Thoái hóa thắt lưng hẹp: Chẩn đoán và Quản lý pptx

Vol 7, No 4, July/August 1999 239 The earliest description of lumbar stenosis is attributed to Antoine Portal, who in 1803 described Òtoo- narrow vertebral canalsÓ in hunch- backs with rickets. The concept of acquired lumbar spinal stenosis was popularized in the 1950s by Verbiest. Since then, advances in diagnostic and therapeutic modali- ties, coupled with the extension of life expectancy, have increased the rate of detection and subsequent operative intervention. The general incidence of degen- erative lumbar spinal stenosis ranges from 1.7% 1 to 8%. 2 Symp- toms typically develop in the fifth or sixth decade of life in association with osteoarthritic changes in the lumbar spine. No sex predomi- nance has been found, although degenerative spondylolisthesis associated with lumbar spinal stenosis is four times more com- mon among women. 3 No associa- tion has been found with occupa- tion or body habitus. The goal of this article is to pro- vide the reader with an under- standing of lumbar stenosis as an element of the degenerative pro- cess of aging commonly seen in the older population, as well as an overview of treatment options Basic Science Anatomically, lumbar stenosis is usually caused by a reduction in the space available for the neural elements due to variant osseous anatomy or filling of the spinal canal with hypertrophic tissue. Three typical canal shapes have been described: round, ovoid, and trefoil (Fig. 1). Trefoil canals have the smallest cross-sectional area and are associated with the highest incidence of symptomatic lumbar stenosis. 4 Arnoldi et al 5 classified lumbar stenosis as congenital, ac- quired, or combined (Table 1). The term Òcentral stenosisÓ is used when compression of the dural sac is the main component. ÒLateral stenosisÓ refers to compression of the nerve root in the lateral recess, in the neural foramen, or lateral to the neural foramen. 6,7 The pathophysiology of degener- ative lumbar stenosis usually begins with disk dehydration, resulting in a loss of disk height and bulging of the anulus fibrosus and ligamentum flavum into the spinal canal. These Dr. Hilibrand is Assistant Professor of Orthopaedic Surgery, Rothman Institute at Thomas Jefferson University, Philadelphia. Dr. Rand is Adjunct Professor of Orthopaedic Surgery and Spine Fellow, Vanderbilt Univer- sity, Nashville, Tenn. Reprint requests: Dr. Hilibrand, Rothman Institute, 5th Floor, 925 Chestnut Street, Philadelphia, PA 19107. Copyright 1999 by the American Academy of Orthopaedic Surgeons. Abstract Degenerative lumbar stenosis is a common cause of disabling back and lower extremity pain among older persons. The process usually begins with degenera- tion of the intervertebral disks and facet joints, resulting in narrowing of the spinal canal and neural foramina. Associated factors may include a developmen- tally narrow spinal canal and degenerative spinal instability. Nonoperative management includes restriction of aggravating activities, physical therapy, and anti-inflammatory medications. If nonoperative treatment has failed, surgical treatment may be appropriate. Decompression should be performed so as to ad- dress all clinically relevant neural elements while maintaining spinal stability. If instability is present, autogenous intertransverse bone grafting is recommended. There may be an advantage to augmenting some of these procedures with in- ternal fixation. Surgical success rates as high as 85% have been reported, but may be compromised by inadequate decompression, inadequate stabilization, or medical comorbidities. Short-term follow-up data indicate that operative man- agement provides more effective relief than nonoperative treatment, but prospec- tive studies comparing the effects of nonoperative and operative interventions on the long-term natural history of lumbar spinal stenosis are needed. J Am Acad Orthop Surg 1999;7:239-249 Degenerative Lumbar Stenosis: Diagnosis and Management Alan S. Hilibrand, MD, and Nahshon Rand, MD changes alter the loading of the facet joints, which together with the intervertebral disk form the three- joint spinal motion segment. Further degeneration leads to facet arthrosis with sclerosis and osteo- phytic overgrowth. The most com- mon result is that as the nerve roots traverse the lateral recesses, they may be encroached on by hyper- trophic facet joints, infolded liga- mentum flavum, and a bulging anu- lus. These degenerative changes can also cause root stenosis in the neural foramen. The anteroposteri- or diameter of the foramen is reduced by the bulging anulus ante- riorly and the hypertrophic facets posteriorly, while the foraminal height is reduced by the loss of intervertebral disk height and asso- ciated facet subluxation. 8 The degenerative process is sometimes accompanied by the development of segmental instabil- ity. Degenerative changes in the supporting structures of the spinal motion segment, including com- promise of the facet joints and cap- sular ligaments, may cause higher mechanical stress across the degen- erated anulus, leading to the devel- opment of dynamic subluxation or spondylolisthesis. 6 As abnormal motion develops within a degener- ated motion segment, it exacerbates nerve root irritation in the stenotic lateral recess and foramen. The sequence of neuropatholog- ic changes that result from stenosis of the lumbar spinal canal have been investigated in animal stud- ies. Delamarter et al 9 employed a dog model in which they created varying degrees of stenosis and demonstrated deleterious effects on the neural elements by increasing the degree of the stenosis. They found that cortical evoked poten- tials were highly sensitive to this compression and were affected long before any clinical signs oc- curred. They also demonstrated venous congestion and arterial con- striction around compressed nerve roots and dorsal root ganglia. The result was blockage of axoplasmic flow, with resulting edema, demye- lination, and wallerian degenera- tion of motor and sensory fibers. Other authors have shown that sensory fibers are more susceptible to pressure and slower to recover than motor fibers, 10 which may explain the presence of subjective sensory changes in the absence of objective physical findings. The mechanism of pain produc- tion in lumbar spinal stenosis re- mains unclear. Although many older patients have degenerative changes in their lumbar spines, few have any symptoms. Arnoldi et al 5 suggested that increased venous pressure may explain the symp- toms of neurogenic claudication. Others have suggested that narrow- ing of the spinal canal may lead to a reduction in blood supply to the cauda equina, resulting in ischemic changes from the diffusion of metabolites. 11 These changes may stimulate the sinuvertebral nerve or lead to secretion of pain mediators, such as substance P, from the dor- sal root ganglion. Perineural in- flammation of unknown origin may also result in pain generation. Clinical Presentation Symptomatic degenerative lumbar stenosis usually has an insidious onset and a slow rate of progres- Degenerative Lumbar Stenosis Journal of the American Academy of Orthopaedic Surgeons 240 Round Trefoil Ovoid Fig. 1 The three typical shapes of the spinal canal. Trefoil canals have the smallest cross- sectional area. Table 1 Classification of Lumbar Spinal Stenosis Congenital (developmental) stenosis Chondrodystrophy Idiopathic Acquired stenosis Degenerative Spondylolytic Iatrogenic Posttraumatic Miscellaneous Combined (degenerative changes superimposed on a congenitally narrow spinal canal) sion. At initial presentation, most patients have a long history of back pain in the lumbar region with recent development or progression of lower-extremity pain. Discom- fort in the back, buttocks, and/or lower extremities is the most com- mon complaint. Symptoms are usually exacerbated by standing, walking, and exercising in an erect posture, which results in the devel- opment of pain, tightness, heavi- ness, and subjective weakness in the legs. 12 This symptom complex, referred to as Òneurogenic claudi- cation,Ó is rapidly relieved by sit- ting down or leaning forward. Cycling, which involves forward flexion, is usually tolerated by pa- tients with lumbar stenosis. Pain- less motor claudication, deficits in proprioception leading to gait dis- turbances, and bowel and bladder dysfunction are relatively uncom- mon. Amundsen et al 12 reported that the most common symptoms in pa- tients with lumbar spinal stenosis were back pain (prevalence of 95%), claudication (91%), leg pain (71%), weakness (33%), and voiding dis- turbances (12%). In 70% of the patients in their study, the pain intensity was equally distributed between the back and the lower extremities; 25% had predominant- ly leg pain. Radicular pain was unilateral in 58% and bilateral in 42%. The radicular pattern corre- sponded to the L5 root in 91%, S1 in 63%, L1-L4 in 28%, and S2-S5 in 5%. In that study, 47% of the patients had double-root involvement, 35% had single-root involvement, 17% had triple-root involvement, and 1% had quadruple-root involve- ment. Even in the presence of symp- toms, there may be few associated physical findings. A thorough examination is required, however, to rule out other conditions that may cause referred pain to the lum- bar region or lower extremities. Abnormalities in gait may be related to the lumbar stenosis, although the possibility of cervical myelopa- thy or an intracranial disorder should also be considered, espe- cially in patients with a positive Romberg test. Patients frequently assume a ÒsimianÓ posture, with translation of the shoulders anterior to the pelvis. Tenderness, if present, is usually noted on palpation of the sciatic notches or the lumbosacral or sacroiliac region. Lumbar lordosis is generally reduced, and range of motion is diminished. There are usually no signs of nerve root ten- sion, although lower-extremity pain may be reproduced by forci- ble lumbar extension. The straight- leg-raising test may be positive if a concurrent disk herniation or nerve root entrapment is present. The remainder of the neurologic exami- nation is usually normal, at least when the patient is at rest. If mus- cle weakness is present, it is most often in the L5 root distribution. Postexercise examination may reveal greater motor weakness and help establish the diagnosis of lum- bar stenosis. Sensory examination should include vibration and propriocep- tion testing in addition to pin-prick testing. Responses may be altered by spinal stenosis or peripheral neuropathy, as may be seen in dia- betic patients. Deep tendon reflexes may be diminished due to involve- ment of the L3, L4, or S1 nerve roots. Loss of deep tendon reflexes is common among the elderly. Evaluation of hip and knee range of motion, joint irritability, and peripheral pulses should be documented. Amundsen et al 12 reported sensory changes in 51% of patients with lumbar stenosis; re- flex changes in 47%; lumbar ten- derness in 40%; reduced spinal mobility in 36%; positive straight- leg-raising test in 24%; weakness in 23%; and perianal numbness in 6%. Differential Diagnosis It is essential to rule out other con- ditions that may present with pain in the low back and/or lower extremity (Table 2). Peripheral vas- cular claudication should be differ- entiated from neurogenic claudica- tion. Both conditions are most prevalent in the older population and may coexist. Symptoms of vas- cular claudication are reproduced at a consistent level of exertion (e.g., walking two blocks). These symp- toms are most commonly described as cramping or tightness in the large muscle groups of the buttock, thigh, and leg. Peripheral vascular disease will result in diminished peripheral pulses; therefore, in patients with buttock, thigh, or leg pain, vascular studies should be obtained, with determination of ankle and brachial indices. Aortic aneurysms are a less common source of low back pain but have serious consequences. They may be palpable on examina- tion, and can be assessed with abdominal ultrasonography. The Alan S. Hilibrand, MD, and Nahshon Rand, MD Vol 7, No 4, July/August 1999 241 Table 2 Differential Diagnosis of Lumbar Spinal Stenosis Vascular conditions Peripheral vascular disease Aortic aneurysm Musculoskeletal diseases Degenerative arthritis of the hip Degenerative arthritis of the knee Pelvic and sacral disorders Neurologic disorders Diabetic neuropathy Peripheral compressive neuropathy Cervical myelopathy Amyotrophic lateral sclerosis Demyelinating disease Other Renal disorder Retroperitoneal tumors Depression Litigation-related issues aorta can frequently be visualized on axial spinal imaging. Osteoarthritis of the hip may be associated with buttock pain, a com- mon complaint in patients with lum- bar stenosis. These patients most commonly relate their symptoms to weight bearing on the affected limb, although radiation therapy to the groin or the anterior aspect of the thigh may be a factor with both hip arthritis and stenosis of the upper lumbar spine. Hip irritability and diminished range of motion, espe- cially in internal rotation and abduc- tion, are commonly seen with degenerative arthritis of the hip. Peripheral neuropathy, most commonly the result of diabetes mellitus, is another cause of lower- extremity complaints in the elderly. Generally, peripheral neuropathy presents with dysesthesias and paresthesias rather than activity- and position-related claudication. The anatomic distribution of symp- toms may help distinguish these two entities: in stenosis, sensation is more likely to be diminished in a dermatomal distribution; in contrast, diabetic neuropathy is characterized by a Òstocking/gloveÓ distribution. Electromyography and nerve con- duction studies are helpful in differ- entiating the two entities. Less common diagnoses, such as renal infections and fevers, retro- peritoneal tumors, and sacral or presacral lesions should be consid- ered as well. It should also be borne in the mind that certain nonorganic factors, such as depres- sion and involvement in litigation, may enhance the pain response. Diagnostic Modalities The primary role of imaging studies is to confirm the clinical diagnosis of lumbar stenosis, although ad- vanced imaging studies are also essential for preoperative planning. Upright plain radiographs are nec- essary to exclude pathologic condi- tions of the spine, such as tumor, infection, and fracture. The films may demonstrate hypertrophic facet joints narrowing the interlaminar space, but these plain-radiographic findings are not diagnostic of spinal stenosis. Dynamic views should be obtained to identify associated in- stability, such as that due to spon- dylolisthesis, scoliosis, or other spinal deformity. In patients with signs and symp- toms consistent with spinal steno- sis, magnetic resonance (MR) imag- ing or postmyelographic computed tomography (CT) is needed to con- firm neural element compression. Until the advent of MR imaging, the most widely utilized radiologic technique for evaluating spinal stenosis was myelography in com- bination with CT. On the myelo- gram, nerve root entrapment in the lateral recess or central canal steno- sis is demonstrated by the level of cutoff of contrast material. The postmyelographic CT images can then be used to identify the bone or soft tissue at each level that must be removed for decompression. Using myelography and CT, Bolen- der et al 4 correlated the degree of symptomatic lumbar stenosis with the anteroposterior diameter of the dural sac. A sac narrower than 10 mm was usually associated with clinical symptoms. Herno et al 13 also found that myelographic evi- dence of complete cutoff of contrast material and severe stenosis corre- lated with better surgical outcome. A disadvantage of myelography is that it requires injection of contrast medium into the spinal canal; post- injection spinal headache is rela- tively common. Magnetic resonance imaging is a noninvasive technique that can also define neural element compression through cross-sectional axial and sagittal imaging. This very sensitive method of evaluation should be used to confirm a clinical diagnosis of lumbar stenosis. In one study, 14 21% of asymptomatic individuals aged 60 to 80 years had MR imaging evidence of lumbar stenosis. Modic et al 15 prospectively found 79% agreement between the severity of lumbar spinal stenosis noted at surgery and the postmyelographic CT findings, compared with 77% agreement with MR findings. Two advantages of plain myelography over MR imaging include the ability to evaluate dynamic neural element compression on flexion-extension views and the improved definition of nerve root anatomy in scoliotic patients. Riew et al 16 recently compared the relative contributions of MR imaging and postmyelographic CT in surgical planning for patients with lumbar stenosis. The four academic spine surgeons who par- ticipated expressed a general pref- erence for MR imaging over post- myelographic CT for preoperative planning. However, the surgical plans derived from review of the postmyelographic CT study alone were much closer to those derived from both studies together than were the plans derived from a review of the MR study alone. Although the authors noted that the two modalities provided com- plementary information and were both important, they concluded that postmyelographic CT is supe- rior to MR imaging as a single study for the preoperative plan- ning of decompression for lumbar spinal stenosis. Electrophysiologic studies are rarely useful, except in identifying the presence and source of a peripheral neuropathy. About 80% of patients with symptomatic lum- bar stenosis will demonstrate elec- tromyographic changes, 17 usually consistent with single or multiple radiculopathies. The presence of these changes supports the diagno- sis of lumbar spinal stenosis, al- though their absence does not Degenerative Lumbar Stenosis Journal of the American Academy of Orthopaedic Surgeons 242 exclude the diagnosis. Somato- sensory evoked potentials (SSEPs) and dermatomal SSEPs may pro- vide additional information by identifying changes in the sensory component of peripheral nerves or in the dermatomes of the lower ex- tremities. 10 Overview of Treatment Options The prevalence of lumbar spinal stenosis is increasing with the aging of the population. However, many persons who have degenerative changes consistent with lumbar stenosis are asymptomatic. For those with symptomatic lumbar stenosis, the natural history has been shown to be relatively stable in mild to moderate cases. Johnsson et al 18 studied the progression of symp- toms over 4 years in 32 patients with lumbar stenosis who either refused or were not medically cleared for decompressive surgery. The symp- toms of 70% of the patients were un- changed at follow-up, with half of the remainder better and half worse. Although the physical findings were improved in almost half of the pa- tients, 38% had progression of elec- tromyographic changes. In the shorter term, patients in the Maine Lumbar Spine Study (Part III) who were treated nonoperatively re- mained symptomatically stable but did not show any significant clinical improvement. 19 Appropriate candidates who do not improve with nonoperative in- terventions are likely to realize major benefits from decompressive surgery if they have moderate to severe lumbar stenosis. The litera- ture supporting operative treatment of lumbar stenosis has been shown to have many methodologic flaws with respect to indications for surgery, surgical approaches, and long-term outcome. 20 Recently, however, the results in the Maine Lumbar Spine Study (Part III) pro- spectively demonstrated superior outcomes at 1 year for operative treatment of symptomatic lumbar stenosis compared with continued nonoperative management. The preliminary results in the same co- hort demonstrated that the opera- tively treated patients maintained their superior status compared with nonoperatively treated patients even at 3-year follow-up. 19 Longer prospective follow-up is needed to prove the long-term superiority of operative treatment of lumbar ste- nosis. Nonoperative Management Nonsteroidal anti-inflammatory drugs (NSAIDs) should be part of the initial management of sympto- matic lumbar stenosis (Fig. 2). At lower doses, the analgesic effect reduces musculoskeletal pain; at higher doses, NSAIDs provide an anti-inflammatory effect on nerve root and joint irritation. Unfortu- nately, many elderly patients can- not tolerate the gastrointestinal and renal side effects. Another option is enteric-coated aspirin, which may be as effective, at lower cost and with fewer gastrointestinal side effects. Patients taking any of these medicines should have their hepatic and renal function monitored. For patients with severe radicu- lar complaints, corticosteroids are potent anti-inflammatory agents that may reduce nerve root irrita- tion. However, their benefits must be weighed against the potential side effects, such as osteonecrosis of the femoral head, hypergly- cemia, and gastritis. In addition, steroids may adversely affect the mental status of elderly patients. Many physicians continue to prescribe narcotic analgesics for the treatment of chronic back pain and neurogenic claudication. These medications are analgesic but not anti-inflammatory. They are addic- tive and have depressant effects on mood and energy level. In the ab- sence of an acute fracture or meta- static cancer, they have a limited role in the treatment of patients with lumbar stenosis. Other medications used to treat stenotic symptoms include muscle relaxants, antidepressants, and calci- tonin. Muscle relaxants may pro- vide short-term relief of muscle spasm, but geriatric patients may have adverse reactions. Tricyclic antidepressants are useful in the treatment of chronic numbness and dysesthetic leg pain, and help estab- lish an effective sleep pattern. Salmon calcitonin has been reported as a medical treatment for neuro- genic symptoms of stenosis. Al- though it may be indicated in the management of spinal stenosis re- lated to Paget disease, its efficacy has not been demonstrated in double- blinded studies. 21 Physical therapy is another com- mon nonoperative intervention for symptomatic lumbar stenosis. A modification of the standard low- back-pain exercise program may be used, wherein postural exercises in flexion are combined with pelvic stabilization and aerobic condition- ing. The stabilization exercises are utilized to strengthen the abdomi- nal and lumbodorsal muscle groups and unload the spinal elements. A program of aerobic conditioning can improve overall muscle tone and truncal balance as well as assist in weight loss, which is important in the treatment of lumbar stenosis in obese individuals. Patients with lumbar stenosis are ideally suited to the exercise bicycle. We also en- courage walking, if tolerated by the patient. Aquatic therapy can be beneficial for some patients limited by medical comorbidities. Brace or elastic-corset immobi- lization for low back pain due to spinal stenosis is another available treatment. The support from a Alan S. Hilibrand, MD, and Nahshon Rand, MD Vol 7, No 4, July/August 1999 243 brace may provide short-term relief when back pain is related to seg- mental instability or spondylolisthe- sis, although it can prevent lumbar flexion, which may limit its regular use. An elastic corset may also assist the posterior musculature by providing a counterforce to contrac- tions, reducing strain. Long-term brace wear leads to truncal decondi- tioning, however, and has therefore fallen out of favor. Epidural steroid injection (ESI) is an invasive modality that may be used to treat patients with lumbar stenosis. It involves delivery of a corticosteroid preparation, such as methylprednisolone, around the stenotic cauda equina and nerve roots in order to relieve lower- extremity pain and neurogenic claudication. Numerous articles have been written regarding the efficacy of ESI in the treatment of lumbar radicular syndromes; how- ever, only a few of these studies included patients with degenera- tive lumbar stenosis. Cuckler et al 22 performed a pro- spective, randomized, double-blind study of epidural steroids in pa- tients with lumbar radicular syn- dromes, approximately half of whom had diagnoses of lumbar spinal stenosis. In the patients with spinal stenosis, no statistically sig- nificant difference was seen in symptomatic improvement be- tween steroid and placebo injec- tions at 24-hour and 1-year follow- up. However, a similar study in 1973 by Dilke et al 23 demonstrated a significant (P<0.05) improvement in short-term pain and functional measures. In a study restricted to patients with symptomatic lumbar stenosis, Hoogmartens and Morelle 24 found that 48% of patients treated with ESI demonstrated functional im- provement from their preinjection status approximately 2 years after treatment. However, the patients were evaluated retrospectively and were not compared with a control group. Furthermore, the authors conceded that the improvement rate was close to that of the placebo effect. Nevertheless, they suggested that ESI is a good alternative to sur- gical treatment in older patients with medical comorbidities. At our center, we reserve ESI for patients with mild to moderate stenosis and major medical comor- bidities for whom medical treat- ments and physical therapy have not been efficacious. If the first in- jection provides symptomatic relief, one or two additional injections are prescribed. We use ESI to facilitate a return to more aggressive truncal strengthening and aerobic condi- tioning, which may provide longer- term relief of symptoms. There are, however, potential complications of ESI, including meningitis, para- paresis, arachnoiditis, and epidural hematoma. 25 Other invasive modalities con- tinue to be widely used despite a lack of scientific support. Facet- joint injections may provide tempo- rary relief in select patients with Degenerative Lumbar Stenosis Journal of the American Academy of Orthopaedic Surgeons 244 No symptoms Symptoms Successful Unsuccessful Successful No further workup No further workup Referrals No further workup Advanced imaging studies Severe stenosis Mild to moderate stenosis Minimal stenosis Unsuccessful at 6-12 weeks Patient with lumbar stenosis NSAIDs Physical therapy Weight loss NSAIDs Physical therapy Weight loss Operative intervention Operative intervention Epidural steroids Reevaluate for other causes Fig. 2 Algorithm for nonoperative management of degenerative lumbar stenosis. low back pain, although their effi- cacy in lumbar stenosis has not been shown. 26 Manipulation is another common treatment modali- ty for back pain that has not been proved efficacious in treating patients with degenerative lumbar stenosis. Operative Treatment Not all patients for whom nonopera- tive treatment has failed will be sur- gical candidates. However, when nonoperative intervention cannot restore the patient to a tolerable quality of life, operative treatment can be considered (Fig. 3). A thor- ough workup should identify any associated degenerative pathologic changes, such as spondylolisthesis, segmental instability, or scoliosis, that might require stabilization in addition to decompression. Decom- pressive surgery is clearly indicated in cases of progressive neurologic deficit, cauda equina syndrome, and disabling lower-extremity weakness. In addition, decompressive treat- ment is appropriate for patients with neurogenic claudication and/or pain that significantly affects the patientÕs quality of life. Previous reviews have suggest- ed that surgical treatment of lum- bar stenosis is successful (defined as significant pain relief with a return to activities of daily living) in 80% to 85% of cases. 17,27 How- ever, other authors have found a much lower rate of successful results, perhaps attributable to dif- fering definitions of a Òsuccessful result.Ó In a retrospective review 4 years after lumbar decompression, Katz et al 28 found Òsuccessful out- comesÓ (defined as relief of pain and no reoperation) in only 57% of cases. Factors associated with un- successful outcomes were multiple comorbidities, single-level decom- pressions, and a 5% annual inci- dence of degeneration at levels ad- jacent to the decompression. In another shorter-term follow-up study of patients who underwent surgery for lumbar stenosis, the authors identified predominating low back pain as another factor associated with poor outcomes after surgical treatment of lumbar stenosis. 29 Katz et al 28 have suggested that surgical results deteriorate with time. However, Herno et al 30 ob- served 67% and 69% successful outcomes in their patients at 7 and 13 years after lumbar decompres- sion, respectively. Preoperative planning for lum- bar decompression begins with a review of the clinical findings and evaluation with MR imaging or postmyelographic CT to precisely identify areas of neural element compression by articular processes, capsular tissue, bulging anulus, and ligamentum flavum in the lateral recesses and foramina. Patients will sometimes present with multi- level disease despite well-defined neurologic deficits referable to one or two motor roots (Fig. 4). Several authors have suggested that pro- duction of symptomatic disease may require compression of the cauda equina at more than one level. 11,31,32 As a result, in some centers it is routine to address all levels of moderate or severe steno- sis visualized with postmyelo- graphic CT at the time of surgical decompression. For surgical decompression, the patient is placed in the prone knee- chest position, although the more lordotic position of the Jackson spinal table may be preferred when internal fixation is planned. The lumbar spine is exposed through a midline incision of skin, subcuta- neous tissue, and thoracolumbar fascia. Subperiosteal exposure is carried out laterally to the facet joint capsules, which are preserved, and the lateral extent of the pars interarticularis is identified. Compressive elements are re- moved by undercutting the facet joints and pars interarticularis, tak- ing care to preserve at least 50% of the facet joint surface area and approximately 1 cm of the dorsal Alan S. Hilibrand, MD, and Nahshon Rand, MD Vol 7, No 4, July/August 1999 245 No spondylolisthesis Spondylolisthesis Stable Fusion ± instrumentation Fusion ± instrumentation Fusion ± instrumentation No or minimal scoliosis Significant scoliosis Lumbar decompression only Fusion in situ Surgical candidate Unstable (>3 mm on dynamic films) StableUnstable (grade II or higher or >3 mm on dynamic films) Fig. 3 Algorithm for operative management of degenerative lumbar stenosis. surface of the pars. 33 The effective- ness of the decompression is checked by carefully mobilizing the compressed root with a No. 4 Penfield elevator and probing the neural foramen with an angled dural elevator. Foraminotomy is usually necessary in patients with significant degenerative stenosis. This may require removal of por- tions of the superior articular pro- cess of the caudal vertebra, a portion of the pars interarticularis, the anu- lus fibrosus, or the pedicle of the superior vertebra. This aspect of the decompression may be accom- plished with a Kerrison rongeur, small curettes, or a high-speed burr. When the root can be easily mobi- lized from the pedicle and an angled dural elevator can be easily passed out the foramen around the root, the decompression is adequate. In the absence of instability, posterolateral grafting has not been shown to improve outcome. 34 Degenerative spondylolisthesis is a radiographic finding often associated with spinal stenosis at the L4-5 level, especially in women (Fig. 5). In 1991, Herkowitz and Kurz 35 published a controlled prospective study of patients with lumbar stenosis and degenerative spondylolisthesis who were ran- domized to either decompression alone or decompression with in situ intertransverse autogenous bone grafting. The authors demon- strated significantly better out- comes in patients who underwent decompression and grafting proce- dures (P = 0.0001). In another study, the same group studied patients with steno- sis and degenerative spondylolis- thesis treated by decompression and grafting, prospectively com- paring outcomes with and without internal fixation. 36 Although radio- graphic evidence of arthrodesis was significantly more likely with internal fixation (P = 0.0015), the clinical outcome was not improved. The authors concluded that even in the absence of a solid fusion, the intertransverse bone grafting pro- cedure stiffened the spine suffi- ciently to relieve nerve root irrita- tion. In some patients with spinal stenosis, dynamic lateral radio- graphs may indicate segmental instability, defined as more than 3 mm of motion between vertebrae on dynamic views. 37 We routinely obtain preoperative prone and supine lateral radiographs of all lumbar stenosis patients. If the films document segmental instability, decompression with intertransverse bone grafting is recommended. The use of internal fixation in this group of patients remains controversial; some surgeons include fixation in the physically active patient with at least 5 mm of motion and localized mechanical complaints. Degenerative Lumbar Stenosis Journal of the American Academy of Orthopaedic Surgeons 246 Fig. 4 Images of a patient who presented with chronic neurogenic claudication, bilateral lower-extremity numbness, and significant weakness of the L4 and L5 muscle groups preventing ambulation. A, Lateral radiograph demonstrates anterolisthesis at L4-5 and retrolis- thesis at L3-4. Note obliteration of L3-4 disk space and narrowing of neural foramen (arrow). B, Myelogram shows cutoff of contrast material from the L2-3 level through the L4-5 level. C, Postmyelographic CT at L3-4 level shows absence of contrast material secondary to proximal cutoff, with hypertrophy of the ligamentum flavum, degenerated facets, and congenitally narrow (trefoil-shaped) spinal canal. D, Anteroposterior radiograph obtained 6 weeks postoperatively demonstrates wide decompression from L2 through L5, with complete resection of the right L3 pars region (arrows) due to severe foraminal stenosis. Note early consolidation of autogenous iliac-crest bone graft in intertransverse gutters. A B C D 2 3 4 5 2 3 4 5 2 3 4 5 Patients undergoing revision lum- bar decompression should receive special attention. Frequently, there is persistent lateral recess and/or foraminal stenosis not addressed at the index procedure, which should be widely decompressed at the revi- sion operation. Some patients may have iatrogenic spondylolisthesis, scoliosis, or postlaminectomy insta- bility. This can lead to significant coronal and/or sagittal decompensa- tion, which must be addressed at reoperation. We routinely evaluate candidates for revision lumbar decompression with 3-ft-long stand- ing radiographs. If patients have coronal decompensation or loss of lumbar lordosis, intertransverse bone grafting and internal fixation may be helpful to restore anatomic sagittal coronal alignment. Whether operative treatment of recurrent lumbar stenosis can be as successful as primary procedures continues to be controversial. Herno et al 38 found that revision procedures were as successful as primary procedures when patients presented at least 18 months after the index procedure with few comorbidities. Overall, however, outcomes were significantly worse for revision procedures than for pri- mary procedures in the study pop- ulation (P<0.0017). Stewart and Sachs 39 reported that at an average follow-up of 4 years, 72% of their patients who had undergone repeat decompression were able to return to their preinjury work status, 83% of those treated with grafting went on to a solid arthrodesis, and none of the 39 patients required another procedure. In general, better out- comes are associated with a pain- free interval after the initial proce- dure, and worse outcomes are asso- ciated with periradicular fibrosis. Complications Although operative treatment can improve the quality of life of patients with lumbar stenosis, it can also be associated with major complications, which should be discussed with the patient. Postop- erative complications may include epidural hematoma, instability fol- lowing wide decompression in the absence of bone grafting, nonunion of the graft with failure of any con- comitant instrumentation, and the need for future surgery due to the development of new disease at adjacent levels. Any lengthy procedure in an elderly patient under general anes- thesia poses risks to the cardiovascu- lar and respiratory systems. These procedures are often associated with major blood loss, which may require transfusion and invasive cardiac monitoring. Despite antibiotic prophylaxis and strict observation of sterile technique, infection and sepsis re- main a risk, especially when inter- nal fixation is applied. A recent historical cohort study of patients undergoing lumbar decompression and arthrodesis, with or without internal fixation, revealed an infec- tion rate of 2% to 3% for degenera- tive spondylolisthesis. 40 In the same study, 40 the incidence of nerve root injury from placement of bone screws in the pedicle for degenerative spondylolisthesis was 0.4%. Nerve root injury can occur as a result of manipulation of a severely Alan S. Hilibrand, MD, and Nahshon Rand, MD Vol 7, No 4, July/August 1999 247 Fig. 5 Images of a patient who presented with lower-back and bilateral lower-extremity numbness with ambulation. A, Lateral radio- graph shows grade I degenerative spondylolisthesis at L4-5. B, Myelogram shows truncation of the column of contrast material and bilat- eral fifth-root cutoff (arrows) at the level of the spondylolisthesis. C, Postmyelographic CT scan demonstrates severe spinal stenosis at the level of the spondylolisthesis. D, Anteroposterior radiograph obtained 6 weeks postoperatively demonstrates L4 decompressive laminec- tomy and intertransverse bone grafting (arrows). The patient was asymptomatic. A B C D 4 5 4 4-5 5 4 5 compressed nerve root or the use of internal fixation. Although there are no large clinical studies that specifi- cally assess the neurologic complica- tions of decompressive lumbar surgery, a recent review by Wang et al 41 of the data on 641 patients revealed an incidence of dural tears of 13.7%; however, half of the opera- tions in that series were revisions. Summary Degenerative lumbar stenosis is a common cause of back and leg pain in older patients. It usually results from degeneration of the lumbar motion segments, which causes insidious compression of the neural elements. Only a small proportion of the elderly population will have symptoms that merit referral to an orthopaedic surgeon. Most patients will obtain symptomatic improve- ment with medication and physical therapy and not require operative intervention. However, it has not been shown that any nonoperative treatment can alter the relatively stable natural history of the disease. The findings from the Maine Lumbar Spine Study 19 suggest that surgical treatment can improve on the natural history of lumbar steno- sis at short- to intermediate-term follow-up. If nonoperative treat- ment has failed and operative man- agement is elected, it is crucial that all pathologic changes be addressed at surgery. An intertransverse grafting (fusion) procedure should be included if there is any preoper- ative or intraoperative evidence of instability. Although the applica- tion of internal fixation increases the likelihood of successful ar- throdesis, further study is needed to determine which patients will benefit. Questions regarding the long-term clinical outcome of surgi- cal treatment of lumbar spinal stenosis require prospective studies for clarification. Degenerative Lumbar Stenosis Journal of the American Academy of Orthopaedic Surgeons 248 References 1.Roberson GH, Llewellyn HJ, Taveras JM: The narrow lumbar spinal canal syndrome. Radiology1973;107:89-97. 2.De Villiers PD, Booysen EL: Fibrous spinal stenosis: A report on 850 myelo- grams with a water-soluble contrast medium. Clin Orthop1976;115:140-144. 3.Rosenberg NJ: Degenerative spondy- lolisthesis: Predisposing factors. J Bone Joint Surg Am1975;57:467-474. 4.Bolender NF, Schšnstršm NS, Speng- ler DM: Role of computed tomogra- phy and myelography in the diagnosis of central spinal stenosis. J Bone Joint Surg Am1985;67:240-246. 5.Arnoldi CC, Brodsky AE, Cauchoix J, et al: Lumbar spinal stenosis and nerve root entrapment syndromes: Definition and classification. Clin Orthop1976;115:4-5. 6.Wiltse LL, Guyer RD, Spencer CW, Glenn WV, Porter IS: Alar transverse process impingement of the L5 spinal nerve: The far-out syndrome. Spine 1984;9:31-41. 7.Kirkaldy-Willis WH, Wedge JH, Yong- Hing K, Reilly J: Pathology and patho- genesis of lumbar spondylosis and stenosis. Spine1978;3:319-328. 8.Zinreich SJ, Heithoff KB, Herzog RJ: Computed tomography of the spine, in Frymoyer JW, Ducker TB, Hadler NM, Kostuik JP, Weinstein JN, Whitecloud TS III (eds): The Adult Spine: Principles and Practice, 2nd ed. Philadelphia: Lippincott-Raven, 1997, vol 1, pp 467-522. 9.Delamarter RB, Bohlman HH, Dodge LD, Biro C: Experimental lumbar spinal stenosis: Analysis of the cortical evoked potentials, microvasculature, and histopathology. J Bone Joint Surg Am1990;72:110-120. 10.Garfin SR, Rydevik BL, Lipson SJ: Spinal stenosis: Pathophysiology, in Rothman RH, Simeone FA (eds): The Spine, 3rd ed. Philadelphia: WB Saun- ders, 1992, vol 1, pp 791-826. 11.Parke WW, Watanabe R: The intrinsic vasculature of the lumbosacral spinal nerve roots. Spine1985;10:508-515. 12.Amundsen T, Weber H, LilleŒs F, Nordal HJ, Abdelnoor M, Magnaes B: Lumbar spinal stenosis: Clinical and radiologic features. Spine1995;20: 1178-1186. 13.Herno A, Airaksinen O, Saari T, Miettinen H: The predictive value of preoperative myelography in lumbar spinal stenosis. Spine1994;19:1335- 1338. 14.Boden SD, Davis DO, Dina TS, Patronas NJ, Wiesel SW: Abnormal magnetic-resonance scans of the lum- bar spine in asymptomatic subjects: A prospective investigation. J Bone Joint Surg Am1990;72:403-408. 15.Modic MT, Masaryk T, Boumphrey F, Goormastic M, Bell G: Lumbar herni- ated disk disease and canal stenosis: Prospective evaluation by surface coil MR, CT, and myelography. AJR Am J Roentgenol1986;147:757-765. 16.Riew KD, Hilibrand AS, Bridwell KH, et al: MR versus CT-myelography: Surgical decision-making for lumbar stenosis, in Proceedings of the 13th Annual Meeting of the North American Spine Society. Rosemont, Ill: North American Spine Society, 1998, pp 117-118. 17.Spengler DM: Degenerative stenosis of the lumbar spine. J Bone Joint Surg Am1987;69:305-308. 18.Johnsson KE, RosŽn I, UdŽn A: The natural course of lumbar spinal steno- sis. Clin Orthop1992;279:82-86. 19.Atlas SJ, Deyo RA, Keller RB, et al: The Maine Lumbar Spine Study, Part III: 1- Year outcomes of surgical and nonsur- gical management of lumbar spinal stenosis. Spine1996;21:1787-1795. 20.Turner JA, Ersek M, Herron L, Deyo R: Surgery for lumbar spinal stenosis: Attempted meta-analysis of the litera- ture. Spine1992;17:1-8. 21.Porter RW, Miller CG: Neurogenic claudication and root claudication treated with calcitonin: A double-blind trial. Spine1988;13:1061-1064. 22.Cuckler JM, Bernini PA, Wiesel SW, Booth RE Jr, Rothman RH, Pickens GT: The use of epidural steroids in the treatment of lumbar radicular pain: A prospective, randomized, double- blind study. J Bone Joint Surg Am 1985;67:63-66. 23.Dilke TFW, Burry HC, Grahame R: Extradural corticosteroid injection in the management of lumbar nerve root compression. BMJ1973;2:635-637. 24.Hoogmartens M, Morelle P: Epidural