microcatheterization of the feeding artery, with particulate embolization of the fistula using polyvinyl alcohol particles. Despite high rates of angiographic success immediately after treatment, this technique was associated with a high recurrence rate due to recanalization of the arterial feeding pedicles, with recurrence rates up to 83%. With the availability of liquid embolic agents (acrylates and Onyx®), the recurrence rate has significantly diminished. The consensus among interventional neuroradiologists at this time is that successful treatment of these malformations consists of penetration of the fistula and the proximal radicular draining vein to obviate the need for future surgery (Figure 14.1).7,9 The treatment protocol used in the series of patients presented by Van Dijk et al.7 used endovascular therapy as the first line of treatment due to its noninvasive nature, low complication rate, and the ability to obtain immediate angiographic control and confirmation of obliteration of the malformation. Using the endovascular criteria that involves the ability to penetrate the fistula and proximal draining vein and the ability to treat the malformation in a single session, only 11 (25%) of the patients were treated via the endovascular route, all of whom demonstrated a clinical success rate and stability equivalent to that of surgery (mean follow-up of 32.3 months) with no permanent complications. Using less stringent criteria, other endovascular specialists using acrylate have reported success rates of up to 90%, but with recurrence rates of up to 23%. Following complete occlusion of the fistula, the progression of the disease can be stopped; however, only 2/3 of all patients have a regression of their motor symptomatology, and only 1/3 show an improvement of their sensory disturbances. Impotence and sphincter disturbances are seldom reversible.10 Intradural (Pial) Arteriovenous Fistula (Ventral Intradural AVF or Type IV) The type IV AVF represents a direct fistula from the anterior spinal artery to the coronal venous plexus (Figure 14.2). Dorsolateral pial network (posterior spinal artery) supply may also be involved and may occasionally
303
304
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
Spinal Vascular Disease
305
Figure 14.2. Schematic illustration of a pial arteriovenous fistula (Type IV). (1) lumbar artery; (2) longitudinal pretransverse anastomosis; (3) nerve root sleeve; (4) dorsospinal branch; (5) dural artery; (6), radiculomedullary artery; (7) dorsal somatic artery; (8) anterior spinal artery; (9) coronal venous plexus; (10) anterior median vein; (11) arteriovenous fistula between anterior spinal artery and the anterior median vein (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
be the sole supply. Most pial fistulae are on the ventral surface of the spinal cord, although dorsal locations are occasionally seen, in which case radiculopial (PSA) supply is the norm11,12 There is a significant male predominance. The original classification was based on the degree of arteriovenous shunting and is outlined next. There are three subtypes: A, B and C. Subtype A (also classified as Merland subtype I) represents a small shunt, with moderate venous hypertension. There is no enlargement of the ASA (occasionally PSA) and only minimal dilatation of the ascending draining vein.1,13 The fistula is located at the point where a vessel caliber change is seen.14 The ASA (or rarely PSA) is the only feeder, and the AVF is typically located along the anterior aspect of the conus
Figure 14.1. Dural arteriovenous fistula. (A) Schematic illustration of a dural arteriovenous fistula: (1) descending aorta; (2) lumbar artery; (3) dural artery; (4) dorsal somatic artery; (5) nerve root sleeve, (6) nerve–arteriovenous fistula complex; (7) radiculomedullary vein, (8) dorsal longitudinal vein. (B) Contrast-enhanced T1-weighted magnetic resonance image of a dural arteriovenous fistula (DAVF), a nonspecific diffuse enhancement, and swelling of the spinal cord (arrow). (C) Three-dimensional TOF magnetic resonance angiogram in coronal plane shows congested radiculomedullary vein (arrow) and dorsal median vein (open arrow). Superselective angiogram of an intercostal artery (D, arrow) shows (E) the DAVF (curved arrow), the retrograde draining and congested radiculomedullary vein (open arrow), and the congested dorsal median vein (heavy black arrow) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
306
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
medullaris or proximal filum terminale.15 This subtype is reported to be the least frequent subtype.12 Subtype B (Merland subtype II) represents a moderate-sized shunt with moderate enlargement of the feeding artery(ies) and the draining veins. The location of the fistula is marked by venous ectasia.14 There are several abnormally dilated feeding arteries, composed of the ASA and one or two arteries from the dorsolateral pial network (PSA), all of which converge on the fistula. These are typically located at the level of the conus. Venous drainage is into tortuous and dilated ascending perimedullary veins.15 Subtype C represents a giant fistula with very large arterial feeder(s) from the ASA and dorsolateral pial network (PSA) converging into the fistula, and draining directly into a giant venous ectasia, often embedded within the substance of the cord. These fistulas are rare,14 although, in at least one large series, they represented the largest subtype of ventral intradural AVFs.15 The location of the fistula is more difficult to ascertain because of the giant ectatic draining vein.14 The giant ectatic draining vein usually drains into the local metameric efferent veins, which are also dilated.15 These lesions are typically located at the thoracic or cervical levels.16 More recent reports have considered this subtype to be most commonly associated with Osler–Weber–Rendu (hereditary hemorrhagic telangiectasia) syndrome.16,17 These are typically found in childhood (HHT patients). If encountered in an adult, the possibility of underlying HHT should be considered. More recently, Rodesch and colleagues have suggested a reclassification of intradural AVFs into the micro-AVF (mAVF) and macro-AVF (MAVF) categories.17 Micro-AVFs would comprise the subtypes A and B outlined above, and macro-AVFs would correspond to the subtype C. Macro-AVFs, especially in the pediatric population, are almost always associated with HHT. Signs and symptoms may be due to vascular steal (more so with higher flow), venous hypertension, mass effect (with venous enlargement/ aneurysms), subarachnoid hemorrhage (SAH)1 or hematomyelia.16 In a series of 32 patients (22 adults, 10 children), hemorrhage (hematomyelia > SAH) was seen in 80% of children and 54% of adults, while nonhemorrhagic progressive myelopathy was the next most common presentation.16 The clinical signs and symptoms almost always appear before age 40 and often present during the first decade (mean age at diagnosis being between 11.5 and 13.5 years). Subarachnoid hemorrhage is the presenting sign in approximately 40% of patients in one series.14,18 Paraparesis or paraplegia is the most common sign, with progressive deterioration over time. Radiculomyelopathy or radiculopathy can also be present, usually due to mass effect from dilated venous structures. These lesions can be seen anywhere along the spine. While the fistula is often ventral, lateral, or posterolateral, locations also occur when there is significant involvement of the dorsolateral pial network (PSA). Therapy In subtype A, the blood supply generally occurs through a minimally dilated anterior (or posterior) spinal artery with slow flow, often making endovascular access and navigation of the artery to the point
Spinal Vascular Disease
of fistulization either impossible, or difficult and possibly dangerous. Newer techniques, such as using a flow-directed microcatheter with a microwire inside (but not extended beyond the microcatheter tip), have made endovascular treatment of some type A fistulae possible, especially in those cases in which some minimal enlargement of the ASA (rarely PSA) is present.12,19 In most cases, surgical obliteration remains the only choice. In case the fistula is located in the ventral surface of the spinal cord and surgical access is difficult, a PVA-particle embolization from a proximal catheter position may be considered. Subtype B shows higher flow within one or multiple dilated pial arteries. Thus, an endovascular approach is feasible with curative obliteration of the arteriovenous fistula using NBCA or Onyx®. In the case of a complex AVF, an intraoperative transvenous embolization has been described.20 In the case of subtype C, the AVF is large and the feeding arteries extremely dilated. Detachable balloons or fibered coils have been used in the past for a permanent obliteration. Under flow control, NBCA may be used safely for a complete closure.18 Extradural/Intradural AVM (Type III) Extradural/Intradural AVM is also known as metameric or juvenile AVM. If all derivatives of the metamere (i.e., skin, muscle, bone, dura, and cord) are involved, it is known as Cobb’s Syndrome. Therapy Because of the complex nature of the malformation, a combined endovascular and surgical approach is recommended.21 However, in rare situations, the authors have been successful with staged endovascular NBCA embolization achieving curative results. Endovascular treatment is generally the mainstay, serving to reduce venous hypertension, vascular steal, and possibly mass effect. Surgery primarily serves to decompress the mass effect on nerve roots and the spinal cord22 Intramedullary AVM (Type II) Intramedullary AVM is also known as Type II or classic AVM. Spinal cord AVMs are the second most common spinal vascular malformation. The angio-architecture of these lesions is similar to the classical brain AVMs, with multiple arterial feeders, a nidus, and draining veins. The nidus can be compact (glomus type) or diffuse (occasionally called juvenile type, not to be confused with the metameric type). The arterial feeders are usually multiple branches of the ventral spinal axis (ASA) and/or dorsolateral pial network (PSAs). These lesions are high flow, high pressure, and low resistance malformations.1 There is a definite male preponderance of patients with this type of SCAVM. In their new proposed classification, Spetzler et al.1 subdivided these lesions into those with compact (glomus type) nidus, and those with diffuse nidus. The natural history is difficult to ascertain, but the majority of patients present before the age of 40.14 The most common presentation is an acute myelopathy due to intramedullary and/or subarachnoid hemorrhage.1,14 A proportion of patients present with intermittent or progressive myelopathy with deterioration of limb
307
308
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
function, or bowel and bladder function. The progressive myelopathy can be due to vascular steal, venous hypertension, or venous compression.1 Pain can also be a common presenting symptom in these patients. If left untreated, patients can be expected to experience an episodic but progressive deterioration due to repetitive bleeding.14 In one 8-year study of 60 patients, 36% of patients less than 41 years of age and 48% of patients aged 41–61 were wheelchair bound within 3 years of diagnosis.23 Based on Djindjian’s original series of 150 patients, 13 % of patients at the 5-year follow-up, 20% of patients at the 10-year follow-up, and 57% of patients at the 20-year follow-up had experienced clinical deterioration.24 Chen and colleagues, in 2007, presented two cases with the atypical initial symptoms of back pain and right chest pain, respectively. They surmised that the pain might have resulted from irritated central and/or radicular nerve fibers from the bleeding of the SCAVM.25 Therapy Early diagnosis of SCAVMs is important because patients are likely to have better functional statuses after treatment if their pretreatment deficits are relatively mild.25 In general, safe surgical resection is often difficult and can bear high morbidity because of the intramedullary location and blood supply from tiny perforators arising from sulcocommissural branches. These may arise as “en passage” arteries, which supply normal cord tissue. A carefully planned staged embolization with obliteration of the most proximal draining pial vein with NBCA may be curative (Figure 14.3). At this point, surgical extirpation usually preceded by embolization remains the mainstay of therapy.22 At very specialized, high volume neurovascular centers, gross total resection with minimized neurological injury can be achieved in up to 92% of patients.22 In cases not amenable to surgery or embolization with liquid embolics, a repeated PVA particle embolization may be the only alternative to reduce the size of the AVM nidus.26–28 Some authors have advocated selective embolizations of identified probable bleeding sites in those SCAVMs that have presented with hemorrhage but cannot be treated for cure.29 There is some recent limited data on the efficacy of Stereotactic Radiosurgery (SRS) in the treatment of SCAVMs. Sinclair and colleagues30 in 2006 reviewed their experience with 15 patients with intramedullary spinal cord AVMs (nine cervical, three thoracic, and three conus medullaris) who were treated by image-guided SRS (Cyberknife®, Accuray, Sunnyvale, CA) between 1997 and 2005. SRS was delivered in two to five sessions with an average marginal dose of 20.5 Gy. Clinical and magnetic resonance imaging follow-up were carried out annually, and spinal angiography was repeated at 3 years. In their series, six of seven patients who were more than 3 years from SRS had significant reduction in AVM size by MRI, and four of five who underwent spinal angiography showed smaller but persistent AVMs. There was one angiographic cure of a conus medullaris AVM. There was no evidence for recurrent hemorrhage or neurological deterioration after SRS in this small series with limited follow-up. The authors concluded that, while the initial experience is promising, the optimal treatment parameters are unclear, and additional study and experience are necessary.30
Spinal Vascular Disease
309
Figure 14.3. Images of the brain of a 21-year-old male who presented with subarachnoid hemorrhage associated with an intramedullary arteriovenous malformation. The AVM was treated in multiple staged sessions with N-butylcyanoacrylate. (A) Computed tomographic image without contrast shows extensive SAH. (B) T1-weighted MRI without contrast shows flow voids within the AVM nidus located at the craniocervical junction (arrows). (C) Gradient echo T2-weighted axial MRI shows the extensive involvement of the anterior and central aspects of the spinal cord and an enlarged anterior median vein with flow void.
Conus Medullaris AVM Conus Medullaris AVM is a new category proposed by Spetzler and colleagues characterized by multiple feeding arteries, multiple niduses, and complex venous drainage. These lesions are composed of multiple direct arteriovenous shunts with feeders from the ventral
310
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
Figure 14.3. Continued. (D–F) Vertebral artery injection in the lateral and anterior–posterior projection shows the extent of the intramedullary AVM, with early drainage of the lower parts of the nidus through the enlarged anterior median vein (D, arrow). (E) Late arterial phase shows both enlarged veins of the middle cerebellar peduncle draining via the superior petrosal vein into the superior petrosal sinus (F, black arrow) and the dilated median anterior pontomesencephalic vein (F, open arrow).
Spinal Vascular Disease
Figure 14.3. Continued. Vertebral artery injection in lateral and anterior– posterior projection. (G) Superselective injection of the anterior spinal artery (open arrow) through a flow-guided micro-catheter, which has been placed over a guide wire (straight black arrow). Multiple sulcocommissural arteries are feeding the AVM nidus (curved arrow). A few of them have already been embolized with acrylate (see subtraction artifact).
311
312
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
Figure 14.3. Continued. (H) Superselective injection of a sulcocommissural artery (thin arrow) shows a compartment of the intramedullary AVM that drains into the enlarged anterior median vein (thick arrow) cranially the vein of middle cerebellar peduncle. The caudal drainage occurs through the anterior median vein (small arrow). (I, J) A staged embolization with complete AVM obliteration was achieved. Note the caliber reduction of the anterior spinal artery because of the shunt reduction. Note the displacement of the anterior spinal artery (J, curved arrow) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Neoplastic Vascular Lesions of the Spinal Cord
spinal axis (ASA), dorsolateral pial network (PSAs), as well as glomustype niduses that are usually extramedullary/pial, but can occasionally be intramedullary.1 These lesions are always located in the conus medullaris and cauda equine, and can extend along the filum terminale all the way down.1 Symptoms can be caused by venous hypertension, venous compression of the cord/cauda equina, or hemorrhage. Unique to this type of spinal vascular malformation is frequent production of radiculopathy in addition to myelopathy.1 Therapy According to Dr. Spetzler and colleagues, “conus medullaris” AVMs are treated with a combined endovascular and microsurgical approach.1 Careful identification of ASA and posterior spinal artery branches separate from the lesion is crucial. Because the venous structures associated with conus AVMs are so hugely dilated, surgical decompression of adjacent spinal cord and nerve roots can relieve neurological symptoms significantly. Conus AVMs are usually easily accessible from a posterior approach. Continuing experience with these entities has demonstrated that aggressive combined treatment can result in good outcomes.22 Isolated Spinal Aneurysms Isolated spinal Aneurysms are aneurysms arising from the ASA, PSA, or occasionally from a radiculopial or radiculomedullary artery, not associated with arteriovenous shunts. These aneurysms are exceedingly rare and usually manifest with sudden onset back or leg pain due to spinal SAH. Those occurring in the cervical spine can also cause intracerebral SAH. A less frequent presentation is that of myelopathy due to mass effect from the aneurysm. Aneurysms arising from the posterior spinal artery have been reported in four cases, all of which occurred in the upper cervical spine.31 Isolated spinal aneurysms are usually fusiform, and often associated with a dissection of the artery.32 Treatment is usually surgical, either by trapping and excision (if significant distal flow is not seen) or by external wrapping.32
Neoplastic Vascular Lesions of the Spinal Cord Cavernous Malformations (Cavernous Angiomas) Cavernous Malformations are slow-flow vascular lesions, consisting of sinusoidal vascular channels lined by a single layer of endothelium and separated by collagenous stroma, and without any normal intervening neural tissue. Grossly, they are well-circumscribed lesions, with a reddish-purple color, often likened to a “mulberry” or “cluster of mulberries.” There is a characteristic gliotic reaction in the surrounding parenchyma, which may form a pseudocapsule. Within the lesion, there is often evidence of hyalinization, thrombosis in various stages of organization, calcification, cholesterol crystals, and cysts.33 The immediate
313
314
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
surrounding parenchyma may contain small, low-flow feeding arteries and draining veins. In a review of 57 reported cases of spinal cavernomas, Canavero et al.34 found that 69% of the patients were women, with a mean age of 36.4 years at diagnosis. They estimated a 1.6% per person per year risk of bleeding, with a higher risk in cervical lesions. MRI was found to be diagnostic in all cases, while angiography was negative in 100%. This latter finding is not strange, considering that cavernomas are one of the classical “angiographically occult” vascular malformations. Typical clinical features of intramedullary cavernous malformations are sensorimotor deficits, usually several hours after the onset of pain. The clinical course is variable, ranging from slowly progressive symptoms to acute quadriplegia. The acute setting is probably caused by new hemorrhage within or around the lesion. Repeated episodes of small bleedings or local pressure effects of the lesion itself on the surrounding spinal cord tissue by capillary proliferation and vessel dilatation may be responsible for slowly progressive symptoms. Once the lesion becomes symptomatic, progressive myelopathy is the most common course.35 Cavernomas can also rarely present as intradural extramedullary36 or epidural lesions.37 Therapy Therapy can consist of a conservative approach or a surgical resection. There is no role for an endovascular approach. Hemangioblastomas Hemangioblastomas are true neoplasms of blood vessels. They can arise either spontaneously, or can be associated with von Hippel–Lindau syndrome. In the spinal cord, they constitute 3.3% of intramedullary tumors, and most commonly present in the fourth decade.38 Up to 30% of patients with spinal cord hemangioblastomas have von Hippel–Lindau syndrome. The majority of spinal hemangioblastomas (79%) are single. The thoracic cord is the most common site, followed by the cervical cord.38 Angiographically, they are very vascular lesions, often with arteriovenous shunting. Rarely, hemangioblastomas can be intraosseous, within the vertebrae, where their imaging appearance is often similar to hemangiomas. In this location, they can involve adjacent vertebrae while sparing the disk and cause extensive local destruction and cord/nerve root compression.39 Therapy A preoperative embolization, usually with particles, significantly reduces the risk of a surgical resection40,41 (Figure 14.4).
Systemic Syndromes Associated with Spinal Vascular Disease Spinal AVMs coexist with other congenital abnormalities, including cutaneous angioma, vertebral anomalies, vertebral hemangioma, Rendu–Osler–Weber Syndrome, Cobb’s Syndrome, and venous and lymphatic dysplasia, in up to 25% of cases.42
Systemic Syndromes Associated with Spinal Vascular Disease
315
Figure 14.4. (A) A hypervascular intramedullary mass of the upper cervical region consistent with a hemangioblastoma. Major blood supply occurs through the C2 radicular artery (large arrow). Note the anterior spinal artery (small arrow). (B) Superselective placement of a microcatheter within the radicular artery for a preoperative PVA particle embolization. (C) Control angiogram after complete devascularization (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
316
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
Osler–Weber–Rendu Syndrome (Hereditary Hemorrhagic Telangiectasia) Hereditary hemorrhagic telangiectasia (HHT, Rendu–Osler–Weber disease) is an autosomal dominant vascular dysplasia with a high penetrance but variable expressivity. It consists of two genotypes (Types 1 and 2). Type 1 is associated with mucocutaneous telangiectasias, pulmonary arteriovenous fistulas, and arteriovenous shunts of the central nervous system. HHT can no longer be regarded as a rare disease; recent epidemiological surveys suggest an incidence of HHT of one in 5–10,000.1–3 Mutations in two different genes are responsible for HHT, and most cases of HHT are caused by mutations in the endoglin gene on chromosome 9q or the activin receptor-like kinase 1 (ALK-1) gene on chromosome 12q. However, a third location on a yet unknown locus must be taken into consideration since some patients with the classical signs of HHT test negative for both the endoglin and ALK-1 gene defects.43 Endothelial cells lacking functioning endoglin or ALK-1 form abnormal vessels and abnormal connections between vessels, resulting in vascular malformations that may occur in multiple organs including the lung, liver, gastrointestinal tract, skin, or the Central Nervous System (CNS). Diagnosis of HHT is based on the four Curacao criteria of: 1. Spontaneous recurrent nosebleeds 2. Mucocutaneous telangiectasia at characteristic sites (lips, oral cavity, fingers, or the nose) 3. Visceral involvement such as pulmonary, hepatic, or CNS arteriovenous malformations (AVMs) 4. An affected first-degree relative. Definite HHT is diagnosed in patients where three criteria are present; in patients with two criteria, HHT can only be suspected. It is estimated that approximately 10–20% of all HHT patients have CNS involvement44 consisting of three types of neurovascular malformations: arteriovenous fistulae (AVFs), small arteriovenous glomerular or nidus-type malformations (AVMs), and micro-AVMs. Approximately one third of neurological complications, including seizures, hydrovenous disorders, and intracranial hemorrhage, are due to cerebral or spinal vascular malformations.45 With regards to cerebrovascular manifestations, the disease displays an age-related expression, with manifestations developing throughout life and varying among affected individuals. While AVFs are present almost exclusively in the age group of young children under 5 years of age, small AVMs are present predominantly in young adolescents, whereas micro-AVMs are present in young adults. Spinal manifestations of HHT may lead to hematomyelia with acute tetraplegia.43 In the retrospective review of neurovascular malformations in children with HHT in 2005 by Krings and colleagues, a total of 31 patients under the age of 16 were included.43 Twenty children presented with 28 arteriovenous (AV) fistulae, including seven children with spinal AV fistulae and 14 children with cerebral AV fistulae. One child had both a spinal and cerebral fistulae. With respect to the
Systemic Syndromes Associated with Spinal Vascular Disease
seven patients with spinal cord AV shunts, all lesions were AVFs. The age range at presentation was 1 month – 6 years. Presenting symptoms were acute tetra-or paraplegia in 5 (due to hematomyelia), progressive tetraplegia (from venous hypertension) in one, and spinal SAH in one patient. The spinal AVFs encountered in these patients typically connect the anterior or posterior spinal artery directly with medullary veins. These large AVFs are always located on the surface of the cord and are always intradural (pial) arteriovenous fistulas, subtype C (ventral intradural AVF, or type IV, subtype C). Cobb’s Syndrome Cobb’s Syndrome is the synonym for the complete manifestation of the metameric type of spinal vascular malformation. Klippel–Trenaunay and Parkes–Weber Syndrome Klippel–Trenaunay (KT) and Parkes–Weber (PW) Syndromes consist of vascular malformations involving the lower limbs primarily, with the following dominant features: (1) cutaneous capillary malformations; (2) varicose veins; (3) limb hypertrophy. KT syndrome is primarily composed of venous anomalies, while PW syndrome has more arteriovenous shunts.46 Spinal cord involvement with pial AVF or AVMs, as well as extradural spinal AV shunts, can be present in these patients.11 Therapy for Osler–Weber–Rendu Syndrome, Cobb’s Syndrome, KT Syndrome, and PW Syndrome Therapy for Cobb’s, KT, and PW Sydrome depends on the type of lesion present and usually consists of embolization alone or with surgical excision. Neurofibromatosis Type-1 NF-1 is a relatively common autosomally dominant inherited disease due to a mutation on the long arm of chromosome 17, with an approximate incidence of 1:3,500. The tissue dysplasia caused by this mutation affects multiple tissues, including the vasculature. Intracranial vascular abnormalities described include aneurysms, Moya-Moya like vessels, as well as arterial occlusions or stenoses. In a very good literature review of spinal vascular anomalies in NF-1 in 2006, Hauck and colleagues described 31 cases of spontaneous spinal AVF associated with NF-l reported through 2005.47 Ninety-seven percent of the lesions were cervical, and all were extradural, usually epidural. The arterial supply most commonly arose from the vertebral arteries, thyrocervical and costocervical trunks, or occipital arteries. The venous drainage was almost always epidural/paraspinous. The patients were 84% female with a mean age of 39 years; only 16% were male with a mean age of 35 years. Clinically, the fistulae constitute a syndromic triad including symptoms of NF-1 (100%), progressive radiculomyelopathy (78%), and a bruit (50%). Other symptoms included tinnitus (10%), cranial neuropathy (3%), or a pulsatile neck mass (3%).
317
318
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
No patient presented with subarachnoid hemorrhage. The symptoms are primarily a function of compression of the spinal cord and/or nerve roots by the dilated epidural veins. The authors made the point that, if confused with a highly vascular neurofibroma, the AVF may put the patient at serious risk of unexpected sudden life-threatening bleeding during resection. Therapy The treatment for neurofibromatosis type-1 is endovascular embolization (transarterial or transvenous), surgical, or occasionally both in case decompression is necessary.
Miscellaneous Vascular Lesions of the Vertebrae Vertebral Hemangiomas Vertebral hemangiomas are benign vascular malformations of the bone, with very well-known and well-described appearance on conventional radiography, CT, and MR. The incidence of hemangiomas is variable, depending on age, but has been reported around 11% with increasing age. Up to 30% of patients have multiple lesions. Pathologically, they are considered to be postcapillary vascular dysembryogenetic malformations. Microscopically, they are divided into capillary, cavernous, and mixed types.48,49 The vast majority of these lesions are asymptomatic and are incidental findings on MRI performed for other reasons. Less than 1% of hemangiomas become symptomatic.48 A review of three series describing the treatment of hemangiomas causing cord compression, with a total of 34 patients, suggests the following characteristics of hemangiomas presenting with neurological symptoms (cord compression or radiculopathy): 1. Twenty-two of 34 patients (65%) had holovertebral (body, pedicles, and laminae) involvement; eight of 34 (23.5%) had partial body and pedicle/posterior element involvement; and four of 34 (11.8%) had involvement of the body only 2. The majority (25 of 34, or 73.5%) are women 3. Young adults form a large portion of patients presenting with cord compression and/or radiculopathy 4. The majority of lesions (17 of 23 in two series, or 74%) are in the thoracic spine.48–50 Fox and Onofrio49 noted that neck or back pain often preceded the neurological symptoms and that thoracic myelopathy was the most common neurological presentation. An additional known risk factor for the development of neurological symptoms is pregnancy, with symptoms developing in the 3rd trimester51 This is possibly due to the role of estrogen and/or increased venous pressure due to abdominal distention and pressure of the growing uterus on the venous structures. The mechanism for cord compression can be epidural extension of the lesion from the bone (vertebral body or post elements) into the spinal canal, expansion of the bony vertebra by the hemangioma, a pathological fracture of the vertebra, epidural
Miscellaneous Vascular Lesions of the Vertebrae
hematoma from bleeding from the lesion, or compression by enlarged feeding arteries or draining veins.51 Djindjian et al.52 divided vertebral hemangiomas into three groups based on clinical and imaging characteristics. Type A: Lesions Causing Cord Compression Type A lesions present with signs and symptoms of cord compression, and imaging demonstrates extraosseous extension of the lesion, usually related to a fracture (insufficiency fracture) due to the presence of the lesion weakening the vertebral body (Figure 14.5). Angiography demonstrates dense opacification of the vertebral body via enlarged osseous (somatic) branches of a normal-sized intercostals/segmental artery. The appearance of the lesion in the vertebral body is described as dense “pools” of contrast appearing in the midarterial phase and persisting into the venous phase. Therapy The usual treatment for these Type A lesions consists of preoperative embolization of the lesion with particles and/or liquid embolic agents (N-butyl-cyanoacrylate or Onyx®, ev3 Neurovascular, Irvine, CA), and operative decompression of the spinal cord/canal, possibly with resection of the lesion and spinal reconstruction and stabilization (Figure 14.5). Doppman et al.48 made the important observation that, even in these lesions with epidural extension, the lesion does not penetrate the dura but is confined by the periosteum, which results in the characteristic bilobed posterior margin of these lesions, indented centrally by the posterior longitudinal ligament. Additional treatment options in these patients in whom timely treatment is a medical necessity is the technique of percutaneous transpedicular injection of ethanol, which they used successfully in the treatment of 11 patients. All the patients in this series had appropriate cross-sectional imaging work-up. The vascularity of the lesions was determined by doing a CT with injection of iodinated contrast through arterial catheters placed selectively in the segmental arteries at the appropriate vertebral level. With current imaging technology, it is possible that dynamic contrastenhanced CT or MR may be adequate in this regard. The needles were placed percutaneously through the pedicle, with the tip of the needle usually positioned at the vertebropedicle junction; these lesions invariably had posterior extension and/or involvement of the posterior elements. Initially, contrast was injected through the needles, and a CT was performed to demonstrate opacification of the lesion. Subsequently, dehydrated ethanol opacified with metrizamide powder was forcefully injected. Due to the pain associated with ethanol injection, MAC anesthesia was recommended. For lower thoracic and upper lumbar lesions, the artery of Adamkiewicz was identified to ensure that it did not arise at the same level. The authors recommended an ethanol volume of less than 15 mL, as higher volumes were associated with subsequent avascular necrosis and compression fractures of the treated vertebrae. Their recommendation was that the injection of ethanol be stopped when no blood could be aspirated from the needle, or when the volume reached 15 mL. In this series of patients, five had complete and
319
320
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
Miscellaneous Vascular Lesions of the Vertebrae
five had partial relief of symptoms (one with no relief). Improvement of symptoms began within 1–2 days. MRI performed prior to discharge demonstrated nonenhancement and shrinkage of the lesions. Recently, Hadjipavlou and colleagues reported on a small case series of six patients with symptomatic hemangiomas. In two patients, they combined ethanol injection immediately followed by kyphoplasty, with good clinical results.53 Type B: Painful Lesions Type B lesions are associated with local pain and tenderness over the involved vertebral body and/or radicular signs. Imaging does not reveal any extraosseous extension. The angiographic appearance is similar to type A lesions. Therapy These lesions are generally large. The first step in the evaluation of these lesions is to exclude the more common causes of back pain, with the help of imaging and physical examination.51 Imaging further helps to exclude involvement of the posterior element, cortical disruption, and epidural spread of the lesion. In the absence of these findings, percutaneous vertebroplasty with polymethylmethacrylate is probably the treatment of choice. Other treatment options include endovascular transarterial embolization of the lesion using particles, N-butyl cyanoacrylate or ethanol. Embolization has been reported to be effective in 60–100% of cases.51 Percutaneous injection of opacified ethanol was described by Doppman and colleagues, but for the treatment of type A lesions.48 Reizine and colleagues suggested that, if a painful lesion is located in the cervical or lumbar spine, without involvement of the posterior elements or cortical disruption, these lesions could be considered nonevolutive (without potential for future growth causing cord compression). On the other hand, if a painful lesion is located in the thoracic spine (especially in a young female) and demonstrates involvement of the posterior elements, cortical disruption, or softtissue extension, it should be considered an evolutive lesion, with serious potential for future cord compression. More recently, balloonassisted vertebroplasty (kyphoplasty) has been developed and has
Figure 14.5. (A) Fast spin echo T2-weighted image shows a vertebral body hemangioma with spinal canal stenosis and cord compression due to extraosseous extension. (B) Contrast-enhanced T1-weighted image shows the enhancing extraosseous epidural extension of the hemangioma with cord compression (arrows). (C) Selective angiogram of a left intercostal artery shows a hypervascular vertebral body with blood supply through perforating somatic branches of the intercostal artery. (D) Preoperative/PVA particle embolization through a microcatheter, which was placed coaxially through the diagnostic catheter. A fibered coil has been placed distal to the origin of somatic branches within the intercostal artery (arrow) to flow direct PVA particles preferentially into the feeding pedicles. The control angiogram shows a nearly complete devascularization. The patient received a high dose of corticosteroids prior to the procedure (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
321
322
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
been reported to be efficacious as a single or adjunct procedure for the management of symptomatic vertebral hemangiomas in at least one small case series.54 Type C: Asymptomatic Hemangiomas Type C represents the vast majority of hemangiomas, which are incidental findings. There are no associated symptoms, and the angiogram is normal.14 Therapy These lesions represent the vast majority of hemangiomas. Generally, unless the patient develops symptoms (i.e., pain and/or neurological deficits), follow-up imaging or additional studies are not necessary.51 An exception can be made for very large lesions (e.g., holobody lesion) in a young woman, in which the chance of further growth of the lesion is higher, and yearly follow-up may be considered. Aneurysmal Bone Cysts Aneurysmal Bone Cysts (ABCs) are benign lesions of bones that primarily affect young people; 80% of patients present under the age of 20. There is no sex predilection. While ABCs can occur at any location, 90% are seen in the spine. Within the spine, most lesions involve the posterior elements, although the vertebral body can also be involved. Additionally, ABCs (in addition to vertebral hemangiomas) can involve two contiguous vertebral bodies.55 The radiographic appearance of ABCs has been well described. Pathologically, the lesions consist of enlarged, communicating spaces within the bone containing venous blood under higher than normal venous pressure. The lining of the spaces consists of a fibro-osseous patchwork and some giant cells.55 Interestingly, up to 1/3 of ABCs are found in conjunction with other lesions, such as fibrous dysplasia, osteoblastoma, or chondrosarcoma,56 and others may be associated with previous trauma.57 With regards to pathogenesis, most authors believe that hemodynamic imbalance or abnormality within the bone is the etiologic factor, especially with regard to impaired venous drainage.57,58 Some have suggested the presence of a congenital vascular abnormality in cases of de novo ABCs and impairment of venous drainage by a secondary factor (associated lesions or trauma) in other cases.57 Angiographically, there is no pathognomonic pattern for ABCs. Findings can vary from faint or moderate vascularity to dense vascularity with a rich network of dilated, tortuous feeding vessels and a dense stain of the lesion within the vertebral body.55 Djindjian and colleagues described arteriovenous shunting in some lesions, while others have described patchy collections of contrast within the cystic spaces, persisting into the late venous phase.24 Therapy The most common approach to symptomatic ABCs is surgery, whether with curettage or with resection of the lesion and reconstruction of the spine if necessary. In many cases, due to the vascularity of the lesion,
Recommended Technique for Spinal Angiography and Intervention
the operating surgeon will request preoperative angiography and embolization of the lesion to decrease intraoperative blood loss, which can be significant (Figure 14.6). At least two separate papers have described the successful use of endovascular embolization as the sole therapy for ABCs. Cigala and Sadile59 described the results of embolization of six large ABCs in children in whom operative therapy was considered difficult. Long-term follow-up showed almost complete healing of the lesions and restoration of the normal shape of the affected bone. None of the patients required subsequent surgery. Radanovic et al.60 described the endovascular embolization of ABCs in five patients, all of whom had relief of their primary symptom (pain) and a decrease in size of the ABC. In those patients with a follow-up period greater than 12 months, sclerosis and recalcification of the lesions were described.
Neoplastic/Metastatic Lesions of the Spine Neoplastic and metastatic lesions can involve the vertebral bodies as well as intra-and extramedullary structures. The goal of endovascular treatment is devascularization prior to a planned surgery or biopsy (Figure 14.7). This significantly reduces the blood loss and improves the surgical resection.61–64 Most often, the embolization is performed using PVA or similar particulate embolic agents, or occasionally using liquid embolics such as NBCA/Onyx® or, rarely, dehydrated ethanol. Needless to say, attention has to be paid to the potential supply of radiculomedullary/pial arteries to the anterior or posterior spinal arteries. An embolization can, on rare occasion, lead to tumor necrosis with subsequent swelling and spinal cord compression. High-dose peri- and postprocedural corticosteroids medication has been suggested to help prevent such a complication.65 On rare occasion and in nonsurgical patients, embolization can be helpful for pain reduction and treatment of radicular compression.65 Although a reduction of tumor growth may be seen, embolization for spinal metastasis and malignant spinal tumors is strictly palliative. An endovascular or direct percutaneous embolization of a vertebral body metastasis or malignant tumor can be achieved. The latter can be performed under CT or fluoroscopic guidance66 and use of NBCA, poly-methylmethacrylate (PMMA), or dehydrated ethanol.67,68 Use of PMMA can additionally provide a biomechanical stability of the vertebral body.69
Recommended Technique for Spinal Angiography and Intervention This is meant to be a brief overview of techniques and intervention and does not serve the purpose of replacing standard textbooks in this field. Generally speaking, with modern catheter techniques in the hands of trained physicians, spinal diagnostic angiography should not have a complication rate higher than that of a diagnostic angiogram of
323
324
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine Figure 14.6. Spinal images of an 11-year-old boy who presented with intractable neck pain associated with an aneurysmal bone cyst after a football match. A preoperative transarterial PVA embolization was performed. (A) Lateral plain spine x-ray film shows a sharply demarcated osteolytic lesion of the posterior part of the C5 vertebral body (arrow) and narrowing of the spinal canal. (B) T1-weighted image shows the C5 lesion with well-defined calcified boundaries (arrow). There is no epidural extension or spinal cord compression visible. (C) Contrast-enhanced CT image shows a hypodense lesion (arrow) with enhancement of the margins. Note the involvement of the vertebral and neuronal foramina and extension into the lateral recess.
Recommended Technique for Spinal Angiography and Intervention Figure 14.6. Continued. (D) Selective right vertebral artery angiogram (lateral plane) shows tortuous feeding posterior and lateral somatic branches arising from two major supplying radicular arteries (arrows); moderate vascularity with a rich network of dilated and tortuous feeding vessels, and patchy collections of contrast material in the cystic spaces, persisting into the late venous phase (see E). (E) Late arterial (lateral plane) phase shows the prominent filling of the epidural venous network and depicts the persisting blush of the C5 vertebral body (arrows).
325
326
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
Recommended Technique for Spinal Angiography and Intervention
the peripheral vascular system. Infrequently, minor asymptomatic iliac or aortic dissections may be encountered in patients with significant arteriosclerosis. Diagnostic angiography of the spine should preferably be a focused study. MRI findings generally guide the invasive diagnostic work-up. Multidetector CT angiography has also recently shown promise for the noninvasive detection of type I spinal dural AVFs.70 Frequently, it might be pertinent to locate the artery of Adamkiewicz or radicularis magna as the major supply to the anterior spinal cord. However, if a vascular lesion, especially a dural arteriovenous fistula, is suspected, a more thorough angiogram may be required. This includes an angiogram of the aortic arch, the descending aorta, the abdominal aorta, and the pelvic system, and, in the case of a cervical spinal cord malformation, the vertebral arteries, the thyrocervical trunk, and the deep and ascending cervical arteries. More recent MRA studies have shown improved sensitivity in depicting dural AVFs and defining the level of the blood supply.71 This will help to focus the time needed for angiography. An aortogram can be accomplished best by using a 5-Fr pigtail catheter and a standard amount of contrast (30–40 mL), which is injected over 2 s using a high-pressure pump. This helps occasionally in finding the level of the feeding arteries of the expected vascular lesion and may serve as a map for the selective spinal angiography, especially in patients with several missing intercostal or lumbar arteries. However, the downside is that a large amount of contrast material is required for the study and may, especially in patients with impaired renal function, stop the study prematurely, to be completed the following day. The recent development of nonionic isomolar contrast agents (Visipaque™ [Iodixanol]; GE Healthcare, Chalfont St. Giles, United Kingdom) has been helpful, as larger amounts can be used. Selective catheterization of intercostal or lumbar arteries is done using a 4-Fr or, on rare occasions, a 5-Fr catheter. The most commonly used
Figure 14.6. Continued. (F) Vertebral artery angiogram (frontal plane) prior to superselective catheterization demonstrates the dilated and tortuous radicular and somatic branches (arrows) and the patchy collection of contrast material in the lateral aspect of the vertebral body. (G) Superselective microcatheter injections of the lower radicular artery (arrow) prior to PVA embolization shows the contrast-filled “lakes” filling within the lateral aspects of the vertebral body. (H) A 5-Fr catheter has been placed into the ascending cervical branch of the thyrocervical trunk (open arrow). The microcatheter is placed through the guide catheter into the radicular artery anastomosis feeding the ABC prior to PVA embolization (arrow). (I) Control angiogram through the vertebral artery after embolization shows nearly complete devascularization. Note that the microcatheter tip is still within the radicular artery (arrow). The mild vasospasm of the vertebral artery noted distal to the second radicular artery origin occurred after a balloon test occlusion (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
327
328
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
Figure 14.7. Continued. Pelvic images of a 50-year-old female who presented with lower back pain and sensory deficit associated with a recurrent giant cell cancer of the sacrum. A preoperative PVA embolization was performed to reduce the intraoperative blood loss. (A) Contrast-enhanced T1-weighted image shows the patchy and irregular enhancement of the sacral body and epidural space (arrows). The nerve roots are encased in the tumor tissue. (B) Pelvic angiogram shows the tumor blood supply from both internal iliac artery branches and the median sacral artery.
Recommended Technique for Spinal Angiography and Intervention
329
Figure 14.7. Continued. (C) Superselective microcatheterization of the right lateral sacral artery (arrow) prior to PVA embolization shows the diffuse tumor blush. (D) Superselective catheterization of the median sacral artery (arrow) prior to embolization shows the significant tumor blood supply through small anterior somatic branches.
330
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine Figure 14.7. Continued.(E) Left internal iliac artery angiogram shows the tumor supply through lateral sacral arteries (black arrow) and the iliolumbar artery (open arrow). (F) Control pelvic angiogram shows a complete tumor devascularization. Fibered coils were used to protect normal distal branches of the iliolumbar arteries (see artifacts superimposing on both internal iliac arteries) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
Neurophysiologic Monitoring During Embolization of Spinal Cord AV Shunts
catheters are the HS-2 and C2 catheter. Other useful shapes are C-1, Simmons or Sidewinder I or II, or, by some experts, a home-steamed 4-Fr catheter with a distal hook-shaped tip. An amount of 2–4 mL is injected within a second, and the angiogram is acquired in anterior– posterior projection. To reduce the time involved in placing the catheter and switching the contrast-filled syringe back and forth, it is recommended to have an assistant (if available) inject the contrast if an injector pump is not available. If injection by hand is preferred, the small syringe should be attached to a three-way stopcock, while another large attached syringe (20 or 30 mL) filled with contrast serves as a reservoir. The DSA acquisition should, especially in vascular malformations, be long enough to depict the normal venous return of the spinal cord. The normal arteriovenous transit time of spinal cord is approximately 18 s.72 In case an intervention is planned and a 6-Fr guide catheter is preferred for the coaxial microcatheter placement, it may be helpful to place a 6-Fr femoral sheath or, if the region of interest (ROI) is located higher, a long femoral sheath bypassing the often tortuous aortic/iliac system. Infrequently, the guide catheter may require to be changed over an exchange wire for a stable position within the intercostal or lumbar artery. It is easier and less traumatic to use hydrophilic-coated exchange glide wires when straightening the proximal part of the segmental arteries. With the introduction of 5-Fr guide catheters with larger lumens, a larger catheter may not be required. A range of microcatheters, including flow-guided catheters and micro-wires, are available for interventional procedures. The selection has to be tailored to the size of the vessel and the embolic material used. For diagnostic purposes, heparin is not given. For interventional procedures, heparin may be given but only on rare occasions to prevent inadvertent thrombosis, especially if catheters are navigated within the spinal cord vasculature. In selected cases of high-flow AVMs that have blood supply from anterior or posterior spinal arteries, the authors put the patient after embolization on aspirin and/or Plavix to prevent a retrograde thrombosis.
Neurophysiologic Monitoring During Embolization of Spinal Cord AV Shunts Neurophysiologic monitoring is becoming a standard of care in spinal surgery, and its availability and ease of use make it a very useful technique in neurointerventional procedures involving the spinal cord as well, especially when general anesthesia is being used. Monitoring of cortical Somatosensory Evoked Potentials (SSEP) and Motor Evoked Potentials (MEP) is very helpful in these cases. If the microcatheter tip cannot be placed close to the nidus of the AV shunt during embolization with liquid embolics, and there is a concern for damage to normal tissue, then provocative testing can be performed using 50 mg of amytal, with or without subsequent injection of 20–40 mg of lidocaine. If no change in the MEP or SSEP is seen, then embolization can proceed safely.73
331
332
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine
Conclusion Endovascular treatment of spinal vascular lesions and highly vascular tumors and metastases of the spine is very effective as a definite therapy or a preoperative intervention to enable surgery or to reduce intraoperative blood loss. A proper understanding of the spinal vascular anatomy, the vascular pathology, and a careful assessment of potential dangerous anastomoses as well as the selection of appropriate embolic materials are prerequisites to avoid ischemia to the spinal cord by inadvertent occlusion of blood supply to the anterior or posterior spinal arteries. References 1. Spetzler RF, Detwiler PW, Riina HA, Porter RW. Modified classification of spinal cord vascular lesions. J Neurosurg 2002;96(2 Suppl):145–156. 2. Zhang H, He M, Mao B. Thoracic spine extradural arteriovenous fistula: case report and review of the literature. Surg Neurol 2006;66 Suppl 1:S18– S23; discussion S23–S24. 3. Chul Suh D, Gon Choi C, Bo Sung K, Kim K-K, Chul Rhim S. Spinal osseous epidural arteriovenous fistula with multiple small arterial feeders converging to a round fistular nidus as a target of venous approach. AJNR 2004;25(1):69–73. 4. Oldfield EH, Bennett A III, Chen MY, Doppman JL. Successful management of spinal dural arteriovenous fistulas undetected by arteriography. J Neurosurg 2002;96(Suppl 2):220–229. 5. Eskandar, EN, Borges LF, Budzik RF Jr, Putman CM, Ogilvy CS. Spinal dural arteriovenous fistulas: experience with endovascular and surgical therapy. J Neurosurg 2002;96(Suppl 2):162–167. 6. Niimi Y, Berenstein A. Endovascular treatment of spinal vascular malformations. Neurosurg Clin North Am 1999;10(1):47–70. 7. Van Dijk JM, Ter Brugge KG, Willinsky RA, Farb RI, Wallace MC. Multidisciplinary management of spinal dural arteriovenous fistulas. Stroke 2002;33:1578–1583. 8. Atkinson JLD, Miller GM, Krauss WE, Marsh WR, Piepgras DG, Atkinson PP, Brown RD Jr, Lane JI. Clinical and radiographic features of dural arteriovenous fistula, a treatable cause of myelopathy. Mayo Clin Proc 2001;76:120–130. 9. Veznedaroglu E, Nelson PK, Jabbour PM, Rosenwasser RH. Endovascular treatment of spinal cord arteriovenous malformations. Neurosurgery 2006;59(5 Suppl 3):S202–S209; discussion S3–S13. 10. Bostroem A, Thron A, Hans FJ, Krings T. Spinal vascular malformations – typical and atypical findings. Zentralbl Neurochir 2007;68(4):205–213. 11. Rohany M, Shaibani A, Arafat O, Walker MT, Russell EJ, Batjer HH, Getch CC. Spinal arteriovenous malformations associated with KlippelTrenaunay-Weber syndrome: a literature search and report of two cases. AJNR Am J Neuroradiol 2007;28(3):584–589. 12. Oran I, Parildar M, Derbent A. Treatment of slow-flow (type I) perimedullary spinal arteriovenous fistulas with special reference to embolization. AJNR Am J Neuroradiol 2005;26(10):2582–2586. 13. Bao Y, Ling F. Classification and therapeutic modalities of spinal vascular malformations in 80 patients. Neurosurgery 1997;40(1):75–81. 14. Hodes JE, Merland JJ, Casasco A, Houdart E, Reizine D. Spinal vascular malformations: endovascular therapy. Neurosurg Clin North Am 1999; 10(1):139–152.
Conclusion 15. Mourier KL, Gobin YP, George B, Lot G, Merland JJ. Intradural perimedullary arteriovenous fistulae: results of surgical and endovascular treatment in a series of 35 cases. Neurosurgery 1993;32(6):885–891. 16. Rodesch G,Hurth M, Alvarez H et al. Spinal cord intradural arteriovenous fistulae: anatomic, clinical, and therapeutic considerations in a series of 32 consecutive patients seen between 1981 and 2000 with emphasis on endovascular therapy. Neurosurgery 2005;57(5):973–983; discussion 973–983. 17. Rodesch G, Hurth M, Alvarez H, Tadié M, Lasjaunias P. Classification of spinal cord arteriovenous shunts: proposal for a reappraisal. Neurosurgery 2002;15(2):374–380. 18. Ricolfi F, Gobin YP, Aymard A, Brunelle F, Gaston A, Merland JJ. Giant perimedullary arteriovenous fistulas of the spine: clinical and radiologic features and endovascular treatment. AJNR Am J Neuroradiol 1997; 18:677–687. 19. Aletich VA, Debrun GM, Koenigsberg R, Ausman JI, Charbel F, Dujovny M. Arteriovenous malformation nidus catheterization with hydrophilic wire and flow-directed catheter. AJNR Am J Neuroradiol 1997;18:929–935. 20. Touho H, Monobe T, Ohnishi H, Karasawa J. Treatment of type II perimedullary arteriovenous fistulas by intraoperative transvenous embolization: case report. Surg Neurol 1995;43(5):491–496. 21. Spetzler RF, Zabramski JM, Flom RA. Management of juvenile spinal AVMs by embolization and operative excision. Case report. J Neurosurg 1989;60:238–247. 22. Kim LJ, Spetzler RF. Classification and surgical management of spinal arteriovenous lesions: arteriovenous fistulae and arteriovenous malformations. Neurosurgery 2006;59(5 Suppl 3):S195–S201; discussion S3–S13. 23. Aminoff M, Logue V. Clinical features of spinal vascular malformations. Brain 1974;97:197–210. 24. Djindjian R. Clinical symptomatology and natural history of arteriovenous malformations of the spinal cord: a study of the clinical aspects and prognosis based on 150 cases. In Pia H, Djindjian R (eds): Spinal Angiomas: Advances in Diagnosis and Therapy. Berlin: Springer, 1978. 25. Chen CC, Wang CM, Chu NK, Wu KPH, Tang SFT, Wong AMK. Spinal cord arteriovenous malformation presenting as chest pain in a child. Spinal Cord 2008;46:456–458. 26. Biondi A, Merland JJ, Reizine D, Aymard A, Hodes JE, Lecoz P, Rey A. Embolization with particles in thoracic intramedullary arteriovenous malformations: long-term angiographic and clinical results. Radiology 1990;177:651–658. 27. Biondi A, Merland JJ, Hodes JE, Pruvo JD, Reizine D. Aneurysms of spinal arteries associated with intramedullary arteriovenous malformations. I. Angiographic and clinical aspects. AJNR Am J Neuroradiol 1992;13: 913–922. 28. Biondi A, Merland JJ, Hodes JE, Aymard A, Reizine D. Aneurysms of spinal arteries associated with intramedullary arteriovenous malformations. II. Results of AVM endovascular treatment and hemodynamic consideration. AJNR Am J Neuroradiol 1992;13:923–931. 29. Cullen S, Krings T, Ozanne A, Alvarez H, Rodesch G, Lasjaunias P. Diagnosis and endovascular treatment of pediatric spinal arteriovenous shunts. Neuroimaging Clin N Am 2007;17(2):207–221. 30. Sinclair J, Chang SD, Gibbs IC, Adler Jr JR. Multisession CyberKnife radiosurgery for intramedullary spinal cord arteriovenous malformations. Neurosurgery 2006;58(6):1081–1089; discussion 1081–1089. 31. Kocak A, Ates O, Cayli S, Sara K. Isolated posterior spinal artery aneurysm. Br J Neurosurg 2006;20(4):241–244.
333
334
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine 32. Gonzalez LF, Zabramski JM, Tabrizi P, Wallace RC, Massand MG, Spetzler RF. Spontaneous spinal subarachnoid hemorrhage secondary to spinal aneurysms: diagnosis and treatment paradigm. Neurosurgery 2005;57(6):1127– 1131; discussion 1127–1131. 33. Maraire JN, Awad IA. Cavernous malformations: natural history and indications for treatment. In Batjer HH (eds): Cerebrovascular Disease. Philadelphia: Lippincott-Raven, 1997:669–677. 34. Canavero S, Pagni CA, Duca S, Bradac GB. Spinal intramedullary cavernous angiomas: a literature metaanalysis. Surg Neurol 1994;41:381–388. 35. Krings T, Mull M, Gilsbach JM, Thron A. Spinal vascular malformations. Eur Radiol 2005;15(2):267–278. 36. Mocco J, Laufer I, Mack WJ, Winfree CJ, Libien J, Connolly Jr ES. An extramedullary foramen magnum cavernous malformation presenting with acute subarachnoid hemorrhage: case report and literature review. Neurosurgery 2005;56(2):E410; discussion E410. 37. Minh NH. Cervicothoracic spinal epidural cavernous hemangioma: case report and review of the literature. Surg Neurol 2005;64(1):83–85; discussion 85. 38. Sze G. Neoplastic disease of the spine and spinal cord. In: Atlas SW (ed): Magnetic Resonance Imaging of the Brain and Spine. Philadelphia: Lippincott-Raven, 1996:1377–1379. 39. Steinmetz MP, Claybrooks R, Krishnaney A, Prayson RA, Benzel EC. Surgical management of osseous hemangioblastoma of the thoracic spine: technical case report. Neurosurgery 2005;57(4 Suppl):E405; discussion E405. 40. Eskridge JM, McAuliffe W, Harris B, Kim DK, Scott J, Winn HR. Preoperative endovascular embolization of craniospinal hemangioblastomas. AJNR Am J Neuroradiol 1996;17:525–531. 41. Tampieri D, Leblanc R, Ter Brugge KG. Pre-operative embolization of brain and spinal hemangioblastomas. Neurosurgery 1993;33(3):502–504. 42. Pia H, Djindjian M. Spinal Angiomas: Advances in Diagnosis and Therapy. New York: Springer, 1978. 43. Krings T, Chng SM, Ozanne A, Alvarez H, Rodesch G. Lasjaunias PL. Hereditary hemorrhagic telangiectasia in children: endovascular treatment of neurovascular malformations: results in 31 patients. Neuroradiology 2005;47(12):946–954. 44. Fulbright RK, Chaloupka JC, Putman CM, Sze GK, Merriam MM, Lee GK, Fayad PB, Awad IA, White Jr RI. MR of hereditary hemorrhagic telangiectasia: prevalence and spectrum of cerebrovascular malformations. AJNR Am J Neuroradiol 1998;19:477–484. 45. Garcia-Monaco R, Taylor W, Rodesch G, Alvarez H, Burrows P, Coubes P, Lasjaunias P. Pial arteriovenous fistula in children as presenting manifestation of Rendu-Osler-Weber disease. AJNR Am J Neuroradiol 1995;37(1):60–64. 46. Jacob AG, Driscoll DJ, Shaughnessy WJ, Stanson AW, Clay RP, Gloviczki P. Klippel-Trenaunay syndrome: spectrum and management. Mayo Clinic Proc 1998;73(1):28–36. 47. Hauck EF, Nauta HJ. Spontaneous spinal epidural arteriovenous fistulae in neurofibromatosis type-1. Surg Neurol 2006;66(2):215–221. 48. Doppman, JL, Oldfield EH, Heiss JD. Symptomatic vertebral hemangiomas: treatment by means of direct intra-lesional injection of ethanol. Radiology 2000;214:341–348. 49. Fox MW, Onofrio BM. The natural history and management of symptomatic and asymptomatic vertebral hemangiomas. J Neurosurg 1993;78:36–45. 50. Jayakumar PN, Vasudev MK, Srikanth SG. Symptomatic vertebral hemangiomas: endovascular treatment of 12 patients. Spinal Cord 1997;35: 624–628.
Conclusion 51. Yuksel M, Yuksel KZ, Tuncel D, Zencirci B, Bakaris S. Symptomatic vertebral hemangioma related to pregnancy. Emerg Radiol 2007;13(5):259–263. 52. Djindjian R, Merland JJ, Djindjian M et al. Vertebral hemangiomas and metameric angiomatosis (Cobb’s Syndrome). In Nadjmi M, Piegras U, Vogelsang H (eds): Angiography of Spinal Column and Spinal Cord Tumors. Stuttgart: Georg Thieme, 1981:141. 53. Hadjipavlou A, Gaitanis I, Kakavelakis K, Katonis P. Balloon kyphoplasty as a single or as an adjunct procedure for the management of symptomatic vertebral haemangiomas. J Bone Joint Surg Br 2007;89(4):495–502. 54. Reizine D, Laouiti M, Guimaraens L, Riche MC, Merland JJ. Vertebral arteriovenous fistulas – clinical presentation, angiographical appearance and endovascular treatment. A review of twenty-five cases. Ann Radiol 1985;28:425–438. 55. Berenstein A, Lasjaunias P. Surgical Neuro-angiography. Vol 5. Berlin: Springer, 1992:125–127. 56. Bonakdarpour A, Levy WM, Aegerter E. Primary and secondary aneurismal bone cyst: a radiological study of 75 cases. Radiology 1978;126:75–82. 57. Ameli NO, Abbassioun K, Saleh H, Eslamdoost A. Aneurysmal bone cysts of the spine; report of 17 cases. J Neurosurg 1985;63:685–690. 58. Lichtenstein L. Aneurysmal bone cyst. A pathological entity commonly mistaken for giant-cell tumor and occasionally for hemangioma and osteogenic carcinoma. Cancer 1950;3:279–289. 59. Cigala F, Sadile F. Arterial embolization of aneurysmal bone cysts in children. Bull Hosp Joint Dis 1996;54(4):261–264. 60. Radanovic B, Šimunić S, Stojanović J, Orlić D, Potočki K, Oberman BB. Therapeutic embolization of aneurysmal bone cyst. Cardiovasc Intervent Radiol 1989;12(6):313–316. 61. Broaddus WC, Grady MS, Delashaw JB, Ferguson RD, Jane JA. Preoperative super-selective arteriolar embolization: a new approach to enhance resectability of spinal tumors. Neurosurgery 1990;27:755–759. 62. Gellad FE, Sadato N, Numaguchi Y, Levine AM. Vascular metastatic lesions of the spine: preoperative embolization. Radiology 1990;176:683–686. 63. Hilal SK, Michelsen JW. Therapeutic percutaneous embolization for extra-axial vascular lesions of the head, neck, and spine. J Neurosurg 1975;43:275–287. 64. King GJ, Kostuik JP, McBroom RJ, Richardson W. Surgical management of metastatic renal carcinoma of the spine. Spine 1991;16:265–271. 65. Jensen ME, Hendrix LE, Dion JE et al. Preoperative and palliative embolization of vertebral body metastases. Proceedings of the 31st Annual Meeting of the American Society of Neuroradiology. Vancouver, BC, Canada, 1993. 66. Gangi A, Kastler BA, Dietemann JL. Percutaneous vertebroplasty guided by a combination of CT and fluoroscopy. AJNR Am J Neuroradiol 1994; 15:83–86. 67. Chiras J, Cognard C, Rose M, Dessauge C, Martin N, Pierot L, Plouin PF. Percutaneous injection of an alcoholic embolizing emulsion as an alternative preoperative embolization for spine tumor. AJNR Am J Neuroradiol 1993;14:1113–1117. 68. Heiss JD, Doppman JL, Oldfield JH. Brief report: relief of spinal cord compression from vertebral hemangioma by intralesional injection of absolute ethanol. N Engl J Med 1994;331:508–511. 69. Cotton A, Deramond H, Cortet B. Preoperative percutaneous injection of methyl methacrylate and n-butyl cyanoacrylate in vertebral hemangiomas. AJNR Am J Neuroradiol 1996;17:137–142.
335
336
Chapter 14 Endovascular Treatment of Vascular and Nonvascular Diseases of the Spine 70. Lai PH, Pan HB, Yang CF, Yeh LR, Su SS, Lee KW, Weng MJ, Wu MT, Liang HL, Chen CK. Multi-detector row computed tomography angiography in diagnosing spinal dural arteriovenous fistula: initial experience. Stroke 2005;36(7):1562–1564. 71. Saraf-Lavi E, Bowen BC, Quencer RM, Sklar EML, Holz A, Falcone A, Latchaw RE, Duncan R, Wakhloo A. Detection of spinal dural arteriovenous fistula with MR imaging and contrast-enhanced MR angiography: sensitivity, specificity, and prediction of vertebral level. AJNR Am J Neuroradiol 2002;23:858–867. 72. Launey M, Chiras J, Bories J. Angiography of the spinal cord: venous phase. AJNR Am J Neuroradiol 1979;6:287–315. 73. Niimi Y, Sala F, Deletis V, Setton A, de Camargo AB, Berenstein A. Neurophysiologic monitoring and pharmacologic provocative testing for embolization of spinal cord arteriovenous malformations. AJNR Am J Neuroradiol 2004;25(7):1131–1138.
15 Kyphoplasty: Balloon Assisted Vertebroplasty John M. Mathis, Charles H. Cho, and Wayne J. Olan
Introduction Pain relief after Percutaneous Vertebroplasty (PV) has been reported in 85–90% of patients with Vertebral Compression Fractures (VCFs),1–4 but the deformity of the vertebral body or the subsequent kyphosis (usually related to multiple compressions) was not a primary focus of this procedure when it was first introduced5 (For a more exhaustive treatment of vertebroplasty, see Chap. 12). Biomechanically, kyphosis shifts the patient’s center of gravity forward, rendering the patient off-balance and at increased risk for a fall. This change in a patient’s center of gravity also creates additional stress on the vertebrae, increasing the risk of additional fractures.6 The kyphosis caused by VCFs in the lumbar or thoracic region decreases vital capacity in the lungs, which in turn accentuates restrictive lung disease.7 Leech et al.8 reported a 9% average decrease in forced vital capacity per osteoporotic compression fracture in the thoracic region. In addition, these fractures can lead to gastrointestinal difficulties. Increasing kyphosis may cause the ribs to increase pressure on the abdomen, creating a sensation of bloating that may lead to early satiety, decreased appetite, and malnutrition.9 There is a significant decrease in the life expectancy of patients with VCFs. In a retrospective study, Cooper et al.10 found that the 5-year survival rate for patients with VCFs was lower than that for patients with hip fractures. A prospective study by Kado et al.11 showed that patients with VCFs had a 23% higher mortality than age-matched controls. The increased mortality was thought to result from pulmonary causes, including pneumonia and chronic obstructive pulmonary disease. Kyphoplasty (KP) is a trade name for the generic procedure best termed “balloon assisted vertebroplasty.” KP was developed in an attempt to reduce the deformity of the vertebral body (Figure 15.1a) and subsequent kyphosis while at the same time providing pain relief
From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_15, © Springer Science + Business Media, LLC 2010
337
338 Chapter 15 Kyphoplasty: Balloon Assisted Vertebroplasty Figure 15.1. (A) Lateral radiograph demonstrates compressed vertebra (black arrow) due to osteoporosis resulting in loss of vertebral height. (B) Lateral fluoroscopic image taken during balloon assisted vertebroplasty. Balloons (overlapping in this image) are inflated (white arrow).
similar to that of PV. These intentions (aside from pain relief) remain largely unproven, though they are avidly touted by the manufacturer. A recent (2007) position statement by multiple societies that routinely use
Patient Selection
the device (American Association of Neurologic Surgery, Congress of Neurologic Surgery, American Society of Interventional and Therapeutic Neuroradiology, and American Society of Spine Radiology) states that these societies find “no proved advantage of KP relative to PV for pain relief, vertebral height restoration, or complication rate.”12 The kyphoplasty procedure consists of inserting a balloon-like device (referred to as a bone tamp) percutaneously into a compressed vertebral body, inflating the device, and attempting to elevate the endplates and restore vertebral body height (Figure 15.1b). This is followed by cement injection, in the same manner the cement is used in PV. In theory, this procedure was expected to improve vital lung capacity and gastrointestinal function by reducing the kyphosis associated with VCFs.13 As stated earlier, these results have not been shown to occur more than that found with PV.
Patient Selection The selection criteria for KP are similar but not equivalent to those for PV (see Chap. 12). As with PV, the patient selection process includes obtaining detailed history and physical examination. The patient’s symptoms need to be linked to the VCF. Evaluation of the patient’s films (radiographs, Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and bone scans) should correlate symptoms with the fracture location and image characteristics. Balloon-assisted vertebroplasty, or KP, has been used to treat osteoporotic VCFs. It may also be used in patients with vertebral body involvement from neoplastic disease such as plasmocytoma or multiple myeloma. The likelihood of restoring vertebral body height depends largely on the density of the bone, acuity of the fracture, and whether the fracture remains mobile (able to be moved by change in position or pressure exerted internally). Fractures treated within 1–3 weeks of the event are much less likely to have experienced substantial healing, and these provide the best opportunity for height restoration. However, height restoration can be seen with PV and generally is found to be of the same order of magnitude as that achieved with KP. The exclusion criteria for balloon kyphoplasty are also very similar to those used for PV and include: 1. VCFs that are not painful or that are not the primary source of pain 2. The presence of osteomyelitis or systemic infection 3. Retropulsed bone fragments, or 4. An epidural extension of tumor. The latter two factors must be considered because balloon inflation for the KP procedure could force material into the spinal canal, causing cord compression. There are also relative contraindications to KP. First, there must be sufficient residual height for the instruments used with kyphoplasty to be inserted into the compressed vertebral body. Thus, although PV may be performed in a severely compressed vertebral body, KP of
339
340 Chapter 15 Kyphoplasty: Balloon Assisted Vertebroplasty Figure 15.2. Sagittal MRI image showing an extreme compression (arrow). There is no room left in this vertebra to insert the balloon for KP (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. ImageGuided Spine Interventions. New York: Springer Science + Business Media, 2004).
the same vertebral body may not be technically feasible (Figure 15.2). Second, small pedicles may also be a limiting technical factor because the instruments used for KP are somewhat larger than those used for PV. When the pedicles appear to be too small to accommodate the KP instruments, a parapedicular approach can be utilized. Kyphoplasty can be performed safely from L5 to T7 in most patients.14
Technique KP is an extension of PV and, as stated earlier, has been termed “balloon-assisted vertebroplasty.” KP relies on the same mechanism for creating pain relief as that achieved by PV (i.e., bone augmentation with the injection of bone cement). The patient is positioned on the fluoroscopic operating table in a prone position. It is important to avoid an antecubital intravenous line. High-resolution C-arm or biplane fluoroscopy is essential when one is performing KP or PV. The patient is positioned so that the spine is located at the isocenter of the C-arm. The fracture is then identified fluoroscopically. The approach is usually bilateral transpedicular; however, a single posterolateral approach can be used for the large lower lumbar vertebrae (almost always at L5, less commonly at L2–L4). An extrapedicular approach must be used when the pedicles are too small to accommodate the kyphoplasty instruments (usually in the mid- or upper thoracic regions). This extrapedicular approach carries additional risks to those found with the transpedicular approach (see Chaps. 1 and 12 for more details).
Technique
Localization of the pedicles is performed in a manner similar to that used for PV. A posterior approach with slight ipsilateral obliquity of 10–25° is preferred. The medial wall of the pedicle must be well visualized and avoided during needle placement. After sterile preparation and draping of the patient and after the fluoroscopy equipment also has been covered in sterile fashion, local anesthetic is injected into the patient’s skin, subcutaneous tissue, and periosteum of the bone. Typically a 25-gauge needle is used, but a longer spinal needle can be used to reach the periosteum. Most patients require only local anesthesia and conscious sedation. As in PV, the key to local anesthesia is the extension of the anesthetic to the periosteum of the pedicle. Patients who cannot lie in a prone position may be candidates for general anesthesia. Prophylactic intravenous antibiotics, typically 1 g of cefazolin, are administered. The KP procedure requires an 8–11 gauge (4–6 in.) bone entry needle, a scalpel, a kyphoplasty kit, inflatable balloon tamps, and bone cement appropriately opacified and approved for KP or PV (Figure 15.3). The procedure begins by directing the entry needle into the bone under fluoroscopic guidance. For a transpedicular approach, the needle is directed through the pedicle to the posterior aspect of the vertebral body (Figure 15.4). For very small pedicles, an extrapedicular approach
Figure 15.3. Some of the materials needed for the kyphoplasty procedure. The inflator and bone cement are not shown (for inflator, see Figure 15.8) (Some instruments: Kyphon, Inc., Sunnyvale, CA).
341
342 Chapter 15 Kyphoplasty: Balloon Assisted Vertebroplasty
F igure 15.4. (A) Slight oblique radiograph (from anteroposterior direction) shows the bone introduction needle (white arrow) penetrating the pedicle (small arrows). (B) Lateral image showing the cannula and K-wire tip just at the pedicle–posterior vertebral body junction (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
can be used. The needle targets a starting point just superior and lateral to the pedicle (Figure 15.5). If a single posterolateral approach is chosen, the trajectory can be established along a posterolateral path similar to that used for discography. This approach is appropriate for the larger lumbar vertebrae, especially L5. One must be cautious to avoid injuring the exiting nerve roots, and the beginning point must not be so far lateral that puncture of the bowel or kidney results. With the posterolateral approach, the drill should cross the midline of the vertebra on anteroposterior and lateral views. Oblique views should also be used to confirm proper positioning. The advantage of the single posterolateral approach is the time saved by placing one balloon instead of two; the disadvantage is a reduction of the working surface area of the inflatable balloon tamp. After needle insertions, the trochar is removed. A Kirschner wire (K-wire) is then directed through the cannula and into the bone. The needle cannula is removed, leaving the K-wire in place. A blunt dissector is then fitted over the K-wire and, under fluoroscopic guidance, into the bone to be situated at the level of the K-wire. In a transpedicular
Technique
343
Figure 15.5. (A) Axial drawing demonstrating the trajectory of the needle for a parapedicular approach. The needle follows the junction of the rib and transverse process of the vertebra and enters the vertebral body along the lateral margin of the pedicle. (B) Lateral view of the parapedicular approach. Note that the needle has a downward angle that allows one to go over the transverse process on the way to the lateral pedicle margin (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
approach, the K-wires and blunt dissector are directed to the posterior third of the vertebral body. One should manipulate the K-wire with the same caution that one would use for a guide wire in the vascular system. The operating physician should always have control of the proximal end of the K-wire because the sharp tip could easily penetrate soft bone and breach the anterior vertebral cortex. A skin incision is then made to accommodate the working cannula, which is advanced through the soft tissues over the blunt dissector and through the pedicle to rest along the posterior aspect of the vertebral body. A plastic handle can be placed on the hub of the cannula to advance it manually into the vertebral body, or a mallet can be used to tap the plastic handle, driving the cannula into the vertebral body. If there is considerable resistance to placing the working cannula, the cannula’s handle can be rotated in an alternating clockwise/counterclockwise (screwing) motion to help breach the cortex and facilitate advancement. If using the mallet, one must be careful to direct the blows onto the handle; inadvertently striking the K-wire or blunt dissector might drive the object deeper into the vertebra. Next, the K-wire and blunt dissector are removed, leaving the working cannula in place. A 3-mm drill is advanced through the cannula,
344 Chapter 15 Kyphoplasty: Balloon Assisted Vertebroplasty Figure 15.6. The bone drill (arrow) in the vertebral body (introduced through the bone cannula) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
and multiplanar fluoroscopy is used to recheck the orientation of the working cannula. Then the drill is directed ideally along a slightly posterolateral to anteromedial trajectory into the vertebra until the tip of the drill is 3–4 mm posterior to the anterior margin of the vertebral body, or at least within the anterior third of the vertebral body (Figure 15.6). If the fracture involves the superior aspect of the vertebral body, the drill must be directed somewhat inferiorly to the midline of the vertebral body. If the fracture is along the inferior aspect of the vertebra, the drill must be directed superiorly to the midline of the vertebra. Extreme caution should be used to avoid breaching the anterior cortex of the vertebral body with the drill. For bilateral transpedicular or extrapedicular approaches, the sequence of events is repeated on the contralateral side. One may avoid this multi-step process by using a newer introductory cannula and stiletto that can initially create a channel large enough to accept the balloon or bone tamp. Express™ needles (KyphX® Express™, Kyphon Inc., Sunnyvale, CA) are available in 11 F, and One-Step™ needles (KyphX® One-Step™, Kyphon Inc., Sunnyvale, CA) are available in 8 F sizes. These needles have an
Technique
345
outer introducer (with a working handle) and inner stylet (diamond or bevel tip). The procedure is very similar to needle placement in vertebroplasty. The introducer system is placed through the pedicle with the tip at the desired position in the vertebral body, similar to the trochar described earlier. The inner stylet is removed. A bone biopsy needle can be introduced through the cannula if a sample is needed. The inflatable balloon tamp is available in different sizes. Each balloon has markers to delineate its distal and proximal extents (Figure 15.7). These markers are also radiopaque and easily visualized under fluoroscopy. Exact™ balloons (KyphX® Exact™, Kyphon Inc., Sunnyvale, CA), inflating to one side of the needle, are available if the needle position is deviated and the desire is to create a cavity toward one side of the needle. The bone tamps are then prepared for inflation. Air is purged from the balloons, and the reservoir of an angioplasty injection device (incorporating a pressure monitor) is filled with 10 mL of diluted iodine contrast material (Figure 15.8). If the patient has an allergy to iodine, gadolinium can be substituted. The uninflated balloon tamps are inserted through the working cannulas under fluoroscopy and directed to the most anterior extent Figure 15.7. Lateral radiograph, with black markers pointing to the distal and proximal markers of the balloon (bone tamp) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
346 Chapter 15 Kyphoplasty: Balloon Assisted Vertebroplasty
Figure 15.8. Inflator (Kyphon, Inc., Sunnyvale, CA) used to expand the balloon. It will be filled with a mixture of radiographic contrast and saline so the balloon can be visualized during inflation.
of the vertebral body. If the clinician feels any resistance in the passageway of the drilled hole, curettes (KyphX® Latitude II™, Kyphon Inc., Sunnyvale, CA) are available if bone trabeculations need to be destroyed prior to balloon expansion. These curette tips can be angled from 0 to 90° with a squeeze of the handle to create the desired angle and rotated side-to-side to create the desired cavity. The balloon tamp can then be inserted without difficulty. Balloon inflation should be performed slowly. Inflation via the injection device is begun under continuous fluoroscopy, increasing balloon pressure to approximately 50 psi to secure the balloon in position. The stiffening wire is withdrawn from the shaft of the bone tamps, and the volume of contrast media in the reservoir is recorded. The balloons are progressively inflated by half-milliliter increments (Figure 15.9), with frequent pauses to check for pressure decay, which occurs as the adjacent cancellous bone yields and compacts. If the bone is osteoporotic, pressure decay may be immediate. If the bone is quite dense, there may be little or no pressure decay, even at pressures up to 180 psi. The balloon system is rated to 180 psi, with a practical maximum of 220 psi. Even with slow inflation, pressures higher than 220 psi have been achieved in dense bone.15 If a balloon ruptures, it is simply withdrawn through the working cannula and replaced. The possible end points of inflation are: 1. Restoration of the vertebral body height to normal 2. Fattening of the balloon against an endplate without accompanying height restoration 3. Contact with a lateral cortical margin
Technique
Figure 15.9. (A) Anteroposterior image reveals two inflated balloons (arrows) during kyphoplasty. (B) Lateral image shows the two inflated balloons (arrows) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
347
348 Chapter 15 Kyphoplasty: Balloon Assisted Vertebroplasty
4. Inflation without further pressure decay, and 5. Reaching the maximum volume of the balloon or maximum pressure. The operating physician must maintain both visual and manual control throughout the entire inflation process and should record the amount of fluid used to inflate the balloon when the end point has been achieved. This volume indicates the size of the cavity that has been created, and it will serve as an estimate of the amount of cement to be delivered. If substantial height restoration has not been achieved, careful repositioning of the bone tamps and reinflation may be helpful. Once adequate inflation has been achieved, the cement is mixed in a manner similar to that for PV. The cement mixture is transferred to a 10-mL syringe that is used to fill a series of 1.5-mL bone filler devices. The volume of cement for injection is approximately 1 mL more than the volume of the cavity created by each inflatable balloon tamp.16 If a quantity of cement is equal to or less than the volume of the cavity, the vertebra will not be reinforced and will recollapse quickly. Without adequate filling, the vertebra treated with KP is actually less strong than one treated with PV, and any height gained may be subsequently lost.17 Once PMMA cement has undergone transition from a liquid to a cohesive, doughy consistency (about 3–4 min after mixing), the bone filler devices are passed through the working cannula and into the anterior aspect of the vertebral cavities. Cortoss (Orthovita; Malvern, Pa) is also approved for KP. This new biologic cement bonds to bone, has less exotherm (compared to PMMA) and is ready to use immediately upon mixing (see “Cement Selection” section in the chapter on PV). The cavity is then filled with cement, proceeding from the anterior to the posterior aspect of the vertebra. Continuous fluoroscopic monitoring is maintained to identify the leakage of cement into the spinal canal, paraspinous veins, inferior vena cava, or disc space. One hypothetical advantage of KP over PV is that the former affords a low-pressure cement delivery into the cavity created by the inflatable balloon tamp. However, laboratory measurements show no difference in the pressure inside the vertebra when KP is compared to PV, proving this was a misconception.18 Some operating physicians prefer to fill one cavity first, leaving the contralateral balloon inflated as a supporting strut. This maneuver may be effective in maintaining any height elevation that has been achieved. When cement filling of the cavity has been confirmed fluoroscopically from both lateral (Figure 15.10a) and anteroposterior views, the bone filler devices are withdrawn partially to allow complete filling of the cavity; then they are used to tamp the bone cement in place before being withdrawn completely. The cannulas are then rotated (so they are not cemented in the bone) and removed, and hemostasis is obtained at the incision site by using manual pressure. Steri-Strips™ (3M, St. Paul, MN) are usually sufficient for wound closure. The patient remains prone on the table and is not moved until the remaining cement in the mixing bowl has hardened completely. The usual time frame for KP is 35–45 min, which compares favorably with the 20–25 min per level required for PV. In denser bone, the balloons may take longer to respond to small incremental increases in pressure. The follow-up and postoperative procedures for KP are identical to those for PV. At some institutions, KP and PV are performed on an outpatient basis unless the patient is extremely frail, or unless the pro-
Results
349
Figure 15.10. (A) Lateral radiograph shows the pretreatment appearance of the compression fracture. (B) Postkyphoplasty image. A small cement leak occurred anteriorly but was asymptomatic. There is mild height restoration between 3 and 4 mm (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM ed. Image-Guided Spine Interventions. New York: Springer Science + Business Media, 2004).
cedure is performed at the end of the day, and staffing issues make it easier to keep the patient overnight for discharge the next morning. Outpatients are observed for 3–4 h after the procedure. General anesthesia and overnight stay for KP have been used to allow hospitals to recover the cost of the KP kit and devices that are approximately ten times the cost of materials to perform PV on the same patient. The higher cost of KP has not been associated with increased safety, increased pain relief, or increased height restoration. Kyphoplasty is a technically demanding procedure. Safe performance requires a high level of skill and high-quality imaging equipment. One should not perform this procedure without being an expert in clinical and radiographic spinal anatomy, without having completed a kyphoplasty course with expert instructors, and without imaging equipment that is capable of clearly delineating key bony landmarks, particularly the pedicles, the cortices, and the spinous processes.
Results The case demonstrated in Figure 15.10 is an average result. The patient had good pain relief (similar to PV) and a modest amount of height was
350 Chapter 15 Kyphoplasty: Balloon Assisted Vertebroplasty
restored (approximately 3–4 mm; Figure 15.10b). The clinical significance of this modest height restoration still needs review. PV may also be associated with mild height restoration (of the same magnitude as that found with KP) and is excellent at relieving pain.19,20 With pain relief following both PV and KP, patients get reduction in kyphosis, as they are able to stand straight and to support their body weight without pain. Reproducible outcome analysis is needed to understand the significance (or lack thereof) of the differences between PV and KP, as the results of pain relief and kyphosis correction seem very similar at present. KP is a relatively new procedure, and good, controlled studies comparing PV to KP or KP to conservative therapy are not readily available. One early outcome study of 70 vertebral bodies treated in 30 patients reported an average restoration of 2.9 mm of height.14 When the treated vertebrae were separated into two groups, 70% gained an average of 4.1 mm (46.8% height restoration), whereas 30% regained no height. Asymptomatic cement extravasations occurred in 8.6% of the levels treated, a rate similar to that reported for PV used for osteoporotic VCFs. Perioperative complications for KP include one myocardial infarction (3.3%) and two patients who sustained rib fracture during positioning (6.7%). Pradhan et al.21 found more height restoration in the center of vertebra treated with KP with the margins achieving similar height gain as that of Lieberman.14 The central height gain was lost in the compliant disc and did not result in significant kyphosis correction.21 Reporting on preliminary results from 340 patients from a multicenter registry, Garfin et al.22 indicated a height restoration similar to that reported earlier. 14 There was a serious complication rate of 1.2% that included permanent cord damage associated with cement leakage (Figure 15.11).22 It should be noted, however, that these results were anecdotally reported in a literature review regarding KP and PV. Nussbaum and colleagues, looking at complications reported to the FDA medical device-related complication website, found that KP actually had a greater than 10 time permanent complication rate when compared to traditional PV.23 These early clinical reports do not offer substantial data for complete evaluation of the procedure’s efficacy. Although KP appears to be able to restore height in some cases (Figure 15.10), it is unknown whether the typically 3–4 mm of height restoration results in clinically significant benefit. Furthermore, it is unknown whether height restoration results in kyphosis reduction and subsequently in increased lung capacity. A long-term follow-up study determining the benefits of KP versus PV is needed, but in reality, this will be a difficult task. Both procedures provide similar pain relief and, in experienced hands, similar risk. In the presence of pain relief, the benefits of height restoration will most likely remain empirical. Although the exact mechanism of pain relief is unknown, it is believed that both procedures provide pain relief secondary to fracture stabilization via cement injection. Though results seem similar between PV and KP, there is no doubt that KP is many times more expensive. The cost of KP, therefore must be evaluated by real gain achieved between the two procedures.
Biomechanical Investigations
Figure 15.11. Computed tomographic image of a complication occurring post Kyphoplasty. There is a large cement leak (white arrows) into the spinal canal. In this case, permanent spinal cord injury occurred due to cord compression by the cement.
Biomechanical Investigations Reports indicate that height restoration has the potential benefit of reducing postfracture kyphosis and its associated sequelae.7–9,24,25 The magnitude of height restoration mentioned in the preliminary clinical reports discussed earlier is similar to that measured ex vivo.15 In the ex vivo study by Belkoff et al.,15 average actual height restoration (average of six height measurements made circumferentially about the vertebral body) was 2.5 ± 0.7 mm7–9,24 It is important to note that in this ex vivo study of osteoporotic vertebral bodies (compressed to create simulated fractures and repaired with PV) half of the compressed height recovers elastically,15 a phenomenon similar to that reported in vivo.26 The actual height restoration seems to range from 2.5 to 3.5 mm, values similar to those reported clinically.14 The vertebral strength after PV has been found to be stronger than that of KP by Wilcox.27 Kim and colleagues also found that height initially gained with KP was lost with cyclic loading.17 This result indicates that vertebra treated with PV are actually taller than those treated with KP (after cyclic loading). One of the stated theoretical advantages of kyphoplasty over standard PV is that the former might permit the injection of cement under lower pressures. This has now been shown to not be the case.18 Indeed, Tomita and colleagues found that pressures inside the vertebra were the same in both PV and KP.28
351
352 Chapter 15 Kyphoplasty: Balloon Assisted Vertebroplasty
Conclusion Both PV and KP seem to provide the same pain relief from vertebral compression fractures and, in experienced hands, approximately the same risk. Because the pain relief from both procedures appears to be similar as are variables such as pulmonary function, gastrointestinal issues, and kyphosis change, it will be difficult to distinguish the two procedures based on clinical outcomes. Any claimed benefits of KP over PV remain to be proven. This similarity has now been addressed by multiple medical associations that use the procedure in a 2007 position statement that reports that they find “no proved advantage of KP relative to PV for pain relief, vertebral height restoration, or complication rate.”17 This leaves the use of KP and its high cost in question for most cases for which PV seems to offer similar benefits but at much lower cost. References 1. Galibert P, Deramond H, Rosat P, Le Gars D. [Preliminary note on the treatment of vertebral angioma by percutaneous acrylic vertebroplasty.] Neurochirurgie 1987;33(2):166–168. 2. Jensen ME, Evans AJ, Mathis JM, Kallmes DF, Cloft HJ, Dion JE. Percutaneous polymethylmethacrylate vertebroplasty in the treatment of osteoporotic vertebral body compression fractures: technical aspects. Am J Neuroradiol 1997; 18(10):1897–1904. 3. Gangi A, Kastler BA, Dietemann JL. Percutaneous vertebroplasty guided by a combination of CT and fluoroscopy. Am J Neuroradiol 1994;15(1):83–86. 4. 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 percentage of lesion filling and the leakage of methyl methacrylate at clinical follow-up. Radiology 1996;200(2):525–530. 5. Mathis JM, Deramond H, Belkoff SM. Percutaneous Vertebroplasty. New York: Springer, 2002. 6. White AA, Panjabi MM. Clinical Biomechanics of the Spine. 2nd ed. Philadelphia: Lippincott, 1990. 7. Schlaich C, Minne HW, Bruckner T, Wagner G, Gebest HJ, Grunze M, Ziegler R, Leidig-Bruckner G. Reduced pulmonary function in patients with spinal osteoporotic fractures. Osteoporos Int 1998;8(3):261–267. 8. Leech JA, Dulberg C, Kellie S, Pattee L, Gay J. Relationship of lung function to severity of osteoporosis in women. Am Rev Respir Dis 1990;141(1):68–71. 9. Silverman SL. The clinical consequences of vertebral compression fracture. Bone 1992;13(Suppl 2):S27–S31. 10. Cooper C, Atkinson EJ, O’Fallon WM, Melton LJ. Incidence of clinically diagnosed vertebral fractures: a population-based study in Rochester, Minnesota, 1985–1989. J Bone Miner Res 1992;7(2):221–227. 11. Kado DM, Browner WS, Palermo L, Nevitt MC, Genant HK, Cummings SR. Vertebral fractures and mortality in older women: a prospective study. Study of Osteoporotic Fractures Research Group. Arch Intern Med 1999;159(11):1215–1220. 12. Jensen ME, McGraw K, Cardella JF, Hirsch JA. Position statement on percutaneous vertebral augmentation: a consensus statement developed by the American Society of Interventional and Therapeutic Neuroradiology, Society of Interventional Radiology, American Association of Neurological
Conclusion Surgeons/Congress of Neurological Surgeons, and American Society of Spine Radiology. J Vasc Interv Radiol 2007;18:325–330. 13. Wong WH, Olan WJ, Belkoff SM. Balloon kyphoplasty. In: Mathis JM, Deramond H, Belkoff SM, eds. Percutaneous Vertebroplasty. New York: Springer, 2002:109–124. 14. Lieberman IH, Dudeney S, Reinhardt M-K, Bell G. Initial outcome and efficacy of kyphoplasty in the treatment of painful osteoporotic vertebral compression fractures. Spine 2001;26(14):1631–1638. 15. Belkoff SM, Mathis JM, Fenton DC, Scribner RM, Reiley ME, Talmadge K. An ex vivo biomechanical evaluation of an inflatable bone tamp in the treatment of compression fracture. Spine 2001;26(2):151–156. 16. Belkoff SM, Mathis JM, Deramond H, Jasper LE. An ex vivo biomechanical evaluation of a hydroxyapatite cement for use with kyphoplasty. Am J Neuroradiol 2001;22:1212–1216. 17. Kim MJ, Lindsey DP, Hannibal M, Alamin TF. Vertebroplasty versus Kyphoplasty: biomechanical behavior under repetitive loading conditions. Spine 2006;31:2079–2084. 18. Tomita S, Malloy S, Abe M, Belkoff SM. Ex vivo measurement of intramedullary pressure during vertebroplasty. Spine 2004;29:723–725. 19. Hiwatashi A, Moritani T, Numaguchi Y. Increase in vertebral body height after vertebroplasty. AJNR Am J Neuroradiol 2003;24:185–189. 20. McKeirnan F, Faciszewski T, Jensen R. Does vertebral height restoration achieved at vertebroplasty matter. J Vasc Interv Radiol 2005;16:973–979. 21. Pradhan BB, Bae HW, Kroph MA, Patel VV, Delamarter RB. Kyphoplasty reduction of osteoporotic vertebral compression fractures: correction of local kyphosis versus overall sagittal alignment. Spine 2006;31:435–441. 22. Garfin SR, Yuan HA, Reiley MA. New technologies in spine: kyphoplasty and vertebroplasty for the treatment of painful osteoporotic compression fractures. Spine 2001;26(14):1511–1515. 23. Nussbaum MS, Gailloud P, Murphy K. A review of complications associated with vertebroplasty and Kyphoplasty as reported to the FDA medical device related web site. J Vasc Interv Radiol 2004;15:1185–1192. 24. Leidig-Bruckner G, Minne HW, Schlaich C, Wagner G, Scheidt-Nave C, Bruckner T, Gebest HJ, Ziegler R. Clinical grading of spinal osteoporosis: quality of life components and spinal deformity in women with chronic low back pain and women with vertebral osteoporosis. J Bone Miner Res 1997;12(4):663–675. 25. Lyles KW, Gold DT, Shipp KM, Pieper CF, Martinez S, Mulhausen PL. Association of osteoporotic vertebral compression fractures with impaired functional status. Am J Med 1993;94(6):595–601. 26. Nelson DA, Kleerekoper M, Peterson EL. Reversal of vertebral deformities in osteoporosis: measurement error or “rebound”? J Bone Miner Res 1994;9(7):977–982. 27. Wilcox RK. The biomechanics of vertebroplasty: a review. Proc Inst Mech Eng H 2004;218(1):1–10. 28. Tomita S, Malloy S, Abe M, Belkoff SM. Ex vivo measurement of intramedullary pressure during vertebroplasty. Spine 2004;29:723–725.
353
16 Sacroplasty Charles H. Cho, John M. Mathis, and Keith E. Kortman
Case Scenario Typical Patient Presentation An elderly, postmenopausal female with multiple underlying medical conditions complains of nonradiating, aching pain in the lower back and sacral area after a mild fall. Preliminary radiographs of the lumber spine show no lumbar compression fractures. The patient is managed conservatively with analgesics and bed rest. One month later, the patient continues to have nonradiating pain, and an MRI of the sacrum shows vertical nondisplaced sacral fractures. Mobility is significantly limited, and there is a component of altered mental status with higher doses of pain medications. The patient has severe osteoporosis and is not considered a candidate for open surgery. Diagnosis Elderly female with sacral insufficiency fractures. Conventional Treatment Historically, patients with sacral insufficiency fractures have been treated with a regimen of bed rest, local warmth, and narcotic analgesics.1 More recently, mobilization and physical therapy have been utilized after a short period of initial bed rest.2 When appropriate, pharmacologic therapy of underlying osteoporosis is promptly initiated. Response to therapy is usually slow with prolonged pain and reduced mobility that can last for months (Figure 16.1). Surgical fracture fixation is rarely used largely due to reduced bone strength and increased risk in these elderly patients with multiple comorbidities.
From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_16, © Springer Science + Business Media, LLC 2010
355
Sacral Fractures and Sacroplasty
Sacroplasty Option Percutaneous Sacroplasty (PS) is the image-guided percutaneous injection of bone cement into the sacral fracture site. Similar to Percutaneous Vertebroplasty (PV), which has been proven effective in alleviating the pain associated with lumbar and thoracic vertebral body compression fractures,3–6 the proposed mechanism of pain relief in PS is fracture stabilization that eliminates motion of the fractured bone. Although the concept of vertebroplasty theoretically applies to sacral fractures as well, even experienced practitioners have been slow to apply vertebroplasty techniques at the sacral level. This is in large part due to constraints imposed by the relatively complex sacral anatomy. The inherent difficulty in fluoroscopic visualization of important sacral landmarks, including the spinal canal and neural foramina, makes the detection of cement leaks into these spaces difficult. Percutaneous Sacroplasty was first described in France in the mid1990s for osteoporosis and in 2000 as a treatment for patients with symptomatic metastatic disease.7,8 Subsequently, a single case9 and a small series10 of patients who underwent PS were reported. In three of the four patients included in these reports, PS was performed with fluoroscopic guidance. The remaining patient was treated with CT guidance. All of the patients reported significantly decreased pain following the procedure. There were no reported complications. PS is ideally suited for Computed Tomography (CT) guidance, with or without fluoroscopy. Localizing images display sacral anatomy and fracture lines. This allows safe and effective placement of two, three, or more needles. After optimal needle tip position is confirmed by additional images, cement can be injected in small aliquots with interval CT imaging to monitor for extraosseous leakage. It is the opinion of the authors that CT is the method of choice for performing PS safely, and, therefore, only its description will appear in this chapter.
Sacral Fractures and Sacroplasty Incidence Spontaneous, pathologic fractures of the sacrum are now well-known but were only recently described by Lourie in 1982.11 Such injuries are often termed insufficiency fractures, indicating that bone strength is
Figure 16.1. (A) Coronal T1 MRI demonstrates a unilateral sacral fracture (white arrow). (B) Coronal TI MRI (several months later than in A) shows a new fracture line developing in the left sacral ala (white arrow). The original fracture in indicated by the white arrowhead. (C) CT scan at the time of treatment demonstrates the same patient as seen in A and B. This image is 6 months later, and the sacral fracture has progressed to bilateral alar components. Also note, the patient has an infusion pump (white arrow) that was put in for pain control (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM, Deramond H, Belkoff SM (eds). Percutaneous Vertebroplasty and Kyphoplasty, Second Edition. New York: Springer Science + Business Media, 2006).
357
358
Chapter 16 Sacroplasty
insufficient to withstand normal mechanical and physiologic forces.12 Sacral fractures may occur spontaneously in patients with osteoporosis, disorders of calcium metabolism, osseous metastatic disease, and prior radiation therapy. There is a strong (10:1) female predominance. Sacral fractures are a frequent cause of debilitating pain, especially in elderly women. The incidence of sacral insufficiency fractures is substantially less than that of osteoporotic fractures involving the lumbar and thoracic spine. However, in a published series evaluating elderly patients with acute onset of low back pain and negative x-rays who underwent bone scintigraphy, 102 sacral fractures were identified over a 2-year period at a single institution.13 In a second study,14 using retrospective review of 1,017 consecutive bone scans in patients over the age of 70, sacral insufficiency fractures were identified in 194 patients (19%). In personal experience of one of the authors (KK) in treating approximately 2000 patients with osteoporotic vertebral fractures at a single institution over a 10-year period, 80 patients with sacral insufficiency fractures were identified and treated, indicating a relative incidence of at least 4% and perhaps as high as 8–10% considering the number of patients whose fractures were not documented and those whose fractures were identified but not treated. Morbidity and Mortality There are relatively few reported series of patients with sacral insufficiency fractures, the largest of which comprised 60 patients who required hospitalization for pain control.15 In this series, the average hospital stay of patients was 45 days. A long-term decrease in selfsufficiency was reported in 50% of patients. Twenty five percent of patients were institutionalized, and there was a 14% 1-year mortality. Better outcomes have been reported in a number of smaller series.16–18 Diagnosis Sacral fractures are often misdiagnosed or go undiagnosed. This may be due to nonspecific symptomatology.19 Patients complain of severe low back and/or buttock pain, often acute in onset. A fall or direct trauma may be hard to elicit as part of the history. The pain is typically exacerbated by weight bearing. Referred pain to the hip or groin is common, especially in patients with concurrent fractures of the ischial and pubic rami. There can be coexistent vertebral compression fractures (Figure 16.2). Physical examination reveals nonspecific sacral tenderness and restricted mobility, indistinguishable from sacroiliac arthropathy and any number of other spinal and pelvic pathologic conditions. Conventional spine radiographs are insensitive in the detection of these fractures. CT offers greater sensitivity (Figure 16.3)20; however, acute, nondisplaced fractures without reactive sclerosis may still be relatively inconspicuous.21 Sacral fractures are typically well shown by bone scintigraphy (Figure 16.2).22 An H-shaped pattern of increased uptake is typical and pathognomonic23; however, radionuclide uptake along symmetric fracture lines may be mistaken for normal variation. Fracture patterns other than the typical H-shaped fracture also occur and may be more difficult to recognize with radionuclide scans.
Figure 16.2. Bone scan showing the typical H shaped sacral insufficiency fracture (black arrowhead) with abnormal activity in both sacral ala and the central body of the sacrum. A vertebral compression fracture co-exists at L1 (black arrow) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM, Deramond H, Belkoff SM (eds). Percutaneous Vertebroplasty and Kyphoplasty, Second Edition. New York: Springer Science + Business Media, 2006).
Figure 16.3. An axial CT scan of the sacrum shows bilateral fractures through the sacral ala (white arrows). The fracture lines are easily seen on CT (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM, Deramond H, Belkoff SM (eds). Percutaneous Vertebroplasty and Kyphoplasty, Second Edition. New York: Springer Science + Business Media, 2006).
360
Chapter 16 Sacroplasty
MRI is both sensitive and specific in the demonstration of sacral fractures.24 Fractures are best shown in either sagittal or angled coronal planes. Marrow edema is conspicuously demonstrated on routine T1-weighted sequences (Figure 16.4a), with slightly greater sensitivity
Figure 16.4. (A) Axial T1 weighted MRI demonstrates the low signal fracture through the right sacral wing (white arrows). The patient presented with a fracture on only one side. (B) Coronal inversion recovery MRI shows high signal in both sacral ala (black arrows) and the central body. In this case, a typical “H” shaped fracture is present (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM, Deramond H, Belkoff SM (eds). Percutaneous Vertebroplasty and Kyphoplasty, Second Edition. New York: Springer Science + Business Media, 2006).
Sacral Fractures and Sacroplasty
afforded by fat-suppressed T2-weighted or STIR sequences (Figure 16.4b). Sacral fractures are often detected as the result of a “corner call” at the bottom of a lumbar spine exam, so lumbar MRIs should include a review of included portions of the sacrum looking for signal abnormality. When suspicion of a sacral fracture is sufficiently high, dedicated images centered on the sacrum with a small field-of-view are indicated. Fracture Anatomy A spectrum of fracture types can occur in the sacrum with the H-shaped fracture being the paradigm. The “H” fracture involves both sacral ala and has a horizontal component that connects the two vertical or alar fractures (Figure 16.5a). The horizontal component can be absent, and the fracture can present with unilateral or bilateral alar components only (Figure 16.5). When a unilateral component is present, it may progress over time to a bilateral or H-shaped fracture (Figure 16.5a). Finally, the horizontal component can present with a single unilateral alar fracture (Figure 16.5c). This complexity of fracture configurations compounds the difficulty of diagnosis. Patient Selection As with PV elsewhere in the spine, appropriate patient selection is critical to ensure favorable outcomes. PS should only be offered to patients with severe pain poorly responsive to conventional medical therapy. Fractures should be documented by scintigraphy, CT, and/or MRI. We prefer MRI for its combination of sensitivity, specificity, and anatomic detail. Healing fractures may remain “hot” on bone scan and appear sclerotic on CT. If marrow edema is limited or absent on MRI, the authors are hesitant to perform a PS. An alternative approach might be a sacroiliac joint injection or nerve block as indicated by the location and distribution of pain. An example of sacral injury not treated is shown in Figure 16.6. This case shows a patient with an acute lumbar compression fracture. The pain, however, is mostly at the sacral region. CT of the sacrum is not convincing. Bone scan shows mild increased uptake in the sacral area. MRI of the sacrum reveals edema in the sacrum bilaterally. Although the symptoms and the MRI findings of microfractures warrant sacroplasty, the fracture lines are not distinctively visible on CT. Sacroplasty can still be performed based on MRI findings and using bone and foramina landmarks. However, in this case, without a visible target, sacroplasty was not performed. Medical management with CT-guided anesthetic and steroid injection was pursued. Fortunately, the patient made an adequate recovery. In addition to appropriate imaging, preoperative evaluation includes a focused history and physical examination, as well as routine laboratory studies. Patients with an active, untreated infection or coagulopathy should be deferred or excluded. Informed written consent should be obtained from eligible patients and should include discussion of nerve injury from extraosseous cement extravasation.
361
362
Chapter 16 Sacroplasty
Figure 16.5. (A) Artist sketch of sacrum showing the typical “H” shaped fracture with bilateral vertical and connecting horizontal components (black arrows). (B) Unilateral sacral fracture (black arrows). (C) Combined unilateral vertical and horizontal fractures. (D) Bilateral vertical sacral fractures without the horizontal component (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM, Deramond H, Belkoff SM (eds). Percutaneous Vertebroplasty and Kyphoplasty, Second Edition. New York: Springer Science + Business Media, 2006).
Technique CT Guidance Patients are positioned prone on the CT table with care taken to avoid additional injury to these fragile patients during transfer. Protective padding may be applied to the shoulders, elbows, hips, knees, and ankles. Procedural sedation can be achieved with small divided doses of IV fentanyl and midazolam, titrated to the patient’s needs. The patient’s vital signs and oxygen saturation levels should be monitored throughout the procedure.
Sacral Fractures and Sacroplasty
Figure 16.6. (A) Sagittal T2 images show compression of the lumbar vertebral body with edema, amendable to vertebroplasty. This patient had sacral pain. (B) CT showed no visible fracture lines. (C) Bone scan showed increased activity in the sacral areas. (D) MRI showed edema (T2 prolongation) in the sacrum on both sides. There is edema in the marrow with discrete fracture lines not visible.
IV antibiotics are recommended and should be given at the onset of the procedure. One gram of cephazolin supplies adequate coverage for most skin pathogens and can be used in all patients except those with a specific allergy to cephalosporins and/or a severe allergy to penicillin. Adding antibiotics to the cement is not required or recommended. The patient’s preprocedure imaging studies should be reviewed and accessible within the CT suite. Localizing, noncontrast CT images are obtained at 3 or 5 mm intervals. Puncture sites are then chosen and, if desired, can be marked on the skin with an indelible marker. The low back and buttocks are then scrubbed with an antiseptic solution, and a surgical drape is applied. The physician and any personnel with direct access to the surgical site or equipment should observe conventional hand washing techniques and wear a surgical gown, cap, mask, and sterile gloves. The choice of puncture sites depends on the fracture pattern and location of fracture lines (Figure 16.7). For H-shaped fractures, one
363
364
Chapter 16 Sacroplasty
Figure 16.7. Sketch of axial section through the sacrum and pelvis. This shows the various needle trajectories that can be used to access the various components of the sacral insufficiency fractures. The central component can be reached by a trans-SI joint approach or a posterolateral approach between the foramina and the spinal canal. The alar fracture can be easily treated by a posterolateral approach which usually parallels the SI joint and is lateral to the foramina (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM, Deramond H, Belkoff SM (eds). Percutaneous Vertebroplasty and Kyphoplasty, Second Edition. New York: Springer Science + Business Media, 2006).
may choose to attempt fixation along the horizontal component of the fracture, but injection of cement into this component of the fracture seems optional for pain relief. This component of the fracture is usually located at the S2 level. The horizontal component of the fracture can be accessed via a needle puncture posterolaterally through the SI joint (Figure 16.8a). An alternate approach places the puncture site over one of the ala with medial angulation of the needle between the spinal canal and the ipsilateral sacral foramen (Figure 16.8b). This needle is advanced into the sacral body to be treated. Puncture sites to treat the lateral fracture components are chosen over each sacral ala at a somewhat medial location to allow mild lateral angulation of the needle along the fracture line which is usually parallel to the sacroiliac joint (Figure 16.8c). At each puncture site, the skin, underlying soft tissues and periosteum are first infiltrated with local anesthetic. A small dermatotomy is made at the intended puncture site with a scalpel blade. The cement delivery needle is advanced to the posterior sacral cortex. The needle tip can be fixed in place by advancing it 2–3 mm into the cortex. Needle position is then checked by additional CT imaging. The cortical entry site and angulation are adjusted as needed. Initially, the needle is advanced in small (5–10 mm) increments. Each advance is monitored by additional CT imaging. The needles can be advanced with manual pressure or by
Sacral Fractures and Sacroplasty
365
Figure 16.8. (A) Axial CT scan demonstrating the trans-SI approach to the central part of the sacrum (black arrows). Care must be taken to avoid the foramina (black arrowhead). (B) Axial CT scan shows the other approach to the central sacrum. This posterolateral approach places the needle (black arrowhead) between the foramina (white arrow) on the entry side and the spinal canal.
small taps with a sterile mallet. When a safe needle course is assured, the needle can be advanced in 10-mm increments. The alar needles can be advanced to within 10 mm of the anterior sacral cortex. Vertical sacral fractures (through the lateral ala) may be unilateral or bilateral. Puncture sites are chosen over each effected ala to allow fracture fixation along the vertical axis. Each puncture site is chosen
366
Chapter 16 Sacroplasty
Figure 16.8. Continued. (C) Axial CT of needles (white arrows) placed into the lateral fracture components bilaterally. This trajectory is parallel to the SI joint and avoids the foramina (black arrowhead) (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM, Deramond H, Belkoff SM (eds). Percutaneous Vertebroplasty and Kyphoplasty, Second Edition. New York: Springer Science + Business Media, 2006).
to allow mild lateral angulation of the needle along the fracture line parallel to the sacroiliac joint, taking care to avoid the neural foramina (Figure 16.7). After satisfactory needle position is confirmed by additional CT imaging, the cement is mixed. Because of the additional time required for CT monitoring, cement with a relatively long set-up time is preferred. Only bone cements approved for PV should be used for PS, as they will contain adequate opacifiers for easy visualization. Working time of the cement can be increased by storing filled syringes in a container of sterile iced saline. A new non-PMMA cement approved for PV, Cortoss (Orthovita; Malvern, Pa) could be used for sacroplasty as it has mix on demand system that allows mixing of only the amount needed for immediate injection. Other cement, not yet mixed, has no time limit on waiting before use. For more information about this cement, see the section on "cement selection" in the chapter on Percutaneous Vertebroplasty. The cement is injected sequentially in 0.5-mL aliquots. After each aliquot, the distribution of cement is monitored by additional CT imaging (Figure 16.9a). If cement extravasates into the spinal canal or a neural foramen, no additional cement is injected through that needle (Figure 16.9b). During the injection, alar needles may be withdrawn along the fracture line in small increments to increase the cross-sectional area of cement fixation.
Sacral Fractures and Sacroplasty
367
Figure 16.9. (A) CT scan showing early cement injections bilaterally (white arrows). Note the anatomic definition and easy visualization of the neural foramina. (B) CT scan in another patient reveals minimal extravasation of cement into the neural foramina (white arrow) long before the quantity is likely to create symptomatic nerve encroachment.
As with vertebroplasty elsewhere in the spine, the volume of cement injection does not directly correlate with pain relief. Typical total cement volumes range from 3 to 8 mL total. The alar component of the fracture will accommodate a higher volume of cement than the central sacral body (Figure 16.9c). The distribution and amount of cement varies from patient to patient and is seldom symmetrical (Figure 16.9d). As cement hardens within the needle, further injection with a syringe may become difficult. Advancing the trocar through it can usually clear
368
Chapter 16 Sacroplasty
Figure 16.9. Continued. (C) Coronal CT demonstrates cement in the central sacral body (black arrowhead) and lateral sacral wings (black arrows). A larger amount of cement was used laterally to secure the fractures in the sacral ala. (D) Coronal CT demonstrates four injections of cement (black arrows) which resulted in good pain relief in this patient following PS (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM, Deramond H, Belkoff SM (eds). Percutaneous Vertebroplasty and Kyphoplasty, Second Edition. New York: Springer Science + Business Media, 2006).
an obstructed cannula. Care must always be taken to avoid extraosseous or foraminal cement extravasations. Procedure time is considerably decreased if CT fluoroscopy is utilized. In addition, “real time” visualization afforded by CT fluoroscopy decreases the likelihood of cement extravasations and neurologic complication.25 However, CT fluoroscopy is not required for procedure safety.
Sacral Fractures and Sacroplasty
A small sample of cement may be used to assess cement hardening on the working table. When the cement is adequately thickened, the needles are removed, and sterile dressings are placed over the puncture sites. The patient is carefully rolled supine onto a gurney. The patient can typically be ambulated and discharged after a period of bed rest and monitoring (usually 1–2 h). Fluoroscopic Guidance Fluoroscopic guidance (with or without CT) has been used for PS.9,10 Fluoroscopy alone may be faster than CT, but it is limited by from a lack of anatomic resolution. Typically, there is difficulty seeing the neural foramina with assurance. Cement can be visualized, but one does not always know whether cement is leaking into the foramina early enough to stop the injection and avoid complications (Figure 16.10). To identify the foramina better, some operators have introduced small gauge spinal needles into the foramina before cement is injected. This aids the operator to identify the foramina location even when the margins are not well seen with fluoroscopy during the procedure. The use of fluoroscopy to date has not resulted in high incidence of reported complications. However, the authors believe that its use alone is problematic and have found CT to offer a greater margin of safety for identifying where cement is going during injection. Outcomes In published reports of sacroplasty, all claim substantial pain relief in treated patients. Frey and colleagues reported results from a prospective multicenter study of 37 patients who underwent fluoroscopically guided sacroplasty for painful insufficiency fractures.26 The procedure was deemed to be both safe and highly efficacious. Sustainable pain relief was verified a full year after the procedure. Whitlow and colleagues reported pain relief in 12 sacroplasty patients comparable to that experienced by a cohort of 21 vertebroplasty patients.27 In personal experience of one of the authors (KK) with more than 75 patients undergoing sacroplasty, all but one experienced dramatic pain relief, allowing rapid return to ambulatory status. Complications Reported complications of sacroplasty are rare. Extraosseous cement extravasation can occur but is usually not manifested clinically. Cement extravasation into a sacral neural foramen may result in radicular pain. This occurred in one of 75 patients in one author’s (KK) series. One patient also developed progressive fracture dislocation 1 week after sacroplasty. Patients with unilateral alar fractures may develop fractures in the opposite ala or elsewhere in the pelvic ring even after sacroplasty. Complete Clinical Picture Patients with sacral insufficiency fractures may have additional injuries or conditions that contribute to their pain. Patients with unilateral
369
370
Chapter 16 Sacroplasty
F igure 16.10. (A) Fluoroscopic image of an injection needle in sacral wing and early cement injection (black arrow). Note the general lack of anatomic landmarks with this imaging method. Neural foramina are hard to see, at best making early detection of a cement leak almost impossible. (B) The same case viewed with CT demonstrates the anatomy in much better detail and allows accurate localization of cement (black arrow) as it is injected (Reprinted with the kind permission of Springer Science + Business Media from Mathis JM, Deramond H, Belkoff SM (eds). Percutaneous Vertebroplasty and Kyphoplasty, Second Edition. New York: Springer Science + Business Media, 2006).
Conclusion
alar fractures have a high incidence of ipsilateral pubic and ischial ramus fractures.17,28,29 The fractures should be documented and the patient informed regarding the likelihood of ongoing pain that may take weeks to resolve. There are early reports of successful percutaneous cement augmentations of theses fractures by a few operators. Concurrent osteoporotic fractures of the thoracic and lumbar spine are not uncommon17 and should be treated with conventional vertebroplasty techniques if the symptoms correlate with the imaging findings. Sacroiliac pain is also common, especially opposite a unilateral alar fracture. If the patient is tender over the contralateral sacroiliac joint, an intra-articular corticosteroid injection can be easily performed with CT guidance at the time of the PS. As previously discussed, patients with healing fractures of the sacrum may also develop sacroiliac joint pain. Distinguishing the exact source of pain is difficult in these patients. Typically, sacroiliac joint pain is less severe than that associated with a recent fracture and less exacerbated by weight bearing. Sacroiliac joint pain is characterized by morning stiffness and pain exacerbation brought on by prolonged standing or sitting. Certainly, in patients with persistent sacral level pain following vertebroplasty, sacroiliac arthropathy should be strongly considered. Occasionally, sacral insufficiency fractures will result in S1 radiculopathy, although nerve root impingement by a fracture fragment is extremely uncommon in this patient population. In patients with the appropriate dermatomal distribution of pain, a selective S1 nerve root block may be considered. CT guidance is ideally suited for that procedure, and it can be performed in conjunction with PS.
Conclusion Sacral insufficiency fractures represent a small but significant component of osteoporotic fractures of the skeleton. These fractures can result in long-term pain and disability for the elderly patient. Traditional conservative therapy fails in a high percentage of these patients. Percutaneous sacroplasty offers a low risk, minimally invasive interventional option for rapid pain relief. It should be considered as a good option in all patients with sustained pain resulting from insufficiency fracture of the sacral region. References 1. Grasland A, Pouchot J, Mathieu A, Paycha F, Vinceneux P. Sacral insufficiency fractures: an easily overlooked cause of back pain in elderly women. Arch Intern Med 1996;156:668–674. 2. Babayev M, Lachmann E, Nagler W. The controversy surrounding sacral insufficiency fractures: to ambulate or not to ambulate? Am J Phys Med Rehabil 2000;79:404–409. 3. Jensen ME, Evans AJ, Mathis JM, Kallmes DF, Cloft HJ, Dion JE. Percutaneous polymethylmethacrylate vertebroplasty in the treatment of osteoporotic vertebral body compression fractures: technical aspects. AJNR Am J Neuroradiol 1997;18:1897–1904.
371
372
Chapter 16 Sacroplasty 4. Deramond H, Depriester C, Galibert P, Le Gars D. Percutaneous vertebroplasty with polymethylmethacrylate: technique, indications, and results. Radiol Clin North Am 1998;36(3):533–546. 5. Cotton A, Boutry N, Cortet B, Assaker R, Demondion X, Leblond D, Chastanet P, Duquesnoy B, Deramond H. Percutaneous vertebroplasty: state of the art. Radiographics 1998;18:311–320. 6. Mathis J, Barr J, Belkoff S, Barr M, Jensen M, Deramond H. Percutaneous vertebroplasty: a developing standard of care for vertebral compression fractures. AJNR Am J Neuroradiol 2001;22:373–381. 7. Dehdashti AR, Martin JB, Jean B, Rüfenacht DA. PMMA cementoplasty in symptomatic metastatic lesions of the S1 vertebral body. Cardiovasc Intervent Radiol 2000;23:235–237. 8. Marcy PY, Palussiere J, Descamps B, Bondiau P-Y, Ciais C, Bruneton J-N. Percutaneous cementoplasty for pelvic bone metastasis. Support Care Cancer 2000;8:500–503. 9. Garant M. Sacroplasty: a new treatment for sacral insufficiency fracture. J Vas Interv Radiol 2002;13:1265–1267. 10. Pommersheim W, Huang-Hellinger F, Baker M, Morris P. Sacroplasty: a treatment for sacral insufficiency fractures. AJNR Am J Neuroradiol 2003;24:1003–1007. 11. Lourie H. Spontaneous osteoporotic fracture of the sacrum. an unrecognized syndrome of the elderly. JAMA 1982;248:715–717. 12. Pentecost RL, Murray RA, Brindley HH. Fatigue, insufficiency, and pathologic fractures. JAMA 1964;187:1001–1004. 13. Hatzl-Griesenhofer M, Pichler R, Huber H, Maschek W. [The insufficiency fracture of the sacrum. An often unrecognized cause of low back pain: results of bone scanning in a major hospital]. Nuklearmedizin 2001;40(6):221–227. 14. Wat SY, Seshadri N, Markose G, Balan K. Clinical and scintigraphic evaluation of insufficiency fractures in the elderly. Nucl Med Commun 2007;28(3):179–185. 15. Taillandier J, Langue F, Alemanni M, Taillandier-Heriche E. Mortality and functional outcomes of pelvic insufficiency fracture in older patients. Joint Bone Spine 2003;70(4):287–289. 16. Rawlings, CE, Wilkins RH, Martinez S, Wilkinson SH. Osteoporotic sacral fractures: a clinical study. Neurosurgery 1988;22:72–76. 17. Weber M, Hasler P, Gerber H. Insufficiency fractures of the sacrum: twenty cases and review of the literature. Spine 1993;18(16):2507–2512. 18. Gotis-Graham I, McGuigan L, Diamond T, Portek I, Quinn R, Sturgess A, Tulloch R. Sacral insufficiency fractures in the elderly. J Bone Joint Surg 1994;76-B(6):882–886. 19. Renner JB. Pelvic insufficiency fractures. Arthritis Rheum 1990;33:426–430. 20. Garetta DJ, Yandow DR. Computed tomography of spontaneous osteoporotic sacral fractures. J Comput Assist Tomogr 1984;8:1190–1191. 21. Grangier C, Garcia J, Howarth NR, May M, Rossier P. Role of MRI in the diagnosis of insufficiency fractures of the sacrum and acetabular roof. Skeletal Radiol 1997;26(9):517–524. 22. Schneider R, Yacavone J, Ghelman B. Unsuspected sacral fractures: detection by radionuclide bone scanning. AJR Am J Roentgenol 1985;144(2):337–341. 23. Ries T. Detection of osteoporotic sacral fractures with radionuclides. Radiology 1983;146(3):783–785. 24. Brahme SK, Cervilla V, Vint V, Cooper K, Kortman K, Resnick D. Magnetic resonance appearance of sacral insufficiency fractures. Skeletal Radiol 1990;19:489–493.
Conclusion 25. Pitton MB, Drees P, Schneider J, Brecher B, Herber S, Mohr W, Eckardt A, Heine J, Thelen M. Evaluation of percutaneous vertebroplasty in osteoporotic vertebral fractures using a combination of CT fluoroscopy and conventional lateral fluoroscopy. Rofo 2004;176(7):1005–1012. 26. Frey ME, DePalma MJ, Cifu DX, Bhagia SM, Daitch JS. Efficacy and safety of percutaneous sacroplasty for painful osteoporotic sacral insufficiency fractures: a prospective, multicenter trial. Spine 2007;32(15):1635–1640. 27. Whitlow CT, Mussat-Whitlow BJ, Mattern CW, Baker MD, Morris PP. Sacroplasty versus vertebroplasty: comparable clinical outcomes for the treatment of fracture-related pain. AJNR Am J Neuroradiol 2007;28(7): 1266–1270. 28. Davies AM, Evans NS, Struthers GR. Parasymphaseal and associated insufficiency fractures of the pelvis and sacrum. Br J Radiol 1988;61(722): 103–108. 29. Peh WC, Khong PL, Ho WY, Yeung HW, Luk KD. Sacral insufficiency fractures. Spectrum of radiological features. Clin Imaging 1995;19(2):92–101.
373
17 Spinal Cord Stimulation: Uses and Applications Stanley Golovac
Introduction Therapeutic electrical stimulation of the nervous system has developed enormously over the last 40 or so years, from the work that was performed by Shealy et al.1 to the point where tens of thousands of units have been implanted every year, and yet one of the biggest criticisms is that there is a lack of high quality evidence of its efficacy. Since the relatively recent enlightenment regarding the neurophysio logy of pain, it has been discovered that electrical stimulation of almost any part of the nervous system can have a dramatic useful purpose of modulating painful conditions such as spinal conditions (Failed Back Spinal Syndrome), CRPS I and CRPS II (Complex Regional Pain Syndrome), IC (Interstitial Cystitis), gastroparesis, chronic pancreatitis, peripheral neuropathic pain, and transformed migraine headaches. It seems that therapeutic stimulation has the potential to continue to add to the knowledge of neurological function; for example, what does relief of central pain afforded by motor cortex stimulation, tell about the integration of “motor” and sensory in the brain? It appears that, at least to some extent, spinal cord stimulation influences intrinsic, already available, modulatory systems, whose function may or may not have been disturbed, to bring about “normalization.” This principle may apply to both “neuropathic” pain and maladaptive changes occurring in ischemic syndromes such as myocardial ischemia, peripheral vascular ischemia, and neuropathic ischemic states.
The History of Neurostimulation Electrical stimulation for the treatment of pain has been around, in one form or another, in every culture for thousands of years. It is said that, from circa 9,000 BC, bracelets were used to prevent headaches and arthritis.2 Each culture acquired and/or discovered a type of fish that could, through electrical stimulation, stun or affect a person in one From: Image-Guided Spine Interventions, Edited by: J.M. Mathis and S. Golovac, DOI 10.1007/978-1-4419-0352-5_17, © Springer Science + Business Media, LLC 2010
375
376 Chapter 17 Spinal Cord Stimulation: Uses and Applications
Figure 17.1. Artist’s impression of the treatment of gout (A) and headache (B) using torpedo fish (Adapted with permission from Perdikis P. Transcutaneous nerve stimulation in the treatment of protracted ileus. S Afr J Surg 1977;17(2):81–86).
way or another. In Africa in the Nile valley, electric catfish use electrical discharge to stun their prey. The ancient Egyptians acknowledged the power of the Nile catfish in tomb paintings. The ancient Greeks called the ray “Narke” or “numbness-producing,” from which the word “narcosis” was coined. The Romans called the ray “torpedo” from the word “torpor,” as the name was synonymous with the effect. Conditions, such as gout and headaches, were treated by having the torpedo ray fish discharge its electrical charge close to the site of the painful condition (Figure 17.1). The therapeutic application of electricity was named “Franklinism” after the American statesman and scientist Benjamin Franklin, who, with his famous kite experiment in 1775, proved that lightening and electrostatic charge on a Leyden jar were identical. John Wesley, the founder of Methodism, extolled the virtues of electricity in his book, The Desideratum, and advocated electrotherapy for angina pectoris, gout, headaches, pleuritic pain, and sciatica.3
Gate Control Theory and Implantable Stimulators The theory by Melzack and Wall in 1965 postulated central inhibition of pain by nonpainful stimuli, a concept that had been predicted half a century ago by the English neurologist Sir Henry Head.4 In 1965, Patrick Wall recruited William Sweet, Head of Neurosurgery at
Gate Control Theory and Implantable Stimulators
Harvard Medical School, to clinically test the gate theory. Quite simply, Melzack and Wall proposed in their theory that nerve impulses from afferent fibers lead to spinal cord transmission (“T”) cells in the substantia gelatinosa. The firing of the projection neuron determines pain. The inhibitory interneuron decreases the chance that the projection neuron will fire. Firing of C-fibers inhibits the inhibitory interneuron (indirectly), thereby increasing the chances that the projection neuron will fire. Firing of the large myelinated A-beta fibers activates the inhibitory interneuron, reducing the chances that the projection neuron will fire, even in the presence of a firing nociceptive fiber. The marriage of electricity to pain control within this paradigm of the gate control theory lies in the age-old principle of “counterirritation.” At first, they experimented on their own infraorbital nerves using needle-stimulating electrodes and on superficial nerves, such as the ulnar nerve, using superficial electrodes. They then used transcutaneous or percutaneous stimulation in three patients who experienced partial or complete relief of pain during stimulation.4,5 Shealy, a neurosurgeon in La Crosse, Wisconsin, and colleagues thought that the “gate” could be best closed by stimulating the dorsal columns and confirmed this assumption experimentally in cats.6 In 1967, Shealy and his coworkers implanted a device (Figure 17.2) in 50-year-old woman suffering from intractable carcinomatous pelvic pain and used a radio-frequency stimulator.6
Figure 17.2. Bottom view of assembled radio-frequency receiver and electrodes as implanted in Shealy’s second patient (RW) on October 8, 1967. The coiled platinum–iridium wires connecting the electrodes to the receiver and the epoxy and glass portions of the receiver were covered with medical-grade silastic. The system was activated by a variable frequency transmitter–stimulator with a fixed pulse width. Frequency and amplitude were controlled by the patient (Reprinted with permission from Mortimer JT. Pain suppression in man by dorsal column electroanalgesia. PhD Thesis. Case Western Reserve University, Cleveland, OH, USA).
377
378 Chapter 17 Spinal Cord Stimulation: Uses and Applications
The circuit design was based on a modified Medtronic (Medtronic, Inc., Minneapolis, MN) device for stimulation of the carotid sinus to control angina and hypertension. The patient experienced approximately 50% relief of her pain, at times almost total control of her pain, and was extensively evaluated until Mortimer (from Shealy’s team) successfully defended his PhD thesis: “Pain suppression in man by dorsal column electro-analgesia” in May 1968.1
Dorsal Roots The majority of dorsal root fibers, upon entering the spinal cord, proceed toward the dorsal columns, where they bifurcate into ascending and descending branches.7,8 In comparison with longitudinal dorsal column fibers, dorsal root fibers have a curved shape, and they differ in orientation with respect to the spinal cord and the implanted electrodes. Dorsal root fibers average 15 mm in diameter. Proximal to the dorsal ganglion, the dorsal root fibers fan out in an ascending, dorsomedial direction to form the rootlets that enter the spinal cord at different angles. Strujik and colleagues have studied the effect of the curvature of the dorsal roots on their electrical threshold.9–11
Chronic Back Pain Back pain is responsible for over 80% of all back pain syndromes affecting Americans daily. One in every 14 people are affected by some kind of either cervical, thoracic, or lumbar pain, meaning that out-of-work days can affect work status for employers. Estimated annual costs for direct and indirect treatments have been estimated at 20–60 billion dollars annually.12 Most back pain is acute or sub-acute resolving within a 6-week period of time normally. However, other estimates suggest less than 30% of patients are completely improved within 3 months of treatment.13 Chronic low back pain represents one of the most widespread and costly medical problems today; it is also a major cause of absenteeism from the workplace. Past analyses have demonstrated that over 5 million people in the United States are afflicted with chronic low back pain. Conservative estimates place the annual cost of treatment at 25 billion dollars.14 A significant fraction of these dollars is attributable to the more than 200,000 US patients yearly who elect for lumbosacral surgery to relieve their pain. Unfortunately, 20–40% of surgical patients will experience persistent or recurrent pain.15 Failed Back Spinal Syndrome One important subset of patients includes those with the so-called “Failed Back Surgery Syndrome”(FBSS). In the literature, this multidimensional syndrome has been used to describe various types of pain including centrally located lumbosacral pain, buttock pain, gluteal pain, extremity pain, and diffuse lower back pain. Many published series emphasize the distinction between back and leg pain; however,
Complex Regional Pain Syndrome
details of the pain syndromes are usually lacking. The etiology of these pains is very difficult to pinpoint. Some of the reasons are: wrong level of surgery, psychological overlay, arachnoiditis, lumbosacral epidural fibrosis, vertebral microinstability, and recurrent disc herniations. While there are several paths as to why one develops a condition such as FBSS, the most common, unfortunately, is poor selection for the surgery. This means that the patient may have had a psychological profile or physical pathology that was contraindicated or not appropriate for the surgical intervention.16,17 Furthermore, if the patient is misdiagnosed, the surgery is obviously incorrect and damaging. The most common misdiagnosis in these cases is arthritis misdiagnosed as lumbar disc disease. Often, improper selection and misdiagnosis follows from inadequate preoperative evaluation and diagnosis work-up. A full diagnostic work-up should include a medical and psychological evaluation. The medical evaluation should include a comprehensive physical examination and history, imaging, and relevant diagnostic procedures, such as radiography, computed tomography (CT), Magnetic Resonance Imaging (MRI), myelography, bone scanning, electromyography, discography, and various diagnostic injections, in order to help delineate the pain generator.18 Surgery that is unnecessary may also be the cause of FBSS. Unnecessary surgery not only fails to treat the problem appropriately, but also may worsen the patient’s condition. An unnecessary surgical excision of the nucleus pulposus from the normal disc is likely to increase the risk of chronic back pain by creating instability and malalignment. Needless surgery places patients at unnecessary risk for injured nerves, torn dura or arachnoid, CSF leakage, and possible wound infection or hemorrhage later.19
Complex Regional Pain Syndrome Complex Regional Pain Syndrome (CRPS), formerly recognized as RSD (Reflex Sympathetic Dystrophy), is frequently misunderstood, misdiagnosed, and mistreated. First and foremost, re-establishing the function of the injured area is of utmost importance. Only one in five are able to return to work after having been diagnosed with the disease. It has been estimated to occur in approximately 1:2,000 traumatic events.20 In 1994, the International Association for the Study of Pain proposed stringent diagnostic criteria and named the “Complex Regional Pain Syndrome Type 1.”21 Pain alleviation is secondary when it comes to the function of the injured area. As one of the author’s colleagues says frequently: “This is not a disease of the extremity; it is a disease of the nervous system.” Initial work performed by the late John Bonica provides an explanation of the disease development stages.22 Stage 1, or the acute phase, is when the extremity is very sensitive to any form of touch, contact, or stimulus. The extremity may appear swollen, discolored, and stiff. Stage 2, or the dystrophic phase, is seen some 3–4 months after the initial injury. During this stage, the extremity begins to become contracted, and remains swollen and very painful. The area now begins to
379
380 Chapter 17 Spinal Cord Stimulation: Uses and Applications
feel cooler with respect to the other extremity. Unfortunately, during Stage 3 or the atrophic phase, the extremity becomes almost useless. If severe atrophy can develop, uncal changes occur. Brittle nails form, and either excessive hair growth or sparse hair can occur. Physical therapy is mandatory in order to even hope for any improvement of the injured area. Functionality is key. The injured area should respond to a combination of treatments. Medications should be started to improve sleep and inhibit neuropathic impulses from the injury; sympathetic inhibition is important in order to restore blood flow to the injury. Conventional pain medication, physical therapy, sympathetic blocks, and transcutaneous electrostimulation of the nerves have all been used to help alleviate pain caused by the initial injury. Spinal cord stimulation introduced by Shealy and colleagues in 1967 has been one of the most successful modalities used in alleviating pain, swelling, and stiffness today.6
Figure 17.3. Multichannel device (Image provided courtesy of St. Jude Medical, St. Paul, MN. Image Copyright St. Jude Medical, all rights reserved).
Arachnoiditis
Of extreme importance in the clinical application of spinal cord stimulation to complex pain syndromes were that multiple arrays of electrodes with defined spacing would allow “capture” of pain (i.e., the overlap of areas of paresthesia over the region of pain perception) better than a single array of electrodes.1,20,21,23–29 Law and Kirkpatrick30 showed that a defined area of the spinal cord, the “physiologic midline,” which differed from the “anatomical midline,” was crucial in the modulation of pain transmission. They also determined that medial dorsal column penetration was improved with bilaterally placed electrode arrays around the physiologic midline, using staggered guarded cathodes with intercontact spacing of 4 mm and surface contact of 3 mm. (A “guarded cathode” is a selection of three adjacent electrodes with the midline electrode, the cathode (“negative”), having opposite polarity from the surrounding two anodal (“positive”) electrodes.) This would enhance the capture of pain in complex dynamic pain syndrome with axial back pain and complex regional pain syndrome. The concept of dual arrays of electrodes led the way for multichannel devices that allow the “steering” of paresthesia (i.e., the moving of the active cathode in both the longitudinal and transverse directions electronically) without the need for surgical intervention. A multichannel device (Figure 17.3) is one that allows for more than one lead to be active simultaneously, thus increasing the area of the spinal cord that can be covered. As soon as the pain improves, swelling, trophic changes, and function should begin to improve. The pseudomotor changes that were initially seen will then begin returning to preinjury states. Care should always be stressed with relation to the use of the extremity, but one should never treat the injured extremity differently with respect to its daily use. One should force the extremity to function normally.
Postherpetic Neuralgia PostHerpetic Neuralgia (PHN) is a very unfortunate disease caused by the well-known virus that most individuals were exposed to as children (Varicella zoster). Approximately 3% of all individuals who are affected by the disease state of shingles develop PHN. Because of this pain syndrome, the use of spinal cord stimulation has been applied effectively to the areas of painful dysesthesia (Figure 17.4). By using a very high frequency to the affected nerve rootlets, pain that is intolerable becomes manageable.
Arachnoiditis Arachnoiditis is a complication that is postsurgical and is believed to be caused by scarring and adhesions located intraspinally. Because of the phenomenon that develops, pain from the adhesions and clumping cause pain with movement of the spine, dysesthetic burning pain in extremities, and loss of functionality. Once again, this unfortunate painful syndrome has very few solutions to its treatment. Intrathecal pain-alleviating medications can be used, but, as has been frequently witnessed, this modality leads only to dependency, tolerance, and habituation.
381
382 Chapter 17 Spinal Cord Stimulation: Uses and Applications Figure 17.4. Stimulation lead configuration.
Phantom Limb Pain Pain that ensues from having either a traumatic amputation or a medically necessary amputation can lead to pain in the missing limb. Pain that is perceived in the limb typically feels dysesthetic and burning in nature. Many of these types of painful syndromes can be improved with the use of various medications, such as tricyclic anti-depressants and neuropathic pain medications. Spinal cord stimulation has been shown to be effective for inhibiting the sensations believed to be painful.
Intercostal Neuralgia Intercostal neuralgia can develop from trauma to the affected nerve rootlets involved in the thoracic level, either via a planned surgical event that may have destroyed or injured a nerve, or with tumor invasion, viral irritation, or stretch injury from a thoracotomy. Either way, the persistent symptoms are the same: burning, pinprick sensation, dysesthesia, hyperalgesia, and allodynia at times. Medication management does help. If it does not, the alternative is to use high-frequency stimulation to overcome the painful dysesthesia. By utilizing a stimulus that blunts the afferent stimulation, one then
Angina Pectoris and Peripheral Vascular Disease
feels the paresthesia over the affected area in order to allow the wearing of clothes and simple touching to occur without pain. All in all, utilizing a stimulus is far better than opioid therapy. Antiviral medications are initially started to avoid any further worsening of the disease and possible spread of the virus.
Occipital Neuralgia (Peripheral Nerve Stimulator) Occipital neuralgia is characterized by paroxysms of pain occurring within the distribution of the greater occipital nerves. The pain may radiate to the ipsilateral frontal or retro-orbital regions of the head. Extreme localized tenderness is often encountered upon palpation over the occipital notches with reproduction of focal and radiating pain. Though known causes include closed head trauma, direct occipital nerve injury, neuroma formation, or upper cervical nerve root compression (spondylosis or ligamentous hypertrophy), most patients have no demonstrable lesions. Treatment options for intractable occipital nerve pain refractory to medications usually involve chemical, thermal, or surgical ablation, following diagnostic local anesthetic blockade. Surgical approaches include neurolysis or nerve sectioning.1
Angina Pectoris and Peripheral Vascular Disease In 1999, the American Heart Association defined angina pectoris as a clinical syndrome characterized by discomfort in the chest, jaw, shoulder, back, or arm, typically aggravated by stress.31 The syndrome is caused by an imbalance between demand and the supply of oxygen to the heart. The decrease in blood supply to the heart is usually the result of vessel occlusion or vasospasm. An angina attack is triggered by an increased demand for oxygen caused by physical stresses. Heart disease remains the leading cause of death in the United States and contributes extensively to healthcare costs to society. The most common group of patients with refractory angina pectoris has coronary artery disease that is not corrected by bypass grafting, stent placement, or aggressive medical management. Another group of refractory patients demonstrates normal coronary angiographies but has significant intermittent angina discomfort. This condition is sometimes referred to as microvascular disease or small vessel disease. On exercise electrocardiogram (EKG), the patients have typical exercise-triggered angina with ST segment depression. Since they generally fail to respond to conventional anti-angina therapy, they remain a treatment dilemma. Some of the theories as to the cause of this syndrome include endothelial dysfunction, abnormal distribution and function of adenosine receptors, and estrogen deficiency. These are treatment dilemmas for the cardiac team. Ischemic changes occur due to the sympathetic inhibition that allows for more blood to flow in the ischemic areas. This then allows for the peripheral circulation to improve cellular activity and overall blood flow.
383
384 Chapter 17 Spinal Cord Stimulation: Uses and Applications
By inhibiting circulating catecholamine, the circulatory system behaves by producing a vasodilatory effect, thereby allowing more nutrients to flow to areas that are deprived of vital nutrients that allow cellular activity to function without forming lactic acidosis and increasing pain in the ischemic area.
Interstitial Cystitis/Pelvic Pain Interstitial cystitis is a chronic debilitating disorder of the urinary bladder characterized by chronic pelvic pain, irritative voiding symptoms, and sterile cytologically sterile urine.32 It affects over 500 million people in the US, and 90% are women.33 A possible reason why males may go undiagnosed is that it may be mistaken for prostodynia and chronic nonbacterial prostatitis. Patients suffering with interstitial cystitis typically have micro-hemorrhagic petechiae. Interstitial cystitis remains the most debilitating nonmalignant disorder seen by urologists. Typically, patients with interstitial cystitis will suffer with the condition for 3–7 years before the diagnosis is even made. Among the most prevalent symptoms of interstitial cystitis are pelvic and perineal pain.34 Cytoscopic and histologic studies do not correlate well with the clinical pathology. Subjective measurements of symptoms are correlated with urodynamic studies. The relationship between chronic interstitial cystitis and neurogenic inflammation suggests that neuromodulation will be a useful therapeutic modality in this situation. Upon histologic examination of bladder biopsies from patients with long-standing findings of interstitial cystitis, the most common findings are marked edema of various elements due to extravasations of plasma and injury to blood vessels and nerves in the muscularis layer, which are all consistent with pathological findings of neurogenic inflammation. If conservative and minimally invasive treatments do not improve the condition, then the consideration of spinal cord stimulation may allow patients to live daily without the restraints of pelvic pain.
Technical Factors in Peripheral Stimulation By placing a lead or leads in various places alongside peripheral nerves, the areas outside of the Central Nervous System (CNS) can be stimulated. This allows direct stimulation over the affected nerve that feels dysesthetic and painful. The threshold for peripheral stimulation is much lower than the threshold for the stimulation of the central nervous system. Examples of peripheral stimulation are ilioinguinal and genitofemoral nerves, median nerve, intercostal nerves, occipital nerves, and supraorbital nerves. Occipital stimulation has a tremendous role in alleviating transformed migraine headaches. Patients who suffer from headaches can use the device each and every time they begin to have a headache; instantaneously, the headache is alleviated. There are various ways to place leads in the occiput. The pain can be band-like in nature, gripping and
Technical Factors in Spinal Cord Stimulation
sometimes stabbing. The headache can last several seconds to hours. Some patients even develop an aura prior to the pain beginning. Pharmacologic treatments are usually the first line in treating headaches. Medications include a variety of types such as NSAIDS, Ergot alkaloids, sumatriptins, neuroleptics, and occasionally opioids. Most practitioners avoid opioids due to the strong habituation tendencies seen with this medication and the frequent occurrence of rebound headaches. Placement of the electrodes is crucial to pain relief. The lead(s) should be placed as close to the anatomical area known to where the nerve(s) cross in order to successfully cover the area with a paresthesia sensation. If not, then stimulating the area will typically not bring about relief. Below the occipital protuberance is the starting point, fluoroscopically noted to be at the C 1–2 intervertebral space. After placement, stimulation helps to localize the point of maximum pain. While the common range of frequency of stimulation (Hz) ranges from 50 to 110, a typical preferred pulse width is between 100 and 400 ms. Depending on the depth of the lead placement, initial stimulation perception can be as low as 0.3 mA. Preferred current levels are usually within 2.0–4.5 mA range. Suturing is performed at the site of lead placement utilizing an anchor to secure the lead(s) to the fascia below and also by placing a strain loop to alleviate the stress upon the lead(s). Tunneling is the next step. Where the chosen site for the battery/generator will be placed will determine other factors, such as length of leads to be used and site of battery implantation. Some typical areas are in the lateral axillary line where a “bra” strap would be located. Some patients choose the anterior subclavian area, and others prefer the buttock/gluteal area (Figure 17.5). Each place has its advantages and disadvantages. The determining factors are ease of recharging the device, comfort, and where one typically wears their belt.
Technical Factors in Spinal Cord Stimulation Spinal Cord Stimulation is a procedure that must be performed under fluoroscopic guidance. Understanding the three-dimensional views of the spinal canal and the anatomy is crucial to proper insertion and alleviation of pain at the site of the target. Proper sterile technique needs to be adhered to. Gown and gloves should be used, and sterile preparation with betadine and alcohol should take place in order to create a sterile field for insertion. Proper antibiotic administration of appropriate medications and proper time allowed for adequate circulating levels need to be followed to ensure tissue level coverage from the antibiotic administered. Mask and sterile technique are essential to prevent infections from developing. The patient is placed on a wellpadded carbon fiber fluoroscopic table to allow full-field visualization of the spine in an anterior–posterior view as well as a lateral view. Understanding of anatomic placement will allow the implanter to access the epidural space with ease. This will allow for insertion of 1, 2, 3, or 4 leads with less difficulty.
385
386 Chapter 17 Spinal Cord Stimulation: Uses and Applications Figure 17.5. Buttock gluteal battery placement.L
With placing, 1, 2, 3, and even 4 leads can be inserted into the epidural space, either from an anterograde or retrograde manner. This will allow stimulation over the sacral plexus. By stimulating the sacral 2, 3, and 4 nerve rootlets, one can inhibit afferent impulses from entering the dorsal cord painfully.
Complications Complications include wound infections, lead fractures, inadvertent spinal taps, epidural punctures, lead migrations, inability to cover the area desired, seromas, hygromas, hematomas, and spinal cord trauma. An infection can develop each and every time the skin barrier is violated. The most common of infections seen and dealt with are those caused by Staphylococcus aureus. Staphylococcus epidermidis is another well-known felon. In this era of resistant bacteria, coverage for resistant strains of bacteria needs to be considered in addition to simple everyday bacterial coverage. Methicillin-resistant Staphylococcus aureus is becoming a “common” bacteria, one that is difficult to eradicate because of the super infections that have developed from overuse/
Techniques and Pearls of Implantation
misuse of antibiotics. Sterile precautions need to be taken each and every time the skin barrier is violated. Proper prepping and draping is essential for the prevention of microbacteria contamination. Gowning and gloving, and adherence to a sterile technique are imperative. Even with these measures, widespread infections are still occurring. Lead fractures and interruption in conductivity are issues that need to be looked into when coverage is interrupted. Pain at the site of the fracture is typically seen. Dysesthesia, burning type pain, is commonly felt by the patient. If impedance is elevated, it is more than likely a fracture has happened. One way to ascertain the integrity of the system is to perform an impedance check. This will help to verify if a problem does exist. Also, a plain X-ray should also be obtained to locate possible disconnections and lead fractures. Lead migration is a major concern because of improper anchoring techniques, and mechanical motion of the spine also contributes to lead migration. By using appropriate anchors, glue, and suturing techniques, the lead(s) should stay in place properly throughout the course of the implantation. Using the supraspinous ligament, spinous process, or deeper fascial tissue should anchor the lead and should keep it steady and not allow it to move either cephalad or caudad. Leads also can migrate laterally in a so-called windshield wiper effect. Usually, all forms of migration can be mitigated by proper anchoring techniques. Fingertrap suturing also can aid in securing the lead(s) in place. Once the system is in place, the integrity of the system should be checked to determine whether any iatrogenic fractures might have occurred. After the process is complete, ample irrigation is performed to help flush out any bacteria contamination, as “the solution to pollution is dilution.” Seromas, hygromas, and hematomas can be avoided by strict hemostasis techniques, and also by creating a pocket that is adequate for the battery device that is being implanted. Not allowing excessive motion and contact between the interface of the battery and subcutaneous tissue will minimize any chance of a seroma from forming. Along with the previously mentioned complications, there are several other complications that can develop including: a dural tear, CSK leak, cord compression from either trauma from the lead insertion or unrecognized spinal canal stenosis that develops from iatrogenic causes, infection, inability to obtain the desired results, lead movement or migration, persistent pain from dural irritation, seromas, hygromas, rejection and allergic reactions, local skin erosions, paralysis or weakness, and pain at implant site.
Techniques and Pearls of Implantation Patient positioning is a major key element in accurately inserting a lead into the epidural space. A large pillow should be placed under the pelvic brim for males and under the abdomen for women. By reversing the lumbar lordosis, the interlaminar spaces open to their widest space. This will allow easier access into the epidural space for the insertion needle. A 30° angle into the epidural space is desired. The typical way to enter is by starting one interlaminar space below
387
388 Chapter 17 Spinal Cord Stimulation: Uses and Applications
the targeted entrance site. Then, after anesthetizing a tract down to the lamina, use an 18-gauge needle or an #11 blade to nick the skin. Following this, use the bubble technique and the “LOR” (Loss of Resistance) technique with a “pulsator” syringe with saline. Enter the epidural space, and use some saline to expand the space. Then guide the lead up to the desired level of dermatomal coverage. Steering can be performed in various ways, and each implanter will develop techniques that will allow the procedure to be performed smoothly. After proper anterior and lateral views to determine posterior epidural lead placement, connect the leads to the testing cables and begin the stimulation process to evaluate the adequate paresthesia coverage over the painful zone and desired relief. If a lead meets resistance and cannot be advanced, then a simple technique of retracting the inner guide wire approximately 2–3 cm from the tip can be performed. Twirl the lead while advancing the lead forward. This may be sufficient to redirect the lead in the direction desired. Lead location is determined by the segment of the spinal cord that correlates to the anatomical area of dorsal column fibers located throughout the spinal cord. Anchoring is another very crucial and important part of the procedure. There are several types of anchors that can be used. Different products advocate the use of a specific types of anchors, from the hard anchor to the bumpy and sleeve types. Each allows for a suture to be placed around the lead. Then it is secured to the fascia between the ligaments. Some products even advocate the use of a glue to help secure the lead to the anchor. There is even a technique of simply securing the lead to the subcutaneous tissue without an anchor to hold the lead. This allows for some movement to occur when the patient either flexes or extends. Tunneling and generator placement are the next steps. The proper depth and location need to be considered in order not to encounter an irritative source such as a belt or an object when the patient sits. Proper depth is crucial with rechargeable units. If the battery is too deep, then contact will not occur, and the battery will not be able to engage for recharging. If the battery is placed too superficial, then the contact area may produce dysesthetic phenomena at the site of contact. Landmarks are crucial for precise placement. First, identify the anatomical midline. Aligning the spinous process, pedicles, and endplates of the entrance site will allow for an easier placement. The lamina should be as open as possible. Remember that reversing the lordosis in the lumbar area allows for this to occur. The 12th rib is also a point of reference. The anatomical midline and the physiologic midline are two different areas. One may not coincide with the other. One should gauge oneself via the anatomical one first, but be ready to steer the lead(s) off the midline if the patient does not perceive stimulation. The spinal cord can twist and turn throughout its course and not correlate to the anatomical midline. After all is said and done, confirm impedance prior to closing the wound in order to save the painstaking time required to revise a freshly placed stimulator. Lead placement and configuration depends on a patient pain pattern (Figure 17.6). Each patient designs their own lead configuration by
Figure 17.6. Lead placement.
390 Chapter 17 Spinal Cord Stimulation: Uses and Applications Figure 17.6. Continued.
indicating to the implanter where his/her pain is perceived and how the distribution of the pain pattern can be covered. Midline cervical pain patterns can be covered with a single lead array. Dual cervical lead placement can also be used for several reasons: lead stabilization, using a guarded cathode configuration, and allowing a broader coverage area with two leads. Mid to upper thoracic lead placement can be used for Angina, Mid epi gastric pain, and Pancreatic pain. Lumbar Lead placement is typically located between the T8and T10 area midline in the thoracic spine.
CPT Coding Device The standard codes that are acceptable and used are 63650 for lead placement, 63685 for battery/generator placement, 63660 for lead removal, 63688 for system revision, and 77003 for fluoroscopic guidance. The typical ICD-9 codes are 724.4 Thoraco-Lumbar Neuralgia, 722.83 for Failed Back Laminectomy Syndrome, and 722.21 and 722.22 for RSD of upper and lower extremity (CRPS Types 1 & 2).
Device Manufacturers
391
Device Manufacturers Three companies offer devices for implantation: Medtronic (Medtronic Inc., Minneapolis, MN) (the oldest), St. Jude Medical (St. Paul, MN), and Boston Scientific (Natick, MA). All have very good products, but each has one item that the other does not have. Medtronic uses a voltage called constant voltage vs. constant current. An example of this is when a car is placed on cruise control and ascends up a hill, the accelerator increases to overcome the resistance. As for constant voltage, the voltage stays constant, the current does not. Medtronic’s handheld programmer (Figure 17.7a) (used with its RestoreULTRA™ neurostimulation system, Figure 17.7b) is simple to use and user friendly. Having a set of controls in hand allows a patient to easily increase or decrease the energy requirements that he/she may need, as well as to move the stimulation up and down the electrodes to maximize pain control.
Figure 17.7. Medtronic’s patient programmer (a) used with its Restore ULTRA™ neurostimulation system (b) (Photos courtesy of Medtronic, Inc., Minneapolis, MN).
392 Chapter 17 Spinal Cord Stimulation: Uses and Applications Figure 17.7. Continued.
Figure 17.8. ANS product family (Image provided courtesy of St. Jude Medical, St. Paul, MN. Image Copyright St. Jude Medical, all rights reserved).
St. Jude Medical offers an array of leads, paddles, steering devices, rechargeable batteries, and constant current energy deliverance (Figures 17.8 and 17.9). The newest addition to an already wellstocked armamentarium is the smallest rechargeable Implantable Pulse Generator (IPG), the EON™ Mini (Figure 17.10), which can allow for up to 16 contacts to cover the area in question that generates pain either
Device Manufacturers
393
Figure 17.9. EON®, EON™ Mini, EON™C (Image provided courtesy of St. Jude Medical, St. Paul, MN. Image Copyright St. Jude Medical, all rights reserved).
Figure 17.10. EON™ Mini (Image provided courtesy of St. Jude Medical, St. Paul, MN. Image Copyright St. Jude Medical, all rights reserved).
in the spinal cord or in the periphery. It allows for longer rechargeable cycles and easier placement for patients comfort. Boston Scientific has the smallest battery and i-sculpt technology (Figure 17.11). This allows the programmer to shift the energy from one lead to another and cover certain areas.
394 Chapter 17 Spinal Cord Stimulation: Uses and Applications
Figure 17.11. Precision Plus™ system (Used with permission of Boston Scientific, Natick, MA).
Energy is the limiting factor, meaning the more the device is used, the less battery life it will have. With the rechargeable system, one can use higher energy levels and not be as concerned with the battery life. All battery lives do end at some point. Medtronic’s ends at the 9-year mark. St. Jude Medical has been able to extend the life of its battery to 10 years or more, and Boston Scientific’s battery ends at approximately the 5-year mark.
Conclusion Spinal cord stimulation can provide significant long-term pain relief and improve quality of life in a variety of benign intractable paingenerating etiologies. The most beneficial effects are noted in cases of FBS, CRPS I and II, pain secondary to PVD, angina, multiple sclerosis, and peripheral neuropathy. In addition to improvements in pain intensity, patients also reported increases in social interactions, mood elevation, and various factors in daily living. SCS using multipolar and multichannel stimulation programs have improved long-term success rates significantly in comparison with the previous decade. The use of multiple leads allowing multiple electrode combinations and sophisticated programming appears to hold much promise in the treatment of patients with predominantly axial back pain, which in the past has been resistant to SCS therapy. The complication rate is low, which
Conclusion
makes SCS a relatively safe and effective approach in the management of long-term pain. The concept of tolerance continues to be the most significant challenge to long-term efficacy of SCS therapy, and further work is required to elucidate its pathophysiology. Prospective, randomized controlled studies that are currently in progress will further confirm the above conclusions. References 1. Shealy CM, Mortimer JT, Reswick JB. Electrical inhibition of pain by stimulation of the dorsal columns. Anesth Analg 1967;46:489-491. 2. Schecter DC. Origins of electrotherapy. Part 1. NY State J Med 1971;71(9): 997–1008. 3. Gadsby JG. Electroanalgesia: historical and contemporary development. PhD Thesis. De Montfort University, Leicester, UK, 1998. 4. Melzack R, Wall PD. Pain mechanisms: a new theory. Science 1965;250:971–978. 5. Wall PD, Sweet WH. Temporary abolition of pain in man. Science 1967;155:108–109. 6. Shealy CN, Taslitz N, Mortimer JT, Becker DP. Electrical inhibition of pain by stimulation: experimental evaluation. Anesth Analg 1967;299–305. 7. Davidoff RA. Handbook of spinal cord, vols. 2 and 3. New York: Marcel Dekker, 1984. 8. Carpenter MB. Human Neuroanatomy. Baltimore: Williams & Wilkins, 1976. 9. Strujik JJ, Holsheimer J, Boom HB. Excitation of dorsal root fibers in spinal cord stimulation: a theoretical study. IEEE Trans Biomed Eng 1993;40:632–638. 10. Strujik JJ, Holsheimer J, van Veen BK, Boom HB. Epidural spinal cord stimulation: calculation of field potentials with special reference to dorsal column nerve fibers. IEEE Trans Biomed Eng 1991;38:104–110. 11. Strujik JJ, Holsheimer J. Transverse tripole spinal cord stimulation: theoretical performance of a dual channel system. Med Biol Eng Comput 1996;34:273–279. 12. Mayer TG, Gatchel RJ. Functional Restoration for Spinal Disorders: The Sports Medicine Approach Philadelphia. Philadelphia: Lea & Febiger, 1988. 13. Croft PR, Joseph S, Cosgrove S, Jordan L, Papageorgiou A, Pope D, Ferry S, Jayson MIV, Silman A. Low back Pain in the community and hospitals. Report to the Clinical Standards Advisory Group of the Department of Health, 1994. 14. Frymoyer JW, Cats-Baril WL. An overview of the incidences and costs of low back pain. Orth Clin North Am 1991;22:263–271. 15. Wilkinson HA. The Failed Back Syndrome: Etiology and Therapy, 2nd ed. Philadelphia: Harper and Row, 1991. 16. Block AR, Gatchel RJ, Deardoff W, Guyer RD. The Psychology of Spine Surgery. Washington: American Psychology Association, 2003. 17. Oaklander Al, North RB. Failed back surgery syndrome. In Loeser JD, Butler SH, Chapman CR, Turk DC (eds): Bonica’s Management of Pain, 3rd ed. Philadelphia: Lippincott Williams & Wilkins, 2001. 18. Green LN. Dexamethasone in the management of symptoms due to herniated lumbar disc. J Neurol Psychiatry 1975;38:1211–1217. 19. Wilkinson HA. The Failed Back Syndrome, 2nd ed. New York: Springer, 1992. 20. Plewes LW. Sudecks atrophy in the hand. J Bone Joint Surg Br 1956;38:195–203. 21. Merskey H, Bogduk N. Classification of chronic pain: descriptions of chronic pain syndromes and definitions of pain terms, 2nd ed. Seattle: IASP, 1994;40–42. 22. Bonica J. Causalgia and other reflex sympathetic dystrophys. In Bonica J, Liebeskinar J, Albe-Fessard D (eds): Advances in Pain Research and Therapy: Proceedings of the Second World Congress on Pain. New York: Raven, 1979;141–166. 23. Law JD. Spinal stimulation: statistical superiority of monophasic stimulation of narrowly separated, longitudinal bipoles having rostral cathodes. Appl Neurophysiol 1983;46:129–137.
395
396 Chapter 17 Spinal Cord Stimulation: Uses and Applications 24. Law JD. A new method for targeting a spinal stimulator: quantitatively paired comparisons. Appl Neurophysiol 1987;50:436. 25. Law JD. Clinical and technical results from spinal stimulation for chronic pain of diverse pathologies. Stereotact Funct Neurosurg 1992;59:21–24. 26. Tulgar M, He J, Barolat G, Ketick B, Strujjik H, Holsheimer J. Analysis of parameters for epidural spinal cord stimulation. Stereotact Funct Neurosurg 1993;61:146–155. 27. Holsheimer J, Strujik JJ, Rijkhoff NJ. Contact combinations in epidural spinal cord stimulation. A comparison by computer modeling. Stereotact Funct Neurosurg 1991;56:220–233. 28. Holsheimer J, Strujik JJ. How do geometric factors influence epidural spinal cord stimulation? Quantitative analyses by computer modeling. Stereotact Funct Nerosurg 1991;56:234–249. 29. He J, Barolat G, Holsheimer J. Perception threshold and electrode position for spinal cord simulation. Pain 1994;59:55–63. 30. Law JD, Kirckpatrick AF. Intractable pain of both arms and legs can be treated with complex spinal cord stimulation. Abstract of the 7th World Congress on Pain. Seattle: IASP, 1993:422. 31. Gibbons RJ, Chatterjee K, Daley J, Douglas JS, Fihn SD, Gardin JM, Grunwald MA, Levy D, Lytle BW et al. ACC/AHA/ACP-ASIm guidelines for the management of patients with chronic stable angina: a report of the American College of Cardiology/Heat Association Task Force on Practice Guidelines. J Am Coll Cardiol 1999;33:2092–2197. 32. Feler CA, Whitworth LA, Brookoff D et al. Recent advances: sacral nerve root stimulation using a retrograde method of lead insertion for the treatment of pelvic pain due to interstitial cystitis. Neuromodulation 1999;2:211–216. 33. Jones CA, Nyberg Jr LM Jr. Epidemiology of interstitial cystitis. Urology 1997;49:2–9. 34. Koziol JA, Clark DC, Gittes RF, Tan EM. The natural history of interstitial cystitisa survey of 374 patients. J Urol 1993;149:465–469.
Index
A Analgesics adjuvant, 36 NSAID, 35 opioids, 35–36 Anesthetic agents, 261–262 amide-linked group, 33 bupivacaine, 34 ester group, 33 lidocaine, 33–34 ropivacaine, 34 Aneurysmal bone cysts (ABCs), 322–323 Angina pectoris, 383–384 Antibiotics, 261 Aortogram, 327 Arachnoiditis, 381–382 Arteriovenous malformations (AVMs), 299 Asympotomatic hemangiomas, 322 Autonomic nerve blockade celiac plexus blockade anatomy, 235–237 contraindications, 239 indications, 236 injection technique, 236–238 common sympathetic blockades, 230 hypogastric plexus blockade, 243–245 impar ganglion blockade, 243–244, 246 lumbar sympathetic blockade anatomy, 239, 240 indications, 239 injection technique, 239 radiofrequency block technique, 239–242 radiofrequency ablation, 230 stellate ganglion blockade anatomy, 230, 231 complications and side effects, 232 indications, 230 injection technique, 230–232 radiofrequency-induced blockade, 232 sympathetic nerve blockade, 230 thoracic and splanchnic sympathetic blockades anatomy, 233, 234 complications and side effects, 235 indications, 233
injection technique, 233–234 radiofrequency block technique, 235 B Balloon assisted vertebroplasty. See Kyphoplasty Bones, anatomy parapedicular approach, 6 percutaneous vertebroplasty (PV)/kyphoplasty (KP), 1 scanning, 46–47 straight sagittal (anterior-to-posterior) orientation, 5 venous system, 6 vertebra size, 1 C Cavernous malformations, 313–314 Celiac plexus blockade anatomy, 235–237 contraindications, 239 indications, 236 injection technique, 236–238 Cement cortoss, 258–259 injection cannula, 267 pain relief, 267–269 small syringes, 266–267 leak risk, 250 polymethylmethacrylate (PMMA) antibiotics, 258, 261 bone reconstruction and augmentation, 249 disadvantages, 258–259 precaution, 258 selection and preparation, 257–258 Cerebrospinal fluid leak, 292 Cervix epidural injections lumbar, 168 procedure, 168–170 selective nerve root blocks, 170–171 facet joints, 210–213 Chronic back pain, 378–379 Cobb’s syndrome, 317
397
398
Index
Complex regional pain syndrome (CRPS), 379–381 Computed tomography (CT) guidance, sacroplasty antibiotics, 363 imaging, 366 puncture sites, 363–364 trans-SI approach, 364 scanning, 45 Constant flow rate infusion pump, 280 Continuous epidural/intrathecal infusion trials, 285–286 Conus medullaris AVM, 309–310 Conventional radiographs, 44–45 Corticosteroids betamethasone, 32–33 methylprednisolone acetate, 32 CPT coding device, 390 Craig system, 97 CRPS. See Complex regional pain syndrome (CRPS) D Dekompressor®, 151, 154 Demographics, 158 Diagnostic angiography, 323 Discography cervical anatomy, 131 anterior paratracheal approach, 134–135 oblique approach, 135–138 post-procedure care, 138 procedure, 131–133 surgical planning, 139 complications, 108–109 conscious sedation, 109–110 disc morphology, 120, 121 discogram report, 110 indications, 108 lumbar anatomy, 111–112 functional anesthetic discography, 122–124 manometry, 121–122 positive predictive value (PPV), 110–111 post-procedure care, 120–121 procedure, 112 surgical outcomes, 124 surgical management, 58–61 thoracic anatomy, 125 post-procedure care, 130 procedure, 125–130 vertebrogenic pain, 141 Dorsal root ganglion (DRG)-PRF procedure billing, 197 complications, 191 contraindication, 185 efficacy, 195–196 equipment and drugs, 185–186 indications, 184–185 laboratory, 186 mechanism of action, 180–182
patient history, 186 patient positioning, 187–188 physical examination, 186 post treatment care, 191 pre-operative medication, 186 PRF lesioning history, 179–180 radicular pain, 177–179 surgical anatomy, 182–184 technique, 188–193 Dural arteriovenous malformations, 301–302 E Electromyography (EMG), 47 Endovascular treatment ABCs, 322–323 aortogram, 327 diagnostic angiography, 323 microcatheter, 331 neoplastic/metastatic lesions, 323 neoplastic vascular lesions cavernous malformations, 313–314 hemangioblastomas, 314 neurophysiologic monitoring, 331 spinal vascular diseases classification and clinical presentation, 300 conus medullaris AVM, 309–313 dural arteriovenous malformation, 301–303 extradural AV shunts, 299–301 extradural/intradural AVM, 307 intradural (pial) arteriovenous fistula, 303–307 intramedullary AVM, 307–308 isolated spinal aneurysms, 313 systemic syndromes Cobb’s syndrome, 317 Klippel−Trenaunay (KT) syndrome, 317 neurofibromatosis type-1, 317–318 Osler–Weber–Rendu Syndrome, 316–317 Parkes−Weber (PW) syndrome, 317 vertebral hemangiomas asympotomatic hemangiomas, 322 characteristics, 318 lesions causing cord compression, 319–321 painful lesions, 321–322 Epidural arteriovenous fistulas (AVS), 301 Epidural injections cervical lumbar, 168 procedure, 168–170 selective nerve root blocks, 170–171 demographics, 158 fluoroscopic guidance, 161–162 historical perspectives, 157–158 patient selection, 158–159 coagulation, 159–161 contraindications, 161 physical examination, 159 steroid, 64–65 techniques catheter, 164–165
Index caudal, 163–165 intralaminar, 162–163 transforaminal, 166 Extradural AV shunts, 299–301 Extradural/intradural AVM, 307 Extrapedicular approach, 340 F Facet blocks, 61–62 Facet joint injections anatomy cervical facet joints, 210–213 lumbar facet joints, 209–210 sagittal and axial diagram, 208–209 thoracic facet joints, 213 clinical presentation, 207–208 contraindications, 214 diagnostic criteria, 213 pathophysiology, joint pain, 210 radiofrequency ablation, 216 techniques, 214–216 lateral cervical facet approach, 219 lumbar intra-articular, 216–217 lumbar periarticular, 217 medial branch block, 217–218 posterolateral cervical facet approach, 219 thoracic facet block, 219–220 Failed back spinal syndrome (FBSS), 378–379 Fine-needle aspiration techniques, 76–77 Fluoroscopic guidance, sacroplasty, 369 Fluoroscopy, 198–199 G Gate control theory, 376–378 Gentamicin, 125–126 Gout and headache, 376 H Headache. See Gout and headache Hemangioblastomas, 314 Hereditary hemorrhagic telangiectasia. See Osler−Weber−Rendu syndrome Hypogastric plexus blockade, 243–245 I Imaging equipments biplane fluoroscopic equipment, 29–30 computed tomography (CT), 31–32 fixed-base C-arm fluoroscopic arrangement, 29–31 Impar ganglion blockade, 243–244, 246 Implanted drug delivery systems advantages, 282 constant flow rate infusion pump, 280 continuous epidural/intrathecal infusion trials, 285–286 device-related complications
allergic reaction, 294 catheter-related problems, 294 pump-related problems, 292–293 SynchroMed® II, 293 infusate-related complications bridge bolus calculation, 295 drug compounding errors, 295 programming errors, 294–295 technician errors, 295–296 intraspinal drug delivery clinic, 282–283 intrathecal therapy, 2–3 outcomes, 289 patient screening, 285 patient selection, 283–285 personal therapy manager (PTM), 282 programmable pump, 280–281 single intrathecal bolus dosing, 285 surgical complications bleeding, 290 cerebrospinal fluid leak, 292 infection, 291–292 neurological injury, 289–290 seroma formation, 290–291 system implantation implant checklist, 288 patient positioning, 286–287 procedure, 287–288 Intercostal neuralgia, 382–383 Interstitial cystitis/pelvic pain, 384 Intradural (pial) arteriovenous fistula (AVF), 303–307 Intramedullary AVM, 307–309 Intraspinal drug delivery clinic, 282–283 Intraspinal opioids, 284 Intrathecal drug delivery device-related complications allergic reaction, 294 catheter-related problems, 294 pump-related problems, 292–293 SynchroMed® II, 293 infusate-related complications bridge bolus calculation, 295 drug compounding errors, 295 programming errors, 294–295 technician errors, 295–296 surgical complications bleeding, 290 cerebrospinal fluid leak, 292 infection, 291–292 neurological injury, 289–290 seroma formation, 290–291 Isolated spinal aneurysms, 313 K Kirschner wire (K-wire), 342 Klippel−Trenaunay (KT) syndrome, 317 Kyphoplasty (KP), 67–68, 271–273 biomechanical investigations, 351 lateral radiograph demonstration, 338 patient selection
399
400
Index
Kyphoplasty (KP) (cont.) relative contraindications, 339 selection and exclusion criteria, 339 results complication rate, 350 lateral radiograph, 349 vs. PV, 350 technique anteroposterior image, 347 cannula and stiletto, 344 extrapedicular approach, 340 inflation end points, 346 Kirschner wire (K-wire), 342 lateral radiograph, 349 materials required, 341 posterolateral approach, 342 technically demanding procedure, 349 VCFs, 337 L Low back pain (LBP), 57 Lumbar discectomy. See Percutaneous lumbar discectomy Lumbar discography anatomy, 111–112 functional anesthetic discography, 122–124 manometry, 121–122 positive predictive value (PPV), 110–111 post-procedure care, 120–121 procedure anterior, posterior tear and posterior bulge, 118, 119 curved tipped needle, 115, 116 disc morphology, 120–121 fluoroscope, 114 lateral and AP imaging, 118 superior articular process (SAP), 114, 115 surgical outcomes, 124 Lumbar facet joints, 209–210 Lumbar sympathetic blockade anatomy, 239, 240 indications, 239 injection technique, 239 radiofrequency block technique, 239–242 M Magnetic resonance imaging (MRI), 45–46, 251–253, 256–257 Microcatheter, 331 Minimally invasive intradiscal therapy, 65–66 Motor evoked potentials (MEP), 331 Myelography, 45 N Neoplastic/metastatic lesions, 323 Neoplastic vascular lesions cavernous malformations, 313–314 hemangioblastomas, 314
Nerve conduction studies (NCVs), 48 Neurofibromatosis type-1, 317–318 Neurolytic (cytotoxic) agents, 37 Neuromodulation/neurodestructive procedures, 54–55 Neuropathic pain treatment algorithm, 41–42 Neurostimulation history, 375–376 Nonsteroidal anti-inflammatory (NSAID), 35 Nuclear medicine (NM), 252–253, 255 O Occipital neuralgia, 383 Opioids, 35–36 Osler−Weber−Rendu syndrome, 316–317 P Pain relief, 273–274 Parkes−Weber (PW) syndrome, 317 Patient evaluation ablative techniques, 40 diagnostic nerve blocks central nerve blocks, 50 false positive results, 48–49 peripheral nerve blocks, 49–50 electrodiagnostics electromyography (EMG), 47 nerve conduction studies (NCVs), 48 imaging studies bone scanning, 46–47 computed tomography scanning (CT), 45 conventional radiographs, 44–45 magnetic resonance imaging (MRI), 45–46 myelography, 45 thermography, 47 ultrasound, 46 laboratory tests, 48 medical history, 42–43 physical examination, 43–44 procedure selection criteria algorithmic approach, 52–53 neuromodulation/neurodestructive procedures, 54–55 pain generators, 51–52 radiofrequency lesioning techniques, 54–55 psychological evaluation, 50–51 Percutaneous lumbar discectomy anatomy, 147–148 complications and consent, 152 historical perspective, 149–150 intervertebral discs, spinal pain, 149 needle placement bulged disc, 150, 151 Dekompressor®, 151, 154 lateral view, 150, 153 P–P proper needle placement, 150, 152 Percutaneous sacroplasty (PS), 357 Percutaneous spine biopsy complications, 76–78 destructive/space-occupying lesions, 76
Index equipment requirements biopsy needle systems, 80–81 choice, 79 fluoroscopy-guided biopsies, 79–80 fine-needle aspiration techniques, 76–77 hemorrhage/hematoma formation, 76, 77 indications, 76 patient preparation axial CT image, 78–79 laboratory parameters, 78 postoperative care, 101 soft tissue dysphagia, 98, 99 presacral abscess, 101–105 sacroiliac osteomyelitis and abscess, 100, 101 techniques approaches, 81–82 Craig system, 97 disc and vertebral endplate, 84 hematoma formation, 76, 77 hypervascular lesions, 97 lesion location and lesion size approach, 81–82 microbiological analysis, 94 needle excursions, 91, 93, 94 parapedicular approach, 83, 84 posterior approach, 83 procedure-related complication, 94, 95 routes, 83 sclerotic/osteoblastic lesions, 94, 96 skin entry site, 85, 86 suprahyoid and infrahyoid compartments, 88, 89 transpedicular approach, 88, 92 vertebral body lesion and posterior element lesion, 82–83 Percutaneous vertebroplasty (PV) anesthesia, 261–262 antibiotics, 261 cement injection, 266–269 cement selection and preparation, 257–259 complications, 272–273 image guidance, 259–261 indications, 250 informed consent, 259 laboratory evaluations, 261 needle introduction and placement, 262–266 pain relief, 273–274 patient selection and workup, 250–257 postoperative care, 268, 270, 271 results, 270–272 venography, 266 vertebral compression fracture (VCF), 249–250 Peripheral nerve PRF procedure complications and efficacy, 201–202 peripheral nerve damage, 197, 198 post surgical pain treatment, surface landmarks, 199, 201 suprascapular nerve treatment, 198–200 technique, 197–198 Peripheral nerve stimulator, 383
401
Peripheral stimulation, technical factors, 384–385 Peripheral vascular disease, 383–384 Personal therapy manager (PTM), 282 Phantom limb pain, 382 Pharmacological agents analgesics adjuvant, 36 NSAID, 35 opioids, 35–36 anesthetic agents amide-linked group, 33 bupivacaine, 34 ester group, 33 lidocaine, 33–34 ropivacaine, 34 antibiotics, 34–35 corticosteroids betamethasone, 32–33 methylprednisolone acetate, 32 neurolytic (cytotoxic) agents, 37 radiographic contrast agents, 36–37 PHN. See Post-herpetic neuralgia Polymethylmethacrylate (PMMA) antibiotics, 258, 261 bone reconstruction and augmentation, 249 disadvantages, 258–259 precaution, 258 selection and preparation, 257–258 Posterolateral approach, 342 Post-herpetic neuralgia (PHN), 381 Post surgical pain treatment, 199, 201 Pulsed radiofrequency (PRF) application, 175 DRG procedure billing, 197 complications, 191 contraindication, 185 efficacy, 195–196 equipment and drugs, 185–186 indications, 184–185 laboratory, 186 mechanism of action, 180–182 patient history, 186 patient positioning, 187–188 physical examination, 186 post treatment care, 191 pre-operative medication, 186 PRF lesioning history, 179–180 radicular pain, 177–179 surgical anatomy, 182–184 technique, 188–193 pain medicine, 176–177 peripheral nerve procedure complications and efficacy, 201–202 peripheral nerve damage, 197, 198 post surgical pain treatment, surface landmarks, 199, 201 suprascapular nerve treatment, 198–200 technique, 197–198 radiofrequency current effect, 175–176
402
Index
R Radiofrequency lesioning techniques, 54–55 Radiographic contrast agents, 36–37 Reflex sympathetic dystrophy (RSD), 379 Retrocrural approach. See Splanchnic nerve denervation Retroneural approach. See Dorsal root ganglion (DRG)-PRF procedure RSD. See Reflex sympathetic dystrophy (RSD) S Sacral fractures anatomy, 361 diagnosis, 358–361 incidence, 357–358 morbidity and mortality, 358 patient selection, 361–362 Sacroiliac joint (SI) injections, 62–63 anatomy, 220–223 clinical presentation, 222 diagnosis, 222–226 innervation, 221–222 techniques, 226 Sacroiliac osteomyelitis, 100, 101 Sacroplasty complete clinical picture, 369–371 complications, 369 conventional treatment, 355–356 CT guidance antibiotics, 363 CT imaging, 366 puncture sites, 363–364 trans-SI approach, 364 diagnosis, 355 fluoroscopic guidance, 369 outcomes, 369 percutaneous sacroplasty (PS), 357 sacral fractures anatomy, 361 diagnosis, 358–361 incidence, 357–358 morbidity and mortality, 358 patient selection, 361–362 typical patient presentation, 355 Selective nerve root blockade, 63–64 SI injections. See Sacroiliac joint (SI) injections Single intrathecal bolus dosing, 285 Somatosensory evoked potentials (SSEP), 331 Spinal cord stimulation angina pectoris and peripheral vascular disease, 383–384 arachnoiditis, 381–382 chronic back pain, 378–379 complications, 386–387 CPT coding device, 390 CRPS, 379–381 device manufacturers, 391–394 dorsal roots, 378 FBSS, 378–379 gate control theory and implantable stimulators, 376–378
gout and headaches, painful condition, 376 intercostal neuralgia, 382–383 interstitial cystitis/pelvic pain, 384 neurostimulation history, 375–376 occipital neuralgia, 383 peripheral stimulation, technical factors, 384–385 phantom limb pain, 382 PHN, 381 technical factors, 385–386 techniques and implantation pearls, 387–390 Spinal vascular diseases classification, 299 conus medullaris AVM, 309–313 dural arteriovenous malformation, 301–303 extradural AV shunts, 301 extradural/intradural AVM, 307 intradural (pial) arteriovenous fistula (AVF), 303–307 intramedullary AVM, 307–309 isolated spinal aneurysms, 313 Spine anatomy anatomical spaces, 14–15 bones parapedicular approach, 6 percutaneous vertebroplasty (PV)/kyphoplasty (KP), 1 straight sagittal (anterior-to-posterior) orientation, 5 venous system, 6 vertebra size, 1 intervertebral discs and joints, 8 macrocirculation embryological development, 15 radicular arteries, 18 radiculomedullary arteries, 18 radiculopial arteries, 18 segmental arteries, 17 microcirculation centrifugal system, 19 centripetal system, 21 nerves lumbar foramina, 12 neural foramina, 7 parasympathic nerves and sympathetic nerves, 13 somatic arterial supply angiography, 22 intervertebral disc, 8–11 venous epidural veins, 26 intrinsic venous system, 23 lumbar level, 19 multiple longitudinal venous channels, 25 thoracic level, 19 Splanchnic nerve denervation, 235 Splanchnic sympathetic blockade. See Thoracic and splanchnic sympathetic blockade Stellate ganglion blockade anatomy, 230, 231 complications and side effects, 232 indications, 230 injection technique, 230–232 radiofrequency-induced blockade, 232
Index Suprascapular nerve treatment, 198–200 Surgical management discography, 58–61 epidural steroid injections, 64–65 facet blockade, 61–62 low back pain (LBP), 75 minimally invasive intradiscal therapy, 65–66 sacroiliac joint (SI) injections, 62–63 selective nerve root blockade, 63–64 vertebroplasty and kyphoplasty, 66–68 Sympathetic nerve blockade, 230 Systemic syndromes Cobb’s syndrome, 317 Klippel−Trenaunay (KT), 317 neurofibromatosis type-1, 317–318 Osler−Weber−Rendu syndrome, 316–317 Parkes−Weber (PW) syndrome, 317 T Thermography, 47 Thoracic and splanchnic sympathetic blockade anatomy, 233, 234 complications and side effects, 235 indications, 233 injection technique, 233–234 radiofrequency block technique, 235 Thoracic discography anatomy, 125 post-procedure care, 130 procedure disc annulus, 126, 128 fluoroscopic views, 127, 129–130
gentamicin, 125–126 lamina, 126 Thoracic facet joints, 213 Transcostovertebral approach, 79, 83, 85, 88 Transpedicular approach, 88, 92, 263–265 Treatment of spine related pain syndromes. See Epidural injections U Ultrasound, 46 V Venography, 266 Ventral intradural AVF. See Intradural (pial) arteriovenous fistula (AVF) Vertebral compression fracture (VCF), 249–250, 337 Vertebral hemangiomas asympotomatic hemangiomas, 322 characteristics, 318 lesions causing cord compression, 319–321 painful lesions, 321–322 Vertebroplasty, 66–67 W Waddell’s signs, 44 Z Zygapophyseal joints, 49, 61
403