Secondary Logo

Journal Logo

SECTION I SYMPOSIUM: New Developments in Lumbar Spinal Stenosis

Lumbar Stenosis

A Clinical Review

Arbit, Ehud, MD; Pannullo, Susan, MD

Author Information
Clinical Orthopaedics and Related Research®: March 2001 - Volume 384 - Issue - p 137-143
  • Free


List of Abbreviations Used: BMP bone morphogenetic protein, TGF-β transforming growth factor-bet

Lumbar stenosis is defined as the reduction in the diameter of the spinal canal, lateral nerve canals, or neural foramina. The stenosis may occur as a part of a generalized disease process and involve multiple areas of the canal and multiple levels or, conversely, may be localized or segmental. The reduction in the diameter of the spinal canal or neural outlets may be attributable to bone hypertrophy, ligamentous hypertrophy, disc protrusion, spondylolisthesis or any combination of these elements. It is important to distinguish between spinal stenosis found incidentally on imaging studies and symptomatic spinal stenosis, because there is a poor correlation between the degree of stenosis and severity of symptoms.

As reported by Wiltse, 31 the first decompressive laminectomy performed for cauda equina syndrome was done by Lane in 1893. In 1900, Sachs and Fraenkel, 24 and in 1911, Baily and Casamajor 2 described the changes in the lumbar spine producing spinal stenosis and nerve root compression including thick laminae, hypertrophied articular facets, and thickened ligamentum flavum. Subsequent reports by Elsberg 8 and by Kennedy et al 14 gave additional credence to this clinical entity. Schlesinger and Taveras 25 described the significance of lumbar stenosis in patients with herniated discs and an atypical picture of multiple nerve root or cauda equina compression. These authors were the first to emphasize that the dimension of the spinal canal was more important than the size of the disc protrusion in the production of symptoms.

Verbiest 27,28 reported the results of his methodical analysis of spinal stenosis. He defined the pathomorphologic changes that take place, specifically the encroachment of the canal by hypertrophied articular processes, and called attention to the characteristic clinical manifestations of the condition including neurogenic claudication.

The progressive pathologic changes of lumbar spinal stenosis that occur in the three joint complex of the disc anteriorly and the zygoapophyseal joints posteriorly and the natural history of the condition were described previously. 15,16,32

Spinal stenosis may be classified by either its etiology or location. Arnoldi et al, 1 who divided lumbar stenosis into two major groups, set forth the most widely used classification: Congenital or developmental stenosis (the terms often are used interchangeably), and acquired stenosis. Congenital stenosis is divided additionally into idiopathic and achondroplastic etiologies. Acquired stenosis is subclassified into degenerative, combined congenital and degenerative, spondylotic and spondylolisthetic, iatrogenic posttraumatic, and metabolic (Table 1).

Classification of Spinal Stenosis

Anatomic classification refers to central canal stenosis, lateral recess stenosis (proximal outlet), or neural foraminal stenosis.

The classification of lumbar stenosis is important because of the implications of the underlying etiology of the condition and when forming a therapeutic strategy, specifically directing surgical approaches. 22


The basic mechanism behind spinal stenosis is remodeling and overgrowth of bone with osteophyte formation. Bone overgrowth either is initiated or accelerated by the degenerative process that affects the three joint complex comprised of the two zygoapophyseal joints and the adjoining disc. 32 The three parts of the joint complex are related intimately and damage that afflicts one part ultimately affects the other two parts. Most often, the degenerative process starts in the disc and affects the articular processes secondarily. Any loss of tissue such as from articular cartilage chaffing, synovitis, or loss of disc height results in relative ligamentous laxity and accelerated joint degeneration. Remodeling of the bone is either a reaction to the excessive joint motion or a physiologic attempt for local arthrodesis. The end result, regardless of the cause, is decreased segmental mobility. This loss of mobility in one segment creates abnormal forces and stresses on adjacent spinal segments, which then degenerate at an accelerated pace. Endochondral ossification can occur remotely from where adjoining periosteum chaffs, (vertebral bodies), such as in the region of the annulus fibrosis after chondrocytic change in the fibrocytes in this area.

The link between the degenerative process and mobility is thought to be the cause for the two lower motion segments (L3-L4, L4-L5) to be most commonly affected by degenerative stenosis. The two lower motion segments are in a region of transit from the rigid sacrum to the mobile lumbar spine. Moreover, their posterior joints have a lesser sagittal orientation that afford more rotation and are more vulnerable to rotatory strains. In many individuals the L5-S1 segment is protected from injury and degeneration by its position below the intercrestal line and by the relative large L5 transverse processes with their strong ligamentous attachments to the iliac crest.

The putative substance for bone overgrowth by chondroid metaplasia and in turn osteophyte formation remains elusive. Bone morphogenetic proteins, cytokines of the TGF-β superfamily have been implicated in the process of disc degeneration. 11,26 Bone morphogenetic proteins are multipotent proteins that regulate the growth, differentiation, and programmed death (apoptosis) of various cell types and are found in abundance in cartilage and bone. Bone morphogenetic proteins and their receptors are greatly expressed during maturation of the intervertebral disc and seem to be related to chondrogenesis within the disc. With progression of degeneration of the disc, BMP and its receptors migrate from the hyaline cartilage of the vertebral endplate to fibrous cells within the annulus and to the calcified cartilage at the site of the enthesis and thus may be related to the formation of osteophytes. It also is suggested that the BMPs, by mediating an effect on cellular apoptosis, contribute to the degenerative process because it has been clearly shown that apoptosis plays a pivotal role in disc degeneration. 9

Pathologic Anatomy

Knowledge of the pathologic anatomy as related to spinal stenosis is important for correlating clinical signs and symptoms with imaging studies and treatment planning.

Central stenosis is caused by hypertrophy of the facet joints, ligamentum flavum, disc protrusion, spondylolisthesis, or by a combination of these. 4,5 Stenosis at multiple levels is more common than strictly segmental stenosis. In approximately 40% of cases central stenosis is caused by soft tissue hypertrophy. On computed tomography (CT) scans, midsagittal lumbar canal diameters less than 10 mm represent absolute stenosis and midsagittal lumbar canal diameters less than 13 mm represent relative stenosis. 29

Lateral spinal stenosis is a common cause of lumbar radiculopathy. The nerve root canal has been divided into three anatomic zones: the entrance zone, the midzone, and exit zone. 17 The entrance zone is the subarticular zone medial to the pedicle and is synonymous with the lateral recess. Its borders consist laterally of the pedicle, posteriorly of the superior articular facet, anteriorly of the posterolateral surface of the vertebral body caudally and the disc rostrally, and medially by the thecal sac. The root sleeve containing cerebrospinal fluid covers the nerve root at the entrance zone. Lateral to the entrance zone the nerve root sleeve coalesces with the nerve root and is devoid of cerebrospinal fluid. The minimal height of a normal lateral recess is 5 mm; a height of 3 to 4 mm is suggestive of lateral recess stenosis and a height of 2 mm is considered pathologic. 6

The majority of cases of lateral recess stenosis are produced by posterolateral disc protrusion or hypertrophy of the superior articular process also referred to as lateral recess syndrome.

The midzone is the part of the canal beneath the pars interarticularis and just inferior to the pedicle where the nerve root takes an oblique downward course from the lateral recess to the foramen. Anteriorly, the midzone is bordered by the posterior aspect of the vertebral body, posteriorly by the pars interarticularis and medially by the opening to the spinal canal.

Computed tomography scans accurately show the pars interarticularis and its relationship to the underlying nerve root. 19 A T1-weighted parasagittal magnetic resonance imaging (MRI) scan defines the pars as a high signal intensity bone marrow surrounded by the lower signal of the cortical bone. The bone marrow signal remains continuous from the superior to the inferior articular process. An interruption of this signal is indicative of a pars defect. 12,13

The most common causes of midzone nerve root compression include a pars defect or pedicular compression. A pars defect such as in isthmic spondylolisthesis with fibrocartilaginous tissue overgrowth can cause nerve root entrapment. More common in patients with rotational deformities or spondylolisthesis is kinking of the nerve root situated inferomedially to the pedicle by one pedicle that is lower than the other because of a rotation deformity or asymmetric disc collapse.

The exit zone corresponds to the intervertebral foramen. It is bordered superiorly and inferiorly by the pedicles of adjacent vertebrae, posteriorly by the pars interarticularis and ligamentum flavum, and anteriorly by the posteroinferior and posterosuperior aspects of the adjacent vertebral bodies and intervening disc. The foramen is shaped like an inverted tear-drop; its normal height varies from 10 to 23 mm and its width at the upper foramen varies from 8 to 10 mm. A foraminal height of less than 15 mm and a disc height of less than 4 mm are associated with nerve root compression 80% of the time. The ventral and dorsal nerve roots occupy 23% to 30% of the foramen and lie anterior to the dorsal root ganglion. The dorsal root ganglion usually is located in the superior lateral aspect of the foramen directly below the pedicle in 90% of lumbar levels. 7

Parasagittal T1-weighted images readily define the integrity of the foramen. The nerve root proper has a low signal and is surrounded by the higher intensity signal of fat. Obliteration of the fat pad often is indicative of foraminal stenosis.

Clinical Presentation

Degenerative spinal stenosis, the most common form of the disease, usually manifests in patients in the sixth or seventh decade of life, with a slight preponderance in women. The congenital form of spinal stenosis or canal and lateral recess stenosis usually manifest in patients in the third or fourth decade of life.

Degenerative spinal stenosis most commonly affects the L3-L4, and L4-L5 segments to cause cauda equina compression. Patients describe their symptoms as a discomfort ranging from a rubberylike feeling, leg weakness to actual pain in the back, buttocks, thighs, and legs. Lower extremity pain is present in approximately 80% of patients and back pain is present in approximately 65% of patients. 10

Pain often is localized poorly and frequently is associated with paresthesias. Symptoms may ascend from the distal lower extremities to the buttock, or alternatively descend the lower extremity. Symptoms generally are bilateral; they may not be symmetric and may affect the entire limb or parts thereof.

A pathognomonic aspect of lumbar stenosis is the relationship between symptoms and function. Symptoms are likely to manifest on prolonged standing or walking and decrease when the individual stops the provoking activity and rests. As the disorder progresses however, the individual’s time of activity before symptoms manifest shortens. Because of the clinical similarity to claudication caused by vascular insufficiency of the lower extremities, the lumbar stenosis pain syndrome has been termed pseudoclaudication or intermittent neurogenic claudication. 20,30

There are three prevailing theories that explain intermittent claudication: the ischemic theory, the mechanical compression theory and the theory of stagnant anoxia. The ischemic theory postulates that as metabolic demands increase during activity such as walking, this increased demand cannot be met because of an insufficient blood flow secondary to segmental compression. A relative nerve root ischemia ensues and may lead to pain, sensory loss, and a motor deficit. 3

In support of the vascular insufficiency theory is the fact that the intrafascicular microvascular matrix is predisposed to decrease in diameter and flow as a result of stretch and vessel angulation caused by bone overgrowth and stenosis. Under normal conditions, nerves are tolerant of traction because the intrafascicular arterial branches have compensating coils that can elongate on traction. In stenosis, acute angulation and tethering of the neural elements restrict the intrafascicular micromovement associated with traction, which results in narrowing of the blood vessels and diminished blood flow. 21

The fact that many patients with claudication attributable to cauda equina dysfunction have symptoms commensurate with posture rather than activity, has advanced the mechanical compression theory. In many such patients assuming a lordotic posture is sufficient to provoke symptoms that are alleviated by flexion.

The theory of stagnant anoxia may reconcile the vascular and mechanical-compressive hypotheses and explain the appearance of symptoms in static and dynamic conditions. In this hypothesis, the mechanical compression by bone and soft tissues may compress the neural elements, the draining veins that exit the canal with the spinal roots or cause cerebrospinal fluid entrapment, thus resulting in interference with venous return. 18 This dynamic entrapment of cerebrospinal fluid and ensuing increase in fluid pressure occurs distal to the site of narrowing or compression within a segment constricted rostrally and caudally by a two level stenosis. This increase in cerebrospinal fluid pressure may impede with the radicular venous return to culminate in relative hypoxia, or cause a reduction in the metabolic exchange and nutritional supply to the roots.

Often, patients with spinal stenosis assume a characteristic posture with the trunk in either an erect position or in a flexed forward position on gait. The assumption of stooped posture occurs gradually and is attributed to two phenomena that affect the cross section of the central canal and nerve outlets. In flexion, the vertebral canal lengthens and the spinal roots stretch. With extension, the canal shortens and the roots undergo an increase in their total volume. In addition, increase in spinal lordosis by trunk extension increases the bulging of the ligamentum flavum and intervertebral discs into the spinal canal and thereby compromises the size of the canal additionally. The onset of symptoms in the lower limbs after prolonged standing and their improvement on sitting, lying with the legs flexed, flexing at the waist, or squatting all are presumably attributable to the same mechanism.

Autonomic-sphincter dysfunction manifesting as recurrent urinary tract infections associated with an atonic bladder, incontinence, and more rarely, episodes of urinary retention are not infrequent and occur in approximately 10% of patients mainly with advanced stages of spinal stenosis. Other and rarer autonomic symptoms have been described, and in general autonomic dysfunction responds favorably to decompression. 23

A paucity of neurologic findings on physical examination, often despite a history of severe disability, is typical for patients with spinal stenosis. Furthermore, characteristic for the condition is the development of neurologic signs when the patient becomes symptomatic after a period of activity (walking) that provokes the symptoms. The most common findings are of deep tendon reflex changes, sensory loss, and muscle weakness. Straight leg raising rarely is positive, but flattening of the lumbar lordosis and a decrease in lumbar extension are common findings.

Some patients with stenosis primarily compressing a nerve root have symptoms of a radiculopathy but most experience a combined mixed symptomatology. Radicular pain is localized better, may be claudicant and intermittent, and can be associated with weakness in specific well-localized muscles or dermatomal sensory changes. One or several nerve roots may be involved, occasionally segmentally separated, and sometimes both lower limbs are affected.

The sparsity of neurologic findings are in marked contrast to the profound changes seen on myelography, CT scans, and MRI scans. The history rather than the objective clinical findings and imaging studies is the decisive factor in establishing the diagnosis. Of the ancillary laboratory studies that may be helpful in the diagnosis, except for imaging studies, is the neurophysiologic investigation. The results of electromyography will be abnormal in the majority of patients. Electromyography is considered to be more sensitive than the neurologic examination. Abnormalities seen on electromyography consist of denervation in muscles innervated by lumbosacral nerve roots. Findings often are bilateral and are located in the paraspinal area.

The differential diagnosis of spinal stenosis includes disc herniation and neoplasia that can be ruled out readily by imaging studies. More perplexing is the differentiation between intermittent neurogenic claudication attributable to cauda equina compression and claudication attributable to peripheral vascular disease. There often is an overlap in the two conditions.

Vascular intermittent claudication causes pain that is more cramplike, there is absence of one or more peripheral pulses, and often there are trophic changes in the extremities. Worsening neurologic symptoms and signs after ambulation or with an increase in the lordotic posture of the spine and/or relief of symptoms with a change in posture alone while exercise continues suggests neurogenic claudication. Walking-induced symptoms of neurogenic claudication often disappear when the patient sits or are relieved after a few minutes of rest. In vascular claudication, lower extremity symptoms often decrease or disappear even simply on standing or walking (Table 2).

Comparison of Vascular and Neurogenic Claudication

Osteoarthrosis of the hips may mimic spinal stenosis because of similar gait disturbance and buttock and proximal thigh pain. Careful examination of the hips is recommended and occasionally radiographs of the hips are warranted.

Gait disturbance and bladder incontinence are prominent symptoms of normal pressure hydrocephalus, a condition affecting patients in the same age range as those patients who are affected with spinal stenosis. Pain is not a feature of normal pressure hydrocephalus, the gait is characteristically shuffling and cognitive dysfunction is common. A noncontrast CT scan or MRI scan of the brain readily rules out this condition.

Lumbar spinal stenosis is defined as the narrowing of the central lumbar canal, the lateral recess, or the intervertebral foramen. The condition is classified as being either congenital or developmental or acquired. Stenosis is the result of bone remodeling and produces symptoms of gait disturbances, radiculopathy, motor or sensory deficits, bowel or bladder dysfunction, and neurogenic claudication. As the longevity of the population increases, clinicians will increasingly confront this clinical entity. Most patients will have significant improvement from decompression of the narrow segment. Age alone is not a contraindication for surgery. However, best results are achieved when a correlation is established between symptoms and radiographic findings and the decompression appropriately addresses the cause of the symptoms.


1. Arnoldi CC, Brodsky AE, Cauchoix J, et al: Lumbar spinal stenosis and nerve root entrapment syndromes: Definition and classification. Clin Orthop 115:4–5, 1976.
2. Baily P, Casamajor L: Osteo-arthritis of the spine as a cause of compression of the spinal cord and its roots: With report of five cases. J Nerv Ment Dis 38:588–609, 1911.
3. Blau JN, Logue V: The natural history of intermittent claudication of the cauda equina. Brain 101:211–222, 1978.
4. Bolander NF, Schonstrom NSR, Spengler DM: Role of computed tomography and myelography in the diagnosis of central spinal stenosis. J Bone Joint Surg 67A:240–246, 1985.
5. Carrera CF, Williams AL: Current concepts in evaluation of lumbar facet joints. CRC Crit Rev Diagn Imaging 21:85–104, 1985.
6. Ciric I, Mikael MA, Tarkington JA, et al: The lateral recess syndrome. J Neurosurg 53:433–443, 1980.
7. Cohen MS, Wall EJ, Frown RA, et al: Cauda equina anatomy II: Extrathecal nerve roots and dorsal root ganglia. Spine 15:1248–1251, 1990.
8. Elsberg CA: Experience in spinal surgery: Observation upon 60 laminectomies for spinal disease. Surg Gynecol Obstet 16:117–132, 1913.
9. Gruber HE, Hanley EN: Analysis of aging and degeneration of the human intervertebral disc. Comparison of surgical specimens with normal controls. Spine 23:751–757, 1998.
10. Hall SH, Bartleson JD, Onofrio BM, et al: Lumbar spinal stenosis: Clinical features, diagnostic procedures, and results of surgical treatment in 68 patients. Ann Intern Med 103:271–275, 1985.
11. Hayashi K, Ishidou Y, Yonemori K, et al: Expression and localization of bone morphogenetic proteins (BMPs) and BMP recetors in ossification of the ligamentum flavum. Bone 21:23–30, 1997.
12. Hasegawa T, An HS, Haughton VM, et al: Lumbar foraminal stenosis: Critical heights of the intervertebral discs and foramina: A cryomicrotome study in cadavera. J Bone Joint Surg 77A:32–38, 1995.
13. Hasegawa T, Mikawa Y, Watanabe R, et al: Morphometric analysis of the lumbosacral nerve roots and dorsal root ganglion by MRI. Spine 21:1005–1009, 1996.
14. Kennedy F, Elsberg CA, Lambert CI: A peculiar undescribed disease of the nerves of the cauda equina. Am J Med Sci 147:645–667, 1914.
15. Kirkaldy-Willis WH, Paine KWE, Cauchoix J, et al: Lumbar spinal stenosis. Clin Orthop 99:30–50, 1974.
16. Kirkaldy-Willis WH, McIvor GWD: Spinal stenosis. Clin Orthop 115:2–144, 1976.
17. Lee CK, Rauschning W, Glenn W: Lateral lumbar spinal canal stenosis: Classification, pathologic anatomy and surgical decompression. Spine 13:313–320, 1980.
18. Madsen JR, Heros RC: Spinal arteriovenous malformations and neurogenic claudication: Report of two cases. J Neurosurg 68:793–797, 1988.
19. McAfee PC, Yuan H: Computed tomography in spodylolisthesis. Clin Orthop 166:62–71, 1982.
20. Paine KWE: Clinical features of lumbar spinal stenosis. Clin Orthop 115:77–82, 1976.
21. Parke WW, Watanabe R: The intrinsic vasculature of the lumbosacral spine nerve roots. Spine 6:508–561, 1985.
22. Postacchini F: Surgical management of lumbar spinal stenosis. Spine 24:1043–1047, 1999.
23. Ram Z, Findler G, Spiegelman R, et al: Intermittent preapism in canal stenosis. Spine 12:377–378, 1987.
24. Sachs B, Fraenkel J: Progressive ankylotic rigidity of the spine. J Nerv Ment Dis 27:1–15, 1900.
25. Schlesinger EB, Taveras JM: Factors on the production of cauda equina syndromes in lumbar discs. Trans Am Neurol Assoc 78:263–265, 1953.
26. Takae R, Matsunaga S, Origuchi N, et al: Immunolocalization of bone morphogenetic protein and its receptors in degeneration of intervertebral disc. Spine 24:1397–1401, 1999.
27. Verbiest H: A radicular syndrome from developmental narrowing of the lumbar vertebral canal. J Bone Joint Surg 36B:230–237, 1954.
28. Verbiest H: Further experience on the pathological influence of a developmental narrowness of the bony lumbar vertebral canal. J Bone Joint Surg 37B:576–583, 1955.
29. Verbiest H: The significance and principles of computed axial tomography in idiopathic developmental stenosis of the bony lumbar vertebral canal. Spine 4:369–378, 1979.
30. Wilson CB: Significance of the small lumbar spinal canal: Cauda equina compression syndromes due to spondylosis, 3. Intermittent claudication. J Neurosurg 31:499–506, 1969.
31. Wilste LL: History of Spinal Disorders. In Frymoyer JW (ed). Adult Spine. New York, Raven Press 33–35, 1991.
32. Yong-Hing K, Kirkaldy-Willis WH: The pathophysiology of degenerative disease of the lumbar spine. Orthop Clin North Am 14:491–515, 1983.

Section Description

Kenneth K. Hansraj, MD; and Patrick F. O’Leary MD Guest Editors

© 2001 Lippincott Williams & Wilkins, Inc.