INTRODUCTION — Lumbar spinal stenosis (LSS) refers to an anatomic condition that includes narrowing of the intraspinal (central) canal, lateral recess, and/or neural foramen. Spondylosis, or degenerative arthritis affecting the spine, is the most common cause of LSS and typically affects individuals over the age of 60 years [1].
LSS is a cause of disability in the aging population [2,3]. Treatment approaches are often not straightforward and require careful consideration of potential risks and benefits.
This topic will discuss the prognosis and treatment of LSS, focusing on lumbar spondylosis. Other topics discuss more general issues related to the evaluation and treatment of low back pain, and also the pathophysiology, clinical features, and diagnosis of LSS. (See "Evaluation of low back pain in adults" and "Subacute and chronic low back pain: Nonsurgical interventional treatment" and "Subacute and chronic low back pain: Nonpharmacologic and pharmacologic treatment" and "Lumbar spinal stenosis: Pathophysiology, clinical features, and diagnosis".)
PROGNOSIS — The natural history of LSS due to degenerative spondylosis is relatively benign. Among 32 patients with symptoms and neuroimaging consistent with LSS who were followed for a mean 49 months without surgical intervention, symptoms remained unchanged in 70 percent, improved in 15 percent, and worsened in 15 percent [4]. However, LSS causes discomfort, often limiting activities of daily living, and can lead to progressive disability. While pain and disability both contribute to decreased quality of life, one survey suggested that patients prioritize reduced pain with walking over improved walking capacity [5].
Radiographic studies suggest that over time, adjacent spinal segments also become affected in almost half of patients [6,7]. Some data indicate that this phenomenon may be a more frequent occurrence in those who undergo surgical fusion and contribute to re-operation rates.
NONSURGICAL TREATMENT — In general, a trial of conservative, nonsurgical treatment precedes surgical treatment. The exception is the rare patient with progressive neurologic deficits, especially the cauda equina syndrome, which indicates more urgent surgical decompression.
Nonsurgical treatments used for LSS have included physical therapy, analgesic and antiinflammatory medications, and epidural steroid injections. High-quality clinical trials supporting the utility of these treatments are lacking [8-10]. Other treatments for low back pain not specific to LSS are discussed separately. (See "Treatment of acute low back pain" and "Subacute and chronic low back pain: Nonpharmacologic and pharmacologic treatment" and "Subacute and chronic low back pain: Nonsurgical interventional treatment", section on 'Glucocorticoid and other injections'.)
Physical therapy — Physical therapy is the mainstay of conservative management, but evidence-based support in the literature is lacking. Regimens are not standardized. Stretching, strengthening, and aerobic fitness are usually recommended [1,11-13]. Goals of therapy include increased muscular stabilization and correction of posture. Techniques to increase lumbar flexion and reduce lumbar lordosis include strengthening of abdominal muscles. Abdominal corsets or braces can also help reduce lordosis and improve exercise tolerance [14]; however, their use is controversial, as overuse can lead to increased weakness of abdominal and other core muscles. (See "Subacute and chronic low back pain: Nonpharmacologic and pharmacologic treatment".)
One randomized study compared two six-week physical therapy regimens in 58 patients with LSS: manual physical therapy and treadmill walking versus lumbar flexion exercises, subtherapeutic ultrasound, and treadmill walking [15]. At six weeks, a greater proportion of patients in the first treatment group reported improvement (79 versus 41 percent); the difference was not statistically significant at one year (62 versus 41 percent). The physical therapy regimens were not well standardized, making specific recommendations difficult. However, this study provides some support for physical therapy programs in managing patients with LSS.
Limited evidence suggests that postoperative physical therapy improves recovery following surgery [16]. A meta-analysis of three randomized studies concluded that active rehabilitation is effective in improving functional status and low back pain at 6 and 12 months postoperative [17]. The clinical magnitude of the benefit is uncertain.
For overweight individuals, exercise for weight loss is an important goal to reduce lumbar lordosis and minimize axial strain on the spinal column. Bicycling and aquatic exercise programs may be good choices for patients with LSS. Dietary modification is also recommended as part of a weight loss program. (See "Obesity in adults: Overview of management".)
Pharmacotherapy — Drug therapy for low back pain includes aspirin, nonsteroidal antiinflammatory drugs (NSAIDs), and opiates. No studies have clearly identified the superiority of one class of drugs over another, but the complication profile of each plays a role in the ultimate choice for the individual patient [18]. Of particular concern in the older age group is the associated cardiovascular risk associated with certain NSAIDs. (See "Treatment of acute low back pain", section on 'Pharmacotherapy'.)
The efficacy of these analgesic medications in patients with neurogenic claudication is not clear. A randomized crossover trial found that pregabalin was not more effective than active placebo (diphenhydramine) in 29 patients with neurogenic claudication [19].
Epidural injections — The available evidence does not support the use of epidural injections of corticosteroids and/or anesthetics in LSS. However, the evidence is limited; most studies of this approach include a heterogeneous patient population with back pain, radicular pain, and/or neurogenic claudication [20-26].
In a randomized trial of 400 patients with LSS and moderate to severe leg pain and disability, those patients who received one or two epidural injections of glucocorticoids plus lidocaine had similar pain and disability outcomes at six weeks as those who received injections of lidocaine alone [27]. Adverse events (such as pain, headache, fever, and infection) were nonsignificantly higher in the glucocorticoid group (21.5 versus 15.5 percent). Since this study did not include a sham control group, it does not rule out a potential benefit from the injection of the local anesthetic. Previously reported uncontrolled observational studies and small trials had suggested that injections of local anesthetic agents with or without corticosteroids were associated with reduced pain in the short term [23,24].
Potential indications for the use of epidural injections for the treatment of low back pain in other settings as well as a summary of associated adverse events are presented separately. (See "Subacute and chronic low back pain: Nonsurgical interventional treatment", section on 'Glucocorticoid and other injections'.)
SURGICAL TREATMENT
Patient selection — Deciding on surgical intervention in a patient with LSS requires a careful consideration of potential risks and benefits. There is limited evidence on which to base treatment recommendations; randomized trials and systematic reviews do not find definitive evidence of a substantial benefit for surgical over nonsurgical treatments [10,28,29]. Patient preferences should weigh heavily in decision-making [30]. Patients should know that any benefits of surgery decline over time and that repeat operations (for same or adjacent level disease) are performed in 15 to 25 percent [31-37].
Several studies have identified varying predictors for outcome of surgical treatment of LSS [33,38-45]. Predictors of a good clinical outcome after surgery are quite variable between studies. A systematic review identified the following predictors in one or more high-quality studies [40]. Negative predictors were:
●Depression
●Concomitant disorder influencing walking capacity
●Cardiovascular comorbidity
●Scoliosis
Positive predictors were:
●Male gender
●Younger age
●Better walking ability
●Better self-rated health
●Less comorbidity
●More pronounced canal stenosis
However, the degree of association between any one of these predictors and surgical outcome varies among studies, and any one variable should not be overemphasized in clinical decision-making [46].
Smoking has also been identified as being associated with negative surgical outcomes, including nonunion, delayed healing, and increased pseudarthrosis [47-50]. Animal studies have shown that nicotine administration reduces bone fusion rates, presumably by impairing revascularization [51,52]. Smoking cessation appears to be beneficial in improving outcomes after surgery. In one review of 357 patients undergoing spinal fusion, the nonunion rate of those who stopped smoking for six months or longer after surgery was similar to nonsmokers and significantly less than those who continued to smoke (17 versus 14 versus 26 percent) [53].
Obese patients have a higher risk of intra- and postoperative complications [54]. However, data suggest that they experience a benefit from surgical treatment that may be comparable to nonobese patients, and they also are more likely to have poorer outcomes than nonobese patients with nonsurgical management [55-57].
Some data suggest that a magnetic resonance imaging (MRI) finding of a positive sedimentation sign (ie, failure of the lumbosacral nerve roots to settle posteriorly with gravity when imaged in the supine position) indicates a patient who is less likely to improve with conservative treatment [58,59]. While subgroup analysis of a randomized study suggests that such patients may have greater surgical treatment efficacy, this result requires independent validation before it can be recommended for patient selection for surgery [60].
Timing — Surgical treatment of LSS is usually elective, in that the benefit of surgery in most cases is to alleviate symptoms and improve function, rather than to prevent neurologic impairment. Surgical approaches are generally considered in patients with disabling pain despite nonsurgical treatments. In the absence of neurologic deficits, delayed surgery appears to produce benefits that are similar to surgery when conservative therapy is selected as the initial treatment choice [31]. Studies comparing surgical and nonsurgical therapy enroll patients with at least three months of symptoms unresponsive to conservative therapy, and many patients have had prolonged symptoms. However, patients on prolonged surgical wait lists (median approximately one year) were more likely to have delayed recovery [61]. In cohort studies of individuals with LSS who initially chose nonsurgical therapy, approximately 30 percent subsequently requested surgical treatment [31,32].
Urgent surgery is required when neurologic deficits are rapidly progressive and/or when bladder dysfunction emerges. While this may occur with cauda equina syndrome or conus medullaris syndrome in the setting of trauma or an intraspinal canal tumor, it is very rarely a manifestation of clinical decompensation in LSS due to spondylosis. In the former cases, the degree of neurologic recovery is related to the speed at which the compressive lesion is corrected and the severity of baseline deficits prior to intervention. (See "Treatment and prognosis of neoplastic epidural spinal cord compression".)
Surgical approach — Surgical approaches for degenerative LSS include single- or multilevel decompressive laminectomy with or without lumbar fusion. Lumbar fusion (with or without instrumentation) is generally reserved for patients with spondylolisthesis, a condition in which one vertebral body translates anteriorly or posteriorly with respect to an adjacent vertebral body.
Spinal stenosis without spondylolisthesis — In the absence of spondylolisthesis (or other complicated spine pathology) decompressive laminectomy without fusion is preferred over fusion, as it is associated with similar pain and function improvement and a lower complication rate [62]. A review of Medicare claims database demonstrated that as the frequency of complex fusion procedures for LSS increased over the 2002 to 2007 time period, so did major complications, 30-day mortality, and resource use [63].
Laminectomy — Laminectomy, the most frequently used surgical procedure in this setting, involves removal of part or all of one or both lamina of the vertebra at the involved level (figure 1). In some cases the spinous process is removed as well; overlying connective tissues, ligaments, and muscle may be transected to gain access to the vertebra.
Interpretation of randomized clinical trials is limited by the absence of blinding, high crossover rates, and small sample size. Though trial results conflict somewhat, in the aggregate they suggest a modest benefit of surgery that diminishes over time:
●In a multicenter study in the United States, 289 patients with symptomatic LSS but no spondylolisthesis were randomized to decompressive surgery versus nonsurgical care and followed for two years [64]. Two-thirds of those randomized to surgical treatment and 43 percent of those assigned to nonsurgical care underwent surgery during the study period. In this study, the intention-to-treat analysis did find a benefit for surgery for a measure of bodily pain, but not for disability or functional outcome. An as-treated analysis, which also included an observational cohort of an additional 365 patients, demonstrated a benefit for surgery on all outcome measures at three months and two years, which was maintained at four years [65]. Surgical benefit persisted through four years, but by eight years, outcomes of pain and functional status were similar among patients treated surgically and nonsurgically; 19 percent of those treated surgically had had a second operation [66].
●Another multicenter study randomly assigned 169 patients to decompressive surgery versus physical therapy [67]. Fifty-seven percent of those assigned to physical therapy received surgery. At two years, physical function was similar between the two patient groups (intention-to-treat analysis). As-treated analysis found a nonsignificantly higher proportion of improved physical function among patients assigned to surgery who received surgery compared with those assigned to physical therapy who crossed over to surgery as well as those assigned to physical therapy who did not receive surgery (61 versus 55 versus 52 percent).
●A smaller randomized trial of 94 patients also reported a benefit from surgical over conservative treatment that was smaller and diminished over time, with no significant benefit in disability at two years [68].
A previously published systematic review of randomized surgical studies of degenerative lumbar spondylosis found some limited evidence supporting the effectiveness of surgical management [28]. Observational cohort studies in general favored surgery, but outcomes varied among studies and centers, and benefits were also noted to decline over time [31-34,38,69,70].
Intraspinous spacer implantation — A potentially less invasive treatment option involves implanting a device between the spinous processes at one or two vertebral levels, relieving compression [71]. This procedure is said to be appropriate for those patients with spinal stenosis without spondylolisthesis who have intermittent claudication symptoms that are exacerbated in extension and relieved in flexion [71,72]. A randomized, multicenter study in 191 patients compared the implantation of the X STOP implant, a titanium alloy device, with nonoperative treatment [73]. At six months, symptoms were relieved in 52 percent of treated patients, compared with 9 percent of controls. Benefit was maintained at two and four years of follow-up and was associated with reduced disability and improved quality of life [74-76].
Subsequent uncontrolled observations have found that implantation of the X STOP device has been efficacious in many patients, if not in as large a proportion as was found in the clinical trial [72,77,78]. While radiologic improvement in spinal canal and neuroforaminal narrowing can be measured after surgery, these changes are not correlated with clinical benefit and are not maintained over time in most patients [79-81]. These procedures appear to be associated with higher rates of subsequent surgery than patients initially treated with laminectomy. Adverse effects also appear to be more commonly reported in general clinical experience; these include discitis/osteomyelitis, device dislocations, spinous process fractures, recurrent disc herniation, hematoma, cerebrospinal fluid fistula, and foot drop [82-85].
It is unclear how this newer procedure compares with the standard surgical procedure, decompressive laminectomy, in terms of effectiveness, side effects, recovery time, and long-term outcomes. This treatment does not appear to be helpful in patients who have spondylolisthesis [85,86].
Minimally invasive decompression — There is long-standing interest in the development of less invasive decompression procedures, such as percutaneous lumbar decompression and/or minimally invasive lumbar decompression, which appear in observational studies to have lower complication rates than traditional surgical techniques [87-94]. It is unclear if these newer procedures offer benefit in terms of improved symptoms and function or fewer complications in routine practice compared with standard decompression with laminectomy.
In a multicenter randomized trial in Norway, 885 patients with LSS had similar outcomes at one year whether treated with microdecompression or laminectomy [95]. Complication rates were also similar between groups after propensity matching, but shorter hospital stays (mean difference 1.5 days) were observed in those who received microdecompression.
Spinal stenosis with spondylolisthesis — Spondylolisthesis, in which one vertebral body translates anteriorly or posteriorly with respect to an adjacent vertebral body, can exacerbate spinal canal narrowing and may call for a different surgical approach [96]. Most surgeons include fusion when operating on patients with spondylolisthesis. The addition of instrumentation (insertion of rigid implants such as a rod or screw) in fusion improves the rate of successful fusion but does not clearly impact clinical outcomes, and is associated with a higher complication rate [96-98]. Administrative data of surgical outcomes from the Nationwide Inpatient Sample between 1993 and 2002 reported a 13 percent complication rate that included hematoma, as well as pulmonary, renal, and cardiac complications [99]. Older patients and those with medical comorbidities are at greatest risk. Outcomes vary significantly among studies and centers, suggesting that local surgical expertise and other procedural factors may influence outcome [70].
A study of 304 patients with degenerative lumbar spondylolisthesis, and either neurogenic claudication or radiculopathy, randomly assigned patients to nonsurgical treatment or to decompressive laminectomy and fusion [100]. Within two years of follow-up, 64 percent of those assigned to surgery and 49 percent of those assigned to conservative treatment underwent operative therapy. Intention-to-treat analysis found no difference in outcome between treatment groups, while an as-treated (nonrandomized) analysis found a benefit from surgery in functional and disability assessments at both two and four years [101].
Randomized studies suggest that the benefit of adding fusion to decompression is of uncertain benefit, particularly in the absence of spinal instability [102]:
●One study of 50 patients with single-level spondylolisthesis compared lumbar decompression surgery with versus without spinal fusion [103]. After a mean follow-up of three years, patients with spinal fusion had significantly less back and lower limb pain and had better physical function. In this study, unsuccessful radiographic fusion did not correlate with poor outcome.
●In a randomized study in 66 patients with stenosis and grade 1 spondylolisthesis (no spinal instability), the addition of fusion was associated with a small benefit on the SF-36 physical-component summary score, but no change in the Oswestry Disability Index score two years after surgery [104]. More blood loss and longer hospital stays occurred in patients undergoing fusion.
●A larger randomized trial compared decompression surgery with fusion to decompression alone in 247 patients with spinal stenosis; 135 had spondylolisthesis [105]. No difference in outcome on the Oswestry Disability Index were noted between groups at five years; outcomes were similar in those with and without spondylolisthesis.
While efficacy is uncertain, the procedure is expensive. In the first study described above, considering the expected benefit over two years of follow-up, surgery was estimated to cost $115,600 per quality-adjusted life-year (QALY), an amount that is generally not considered highly cost effective [106].
Intraoperative implantation of bone morphogenetic proteins during fusion surgery has been shown to improve fusion rates, but has not been shown to improve clinical outcomes and may increase complications [107-112].
Surgical complications — Rates of surgical complications are associated with the complexity of the surgical procedure. Laminectomy without fusion has lower complication rates than laminectomy with fusion; multilevel fusion has a particularly high rate of complications. Surgical complication rates include mortality in 0.5 to 2.3 percent [69,113]. Other serious complications (eg, infection and deep venous thrombosis) occur in approximately 12 percent of patients. The patient's age and number of comorbidities impact significantly on surgical risk, as does the number of procedures performed by the surgeon and hospital, with more procedures associated with lower mortality and complication rates [69,113-116]. Features of the surgical procedure, including the use of fusion and/or instrumentation and the number of levels operated on, may also impact complication rates [63,96,117-120]. Reoperation for recurrent stenosis or for adjacent disease is not rare, but rates vary widely [102,121].
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topics (see "Patient education: Spinal stenosis (The Basics)")
SUMMARY AND RECOMMENDATIONS — Lumbar spinal stenosis (LSS) is most often due to degenerative spondylosis, which typically affects individuals over the age of 60 years.
●The prognosis of LSS is benign in that neurologic disability is rare and most patients remain stable over several years of follow-up. However, for some patients, the symptoms become disabling and restrict activity. (See 'Prognosis' above.)
●For patients with LSS who do not have fixed or progressive neurologic deficits, we suggest conservative treatment (Grade 2C). Physical therapy and/or oral pain medication are often used, although their efficacy has not been rigorously evaluated. (See 'Nonsurgical treatment' above.)
The available evidence does not support the use of epidural injections of corticosteroids. (See 'Epidural injections' above.)
●We suggest consultation for surgical therapy for patients who do not have an adequate clinical response to conservative therapy and who are functionally disabled by their symptoms, and for patients who have a progressive neurologic deficit (Grade 2C). A shared decision-making process with information about treatment options and assessing patient preferences is recommended. The risks and benefits of treatment options should be reviewed, including benefits of surgery that decline over time and the possibility of repeat operations (for same or adjacent level disease) in 15 to 25 percent.
●In the absence of unstable spondylolisthesis (or other complicated spine pathology), decompressive laminectomy without fusion is preferred over more complicated surgical techniques involving fusion with or without instrumentation. (See 'Surgical treatment' above.)
●We recommend urgent surgical consultation for rare patients with a rapidly progressive cauda equina or conus medullaris syndrome or newly emerging bladder dysfunction (Grade 1B). (See 'Surgical treatment' above.)