Diagnosis and Management of Lumbar Spinal Stenosis

EVERY YEAR AN ESTIMATED 90 OF 100 000 PERSONS OLDER than 60 years undergo lumbar fusion surgery. The diagnosis of lumbar spinal stenosis, defined as “a clinical syndrome of buttock or lower extremity pain, which may occur with or without back pain, associated with diminished space available for the neural and vascular elements in the lumbar spine,” is an important driver of the exponential increase in this procedure. Most surgeons rely on imaging for diagnosing spinal stenosis and for determining the need for surgery. However, the assumption that radiological measures confirm the diagnosis of the clinical syndrome of stenosis has been questioned. Without a clear diagnostic standard, a management strategy that minimizes the potential of harm from an incorrect diagnosis needs to be developed. Evidence-basedguidelinesoftenuseacircuitouslogic insupporting imaging as the key diagnostic test for stenosis. For instance, the North American Spine Society (NASS) guidelines conclude that imaging is the key noninvasive test for diagnosis. The guidelines do not specify radiological criteria for stenosis and yet exclude from review any studies that do not use imaging as an inclusion criterion. Research on spinal stenosis almost always uses imaging to establish the diagnosis. However, the inclusion criterion is typically based on an opinion ofaradiologistorsurgeonwhoreviewedthe imagesrather than some stated definition. The interrater reliability of the clinical impression is seldomestablished,maskingclinicaldata is rare, andcontrolpopulationsare seldomincluded.Anatomical cutoffs relating to anterior-posterior spinal canal diameter or thecal sac diameter have been proposed as diagnostic criteria for spinal stenosis; however, these measurements are not necessarily tied to clinical evidence or statistical norms. Several studies have examined this issue. In one study, radiologists masked to clinical status diagnosed stenosis in 65% of magnetic resonance imaging (MRI) results performed on 32 asymptomatic volunteers. In another report, the relationship between clinical presentation of stenosis and measurement of anterior-posterior spinal canal diameter was statistically significant but MRI had no discriminant value in separating patients with clinical stenosis from asymptomatic volunteers. Among patients older than 60 years in the Framingham cohort, 47% of lumbar computed tomography studies met accepted criteria for moderate or severe stenosis. Another test used for spinal disorders is electromyography (EMG), although the NASS guidelines did not support the use of EMG. In one study, a newer quantitative EMG protocol called paraspinal mapping significantly discriminated persons with clinical stenosis from those with mechanical back pain and asymptomatic volunteers. In another study, 15 of 16 persons (94%) with “stenosis” on MRI but no symptoms had normal paraspinal mapping EMG, whereas 26 of 28 persons (93%) with both symptoms and stenosis on MRI had abnormal paraspinal mapping EMG. It appears that the clinical syndrome of stenosis may be related to pathophysiological alterations demonstrated on EMG—in contrast to MRI metrics that do not appear to distinguish the relevant lesion. Lack of a clear relationship between imaging findings and the clinical presentation of spinal stenosis can be explained. It is possible that severe radiological stenosis relates to clinical stenosis. Also, imaging studies are conducted with patients supine, whereas symptoms of stenosis generally are precipitated by standing or walking. In these upright positions, the spinal canal can be made smaller by segmental instability, compression by soft tissue structures (facet joint cysts, ligamentum flavum, intervertebral disk, and posterior epidural fat), or venous congestion. In evaluating older patients with suspected spinal disorders, 3 key steps should be taken: Find and treat what is not stenosis, define and treat the effects of stenosis, and treat presumed stenosis without a definitive diagnosis. Failing all of these steps, a positive diagnosis is an important consideration before surgery. First, find and treat what is not stenosis. The presence of leg pain does not necessarily mean that the clinically most relevant symptoms are the result of nerve root compression. Because mechanical back pain is ubiquitous, persons who have leg disorders ranging from diabetic neuropathy to peripheral vascular disease to polyarthritis may be misdiagnosed as having spinal stenosis. The back pain component of clinical “stenosis” may result from mechanical pain generators such as the sacroiliac joints, facet joints, hip joints,