General Concepts

The lumbar spine has a dichotomous role in terms of function, which is strength coupled with flexibility. The spine performs a major role in support and protection (strength) of the spinal canal contents (spinal cord, conus, and cauda equina) but also give us inherent flexibility, allowing us to place our limbs in appropriate positions for everyday functions.

The strength of the spine results from the size and arrangements of the bones, as well as from the arrangement of the ligaments and muscles. The inherent flexibility results from the large number of joints placed so closely together in series. Each vertebral segment can be thought of as a three-joint complex—one intervertebral disk with vertebral end plates, and two zygapophyseal joints. The typical lordotic framework of the lumbar spine assists with this flexibility but also increases the ability of the lumbar spine to absorb shock.

The Vertebrae

The bony anatomy of the lumbar spine consists of five lumbar vertebrae. A smaller percentage of the population has four (the fifth vertebrae is sacralized) or six (the first sacral segment is lumbarized). Anatomic variants also exist consisting of a partially lumbarized S1. The lumbar vertebrae have distinct components, which include the vertebral body, the neural arch, and the posterior elements (Figure 40-1). The vertebral bodies increase in size as you travel caudally in the spine. The lower three are typically more wedge-shaped (taller anteriorly), which helps create the normal lumbar lordosis. The structure of these large vertebral bodies serves its weight-bearing function well to support axially directed loads; however, they would fracture more routinely were it not for the shock-absorbing intervertebral disks placed strategically between the vertebral bodies.

FIGURE 40-1 Lateral view of the lumbar vertebrae.

(Modified from Parke WW: Applied anatomy of the spine. In Rothman RH, Simeone FA, editors: The spine, ed 4, Philadelphia, 1999, Saunders.)

The sides of the bony neural arch are the pedicles, which are thick pillars that connect the posterior elements to the vertebral bodies. They are designed to resist bending and to transmit forces back and forth from the vertebral bodies to the posterior elements. The posterior elements consist of the laminae, the articular processes, and the spinous processes. The superior and inferior articular processes of adjacent vertebrae create the zygapophyseal joints. The pars interarticularis is a part of the lamina between the superior and inferior articular processes (Figure 40-2). The pars is the site of stress fractures (spondylolysis) because it is subjected to large bending forces. This occurs as the forces transmitted by the vertically oriented lamina undergo a change in direction into the horizontally oriented pedicle.²³

FIGURE 40-2 An oblique dorsal view of an L5 vertebra, showing the parts of the vertebral arch: 1, pars interarticularis (crosshatched area); 2, pars laminaris; and 3, pars pedicularis. The dotted line indicates the most frequent site of mechanical failure of the pars interarticularis.

(Modified from Parke WW: Applied anatomy of the spine. In Rothman RH, Simeone FA, editors: The spine, ed 4, Philadelphia, 1999, Saunders.)

The Joints

The Intervertebral Disk

The intervertebral disk and its attachment to the vertebral end plate are considered a secondary cartilaginous joint, or symphysis. The disk consists of the internal nucleus pulposus and the outer annulus fibrosus. The nucleus pulposus is the gelatinous inner section of the disk. It consists of water, proteoglycans, and collagen. The nucleus pulposus is 90% water at birth. Disks desiccate and degenerate as we age and lose some of their height, which is one reason we are slightly shorter in our geriatric years.

The annulus fibrosus consists of concentric layers of fibers at oblique angles to each other, which help to withstand strains in any direction. The outer fibers of the annulus have more collagen and less proteoglycans and water than the inner fibers.¹⁹ The varying composition supports the functional role of the outer fibers in acting like a ligament to resist flexion, extension, rotation, and distraction forces.

The main function of the intervertebral disk is shock absorption (Figure 40-3). It is primarily the annulus, not the nucleus, that acts as the shock absorber. (The nucleus is primarily a liquid and is incompressible.) When an axial load occurs, the increase in force in the incompressible nucleus pushes on the annulus and stretches its fibers. If the fibers break, then a herniated nucleus pulposus results.

FIGURE 40-3 The mechanism of weight transmission in an intervertebral disk. A, Compression increases the pressure in the nucleus pulposus. This is exerted radially onto the annulus fibrosus, and the tension in the annulus increases. B, The tension in the annulus is exerted on the nucleus, preventing it from expanding radially. Nuclear pressure is then exerted on the vertebral end plates. C, Weight is borne, in part, by the annulus fibrosus and by the nucleus pulposus. D, The radial pressure in the nucleus braces the annulus, and the pressure on the end plates transmits the load from one vertebra to the next.

(Modified from Bogduk N: The inter-body joint and the intervertebral discs. In Bogduk N, editor: Clinical anatomy of the lumbar spine and sacrum, ed 3, Edinburgh, UK, 1977, Churchill Livingstone.)

The Zygapophyseal Joints

The zygapophyseal joints (also known as Z joints and facet joints) are paired synovial joints, that is, they have a synovium and a capsule (Figure 40-4). Their alignment or direction of joint articulation determines the direction of motion of the adjacent vertebrae. The lumbar zygapophyseal joints lie in the sagittal plane and thus primarily allow flexion and extension. Some lateral bending and very little rotation are allowed, which limits torsional stress on the lumbar disks. Rotation is more a component of thoracic spine motion. The majority of spinal flexion and extension (90%) occurs at the L4–L5 and L5–S1 levels, which contributes to the high incidence of disk problems at these levels.

FIGURE 40-4 A posterior view of the L3–L4 zygapophyseal joints. On the left, the capsule of the joint (C) is intact. On the right, the posterior capsule has been resected to reveal the joint cavity, the articular cartilages (AC), and the line of attachment of the joint capsule (dashed line). The upper joint capsule (C) attaches further from the articular margin than the posterior capsule does.

(Modified from Bogduk N: The zygapophysial joints. In Bogduk N, editor: Clinical anatomy of the lumbar spine and sacrum, ed 3, Edinburgh, UK, 1977, Churchill Livingstone.)

Biomechanics

Because flexion loads the anterior disk, the nucleus is displaced posteriorly.¹¹¹ If the forces are great enough, the nucleus can herniate through the posterior annular fibers. The lateral fibers of the posterior longitudinal ligaments are thinnest, however, making posterolateral disk herniations the most common (Figure 40-5). The posterolateral portion of the disk is most at risk when there is forward flexion accompanied by lateral bending (i.e., bending and twisting). The zygapophyseal joints cannot resist rotation when the spine is in flexion. This increases torsional shear forces in the lumbar spine, making rotary movements in a forward-flexed posture probably the most risky for lumbar disks.

FIGURE 40-5 Posterolateral intervertebral disk herniation.

The Ligaments

The two main sets of ligaments of the lumbar spine are the longitudinal ligaments and the segmental ligaments. The two longitudinal ligaments are the anterior and posterior longitudinal ligaments. They are named according to their position on the vertebral body. The anterior longitudinal ligament acts to resist extension, translation, and rotation. The posterior longitudinal ligament acts to resist flexion. Disruption of either ligament primarily occurs with rotation rather than with flexion or extension. The anterior longitudinal ligament is twice as strong as the posterior longitudinal ligament.

The main segmental ligament is the ligamentum flavum, which is a paired structure joining adjacent laminae. It is the ligament that is pierced when performing lumbar punctures. It is a very strong ligament but is elastic enough to allow flexion. Flexing the lumbar spine puts this ligament on stretch, decreasing its redundancy and making it easier to pierce during a lumbar puncture.

The other segmental ligaments are the supraspinous, interspinous, and intertransverse. The supraspinous ligaments are the strong ligaments that join the tips of adjacent spinous processes and act to resist flexion. These ligaments, along with the ligamentum flavum, act to restrain the spine and prevent excessive shear forces in forward bending. This is supported by electromyographic studies that have shown that there is no active contraction of the erector spinae and hip extensor muscles when resting in lumbar flexion.

The Muscles

Muscles With Origins on the Lumbar Spine

These muscles can be divided anatomically into posterior and anterior muscles. The posterior muscles include the latissimus dorsi and the paraspinals. The lumbar paraspinals consist of the erector spinae (iliocostalis, longissimus, and spinalis), which act as the chief extensors of the spine, and the deep layer (rotators and multifidi) (Figures 40-6 and 40-7). The multifidi are tiny segmental stabilizers that act to control lumbar flexion because they cannot produce enough force to truly extend the spine. Their most important function has been hypothesized to be that of a sensory organ to provide proprioception for the spine, given the predominance of muscle spindles seen histologically in these muscles.

FIGURE 40-6 The intermediate layer of back muscles: the erector spinae.

FIGURE 40-7 The deep back muscles: the multifidi.

The anterior muscles of the lumbar spine include the psoas and quadratus lumborum. Because of the direct attachment of the psoas on the lumbar spine, tightening this muscle accentuates the normal lumbar lordosis. This can increase forces on the posterior elements and can contribute to zygapophyseal joint pain. The quadratus lumborum acts in side bending and can assist in lumbar flexion.

Abdominal Musculature

The superficial abdominals include the rectus abdominis and external obliques (Figure 40-8, A). The deep layer consists of internal obliques and the transversus abdominis (Figure 40-8, B). The transversus abdominis has received significant attention recently as an important muscle to train in treating low back pain. Its connection to the thoracolumbar fascia (and consequently its ability to act on the lumbar spine) has probably been the major reason it has received such attention of late.

FIGURE 40-8 A, The superficial abdominal muscles. B, The deep abdominal muscles.

Biomechanical Lifting in Relation to Muscular Activity and Disk Loads

The activity of the lumbar muscles correlates well with intradiskal pressures (i.e., when back muscles contract, there is an associated increase in disk pressure). These pressures change depending on spine posture and the activity undertaken. Figure 40-9 demonstrates the changes in L3 disk pressure under various positions and exercises.^149,150 Adding rotation to the already flexed posture increases the disk pressure substantially. Comparing lifting maneuvers, it has been shown that there is not a significant difference in disk pressure when lifting with the legs (i.e., with the back straight and knees bent) versus lifting with the back (i.e., with a forward-flexed back and straight legs).^6,7 What decreases the forces on the lumbar spine is lifting the load close to your body because the farther the load is from the chest, the greater the stress on the lumbar spine.⁷

FIGURE 40-9 A, Relative change in pressure (or load) in the third lumbar disk in various positions in living subjects. B, Relative change in pressure (or load) in the third lumbar disk during various muscle-strengthening exercises in living subjects. Neutral erect posture is considered 100% in these figures; other positions and activities are calculated in relationship to this.

(Modified from Nachemson AL, Waddell G, Norlund AI: Epidemiology of neck and low back pain. In Nachemson AL, Johnsson B, editors: Neck and back pain: The scientific evidence of causes, diagnosis, and treatment, Philadelphia, 2000, Lippincott Williams & Wilkins.)

The Nerves

The conus medullaris ends at about bony level L2, and below this level is the cauda equina. The cauda equina consists of the dorsal and ventral rootlets, which join together in the intervertebral neuroforamen to become the spinal nerves (Figure 40-10). The spinal nerve gives off the ventral primary ramus. The ventral primary rami from multiple levels form the lumbar and lumbosacral plexus to innervate the limbs. The dorsal primary ramus, with its three branches (medial, intermediate, and lateral), innervates the posterior half of the vertebral body, the paraspinal muscles, and the zygapophyseal joints, and provides sensation to the back. The medial branch is the most important to remember because it innervates the zygapophyseal joints and lumbar multifidi and is the target during radiofrequency neurotomy for presumed zygapophyseal joint pain (Figure 40-11).²⁴

FIGURE 40-10 A lumbar spinal nerve, its roots, and meningeal coverings. The nerve roots are invested by pia mater, and covered by arachnoid and dura as far as the spinal nerve. The dura of the dural sac is prolonged around the roots as their dural sleeve, which blends with the epineurium of the spinal nerve.

(Modified from Bogduk N: Nerves of the lumbar spine. In Bogduk N, editor: Clinical anatomy of the lumbar spine and sacrum, ed 3, Edinburgh, UK, 1977, Churchill Livingstone.)

FIGURE 40-11 Observe that the innervation of the zygapophyseal joints derives from the medial branch off the dorsal primary ramus.

The Degenerative Spine Cascade

Kirkaldy-Willis et al.¹⁰⁸ have supplied us with the most accepted theory describing the cascade of events in degenerative lumbar spine disease that results in spondylotic changes, disk herniations, and eventually multilevel spinal stenosis (Figure 40-12). At the heart of this theory is the fact that, although the posterior zygapophyseal joints and the anterior intervertebral disks are separated anatomically, forces and lesions affecting one certainly alter and affect the other. For example, axial compressing injuries can damage the vertebral end plates, which can lead to degenerative disk disease, which eventually stresses the posterior joints, leading to the common degenerative changes seen in them over time. Torsional stress can injure the posterior joints and the disks, which in turn leads to increased stress on both these elements, resulting in further degenerative changes over time. When these degenerative changes affect one level, such as L4–L5, a chain reaction occurs, placing stress on the levels above and below the currently affected level, and eventually resulting in more generalized multilevel spondylotic changes.

FIGURE 40-12 The spectrum of degenerative change that leads from minor strains to marked spondylosis and stenosis.

(Modified from Kirkaldy-Willis WH, Wedge JH, Yong-Hing K, et al: Pathology and pathogenesis of lumbar spondylosis and stenosis, Spine 3:319-328, 1998, with permission of Lippincott Williams & Wilkins.)

In studying lumbar degenerative disease, the question of which came first (disk degeneration or zygapophyseal joint degeneration) always arises. Fujiwara⁶⁷ has answered this by studying multiple magnetic resonance images (MRIs) of aging spines. He hypothesizes that disk degeneration precedes zygapophyseal joint osteoarthritis, and that it might take 20 years for zygapophyseal joint disease to occur after the onset of disk degeneration.

To describe the degenerative cascade in more detail, we will separate our discussion of the changes that occur in the posterior joints from those in the disk, but fully realizing that they both can occur simultaneously and affect each other (see Figure 40-12). The degenerative changes that occur in the zygapophyseal joints from aging and repetitive microtrauma are similar to those that occur in the appendicular skeletal joints. The process begins with synovial hypertrophy, which eventually results in cartilage degeneration and destruction. With lessened and weakened cartilage and capsular laxity, the joint can become unstable. With the repetitive abnormal joint motion that results from this instability, the bony joint hypertrophies. This narrows the central canal and lateral recesses, potentially impinging nerve roots.

A similar process occurs anteriorly at the disk level from repetitive microtrauma of primarily shearing forces. Tears in the annulus are thought to be the first anatomic sign of degenerative wear. When the annulus is weakened enough, typically posterolaterally, the internal nucleus pulposus can herniate. Internal disk disruption can occur without herniation, however, because age and repeated stresses acting on the spine cause the gelatinous nucleus to become more fibrous over time. Tears in the annulus can progress to tears in the fibrous disk material, resulting in “internal disk disruption” without frank herniation. All this results in a loss of disk height, which causes instability (because the end-plate connection to the disk is degenerated), as well as lateral recess and foraminal narrowing and potential nerve root impingement. The loss of disk height also places new stresses on the posterior elements, resulting in further instability of the zygapophyseal joints and further degeneration and nerve root impingement.

Although this theory offers an explanation as to how the spine ages, it is still unclear why there is such a marked disconnect between the anatomic changes in the spine associated with aging, and back pain. Many patients with normal spine anatomy suffer from back pain, occasionally disabling pain, and many patients with marked degenerative changes on imaging are nearly or fully pain free. As a result, several theories have been developed to explain the occurrence of back pain.

Radiculitis and Radiculopathy

Many patients with radicular pain have no neural impingement noted on MRI. Studies have shown that disk herniations can cause an inflammatory response.^131,136,187 The mechanism stems from the fact that the nucleus pulposus is highly antigenic as a result of being in an immunoprotected setting in nonpathologic states. When the fluid of the nucleus pulposus is exposed to neural tissue of the spinal canal and neuroforamen through a defect in the annular fibers, an autoimmune-mediated inflammatory cascade begins. The inflammatory mediators generated can cause swelling of the nerves. This can alter their electrophysiologic function, sensitizing these neurons and enhancing pain generation without specific mechanical compression.

The mechanism of mechanical compression of the nerve roots has been studied as well.^13,184,185 Compression of nerve roots can induce structural and vascular changes as well as inflammation.¹⁵⁰ Neural compression can result in impairment of intraneural blood flow, subsequently decreased nutrient supply to the neural tissue, local ischemia, and formation of intraneural edema. This can set off an inflammatory cascade similar to that described above. Mechanical stimulation of lumbar nerve roots has also been shown to stimulate production of substance P, the neuropeptide known to modulate sensory nociceptive feedback.¹³ With these biochemical reactions, the local structural effects of mechanical compression (demyelination and axonal transport block) just compound the symptomatic response.

Pain of Spinal Stenosis

These mechanisms could also be responsible for neurogenic claudication symptoms of lumbar stenosis. Newer theories, however, also support a spinal vascular role in stenosis symptoms.

If mechanical compression were the sole problem in spinal stenosis, decompressive surgeries would be the only needed cure. We know that this is untrue, and consequently alternative theories on the pathogenesis of symptomatic spinal stenosis have been studied. Two theories supporting a vascular component to symptoms of spinal stenosis are the venous engorgement and arterial insufficiency theories.¹

In the venous engorgement theory, the spinal veins of patients with stenosis dilate, causing venous congestion and stagnating blood flow.⁴⁵ This pooling of blood in the spinal veins increases epidural and intrathecal pressures, leading to a microcirculatory, neuroischemic insult (i.e., an ischemic neuritis), which in turn leads to the typical neurogenic claudication symptoms of stenosis.

The arterial insufficiency theory of spinal stenosis is based on the arterial dilatation of the lumbar radicular vessels during lower limb exercise to provide increased blood flow and nourishment to the nerve roots. In patients with spinal stenosis, this reflex dilatation might be defective.¹⁴Because patients with spinal stenosis are typically elderly, they are also at higher risk for atherosclerosis, which in turn just amplifies the arterial insufficiency.

Pain Generators of the Lumbar Spine

The low back is an anatomically diverse set of structures, and there are many potential sources of pain. One useful strategy to clarify these potential sources of pain is learning what low back structures are innervated (and can transmit pain through neural pain fibers) and what structures have no innervation (Box 40-1).

The sinuvertebral nerve innervates the anterior vertebral body, the external annulus, and the posterior longitudinal ligament. The posterior longitudinal ligament is a highly innervated structure and can play a significant role in low back pain perception with lumbar disk herniations. The medial branch of the dorsal primary ramus innervates the zygapophyseal joints and interspinous ligaments, as well as the lumbar multifidi. The other small branches of the dorsal primary ramus innervate the posterior vertebral body and other lumbar paraspinal musculature and fascia. The anterior longitudinal ligament is innervated by the gray rami communicans, which branch off the lumbar sympathetic chain. The internal annulus fibrosus and nucleus pulposus do not have innervation and in nondisease states cannot transmit pain.

Because many structures are potentially a source of pain, theories have been developed to help practitioners determine the cause of a particular patient’s pain. Some of the most common ones are described below.

The History

As with any pain history, features of back pain that should be explored include location; character; severity; timing, including onset, duration, and frequency; alleviating and aggravating factors; and associated signs and symptoms. Each of these features can assist the clinician in obtaining a diagnosis and prognosis and determining the appropriate treatment. The causes of back pain are often very difficult to determine. For as many as 85% of patients, no specific cause for back pain is found.⁵⁰ One of the main purposes of the history is to rule out rare but serious causes of back pain. Elements of historical information that suggest a serious underlying condition as the cause of the pain such as cancer, infection, long tract signs, and fracture are called red flags (Box 40-2). When these are present, further workup is necessary (Table 40-1).

Besides determining a diagnosis, a purpose of the history is to explore the patient’s perspective and illness experience. Certain psychosocial factors are valuable in determining prognosis (Box 40-3). Factors such as poor job satisfaction, catastrophic thinking patterns about pain, the presence of depression, and excessive rest or downtime are much more common in patients in whom back pain becomes disabling. These are called yellow flags because the clinician should proceed with caution, and further psychologic evaluation or treatment should be considered if they are present. Some of these psychosocial factors are addressed by specific questions, and some become evident through statements that patients make during the history as they describe their illness experience. Questions about, for example, what patients believe is causing the pain, their fear and feelings surrounding this belief, their expectations about the pain and its treatment, and how back pain is affecting their lives (including work and home life) can yield valuable information. Many of these yellow flags are better prognostic indicators than the more traditional medical diagnoses.²³⁵

The Physical Examination

Table 40-2 outlines a thorough examination of the lumbar spine.

Table 40-2 Physical Examination for Low Back Pain

Observation

Observation should include a survey of the skin, muscle mass, and bony structures, as well as observation of overall posture (Figures 40-13 and 40-14), and the position of the lumbar spine in particular. Gait should also be observed for clues regarding etiology and contributing factors.

FIGURE 40-13 The four types of postural alignment: Ideal alignment (A), kyphosis–lordosis posture (B), flat back posture (C), and sway-back posture (D).

(Modified from Kendall FP, McCreary EK: Trunk muscles in muscle testing and function, Philadelphia, 1983, Williams & Wilkins.)

FIGURE 40-14 Faulty pelvic alignment as a result of weak and long abdominal muscles (A), short and stiff hip flexors (B), apparent anterior tilt (C), and posterior tilt (D). The effect of pelvic tilting on the inclination of the base of the sacrum to the transverse plane (sacral angle) during upright standing is shown. E, Tilting the pelvis backward reduces the sacral angle and flattens the lumbar spine. F, During relaxed standing, the sacral angle is about 30 degrees. G, Tilting the pelvis forward increases the sacral angle and accentuates the lumbar spine.

(Modified from Sahrmann SA: Movement impairment syndromes of the lumbar spine: diagnosis and treatment of movement impairment syndromes, St Louis, 2002, Mosby.)

Palpation

Palpation should begin superficially and progress to deeper tissues. It can be done with the patient standing. To ensure that the back muscles (Figure 40-15) are fully relaxed, palpation is often done with the patient lying prone, perhaps with a pillow under the abdomen to slightly flex the spine into a position of comfort. It should proceed systematically to determine what structures are tender to palpation.

FIGURE 40-15 Anatomy of the low back surface anatomy.

Orthopedic Special Tests to Assess for Relative Strength and Flexibility

Back pain can be caused by deconditioning, poor endurance, and muscle imbalances. This makes it important to identify any inefficient or abnormal movement patterns of muscles that control the movement of the spine and the position of the pelvis.

Because of their stabilizing effect on the spine, abdominal muscle strength and endurance is important. Several different ways can be used to measure abdominal muscle strength and control (Figures 40-16 and 40-17). One grading system assesses whether the patient is able to maintain a neutral spine position while adding increasingly more challenging leg movements(Figure 40-18).

FIGURE 40-16 Trunk raising forward: grading. The curl trunk sit-up is performed with the patient lying supine and with the leg extended. The patient posteriorly tilts the pelvis and flexes the spine, and slowly completes a curled trunk sit-up. Kendall states that the “crucial point in the test for the abdominal muscle strength is at the moment the hip flexors come into strong action. The abdominal muscle at this point must be able to oppose the force of the hip flexors in addition to maintain the trunk curl.” At the point where the hip flexors strongly contract, patients with weak abdominal muscles will tilt the pelvis anteriorly and extend the low back. A, A 100% or normal grade is the ability to maintain spinal flexion and come into the sitting position with the hands clasped behind the head. B, An 80% or good grade is the ability to do this with the forearms folded across the chest. C, A 60% or fair grade is the ability to do this with the forearms extended forward. A 50% or fair grade is the ability to begin flexion but not maintain spinal flexion with the forearms extended forward.

(Modified from Kendall FP, McCreary EK: Trunk muscles in muscle testing and function, Philadelphia, 1983, Williams and Wilkins.)

FIGURE 40-17 Leg lowering: grading. In the second test, the patient raises the legs one at a time to a right angle, and then flattens them back on the table. The patient slowly lowers the legs while holding the back flat. A 100% or normal grade is the ability to hold the low back flat on the table as the legs are lowered to the fully extended position. An 80% or good grade is the ability to hold the low back flat and lower the legs to a 30-degree angle. A, A 60% or fair plus grade is the ability to lower the legs to 60 degrees with the low back flat. B, The pelvis tilted anteriorly and the low back arched as the legs were lowered. C, The final position. Kendall notes that this second test is more important than the first (see Figure 40-16) in grading muscles essential to proper posture, and that often patients who do well on the first test do poorly on the second.

(Modified from Kendall FP, McCreary EK: Trunk muscles in muscle testing and function, Philadelphia, 1983, Williams and Wilkins.)

FIGURE 40-18 Abdominal strength grading. A, The patient lies supine with the knees bent (supine crook lying). The physician cues the patient to activate the transversus abdominis (“Pull your belly button toward your backbone”), and a very slight lumbar lordosis is maintained in a neutral position in which the spine is neither flexed nor extended. The ability to maintain the neutral spine is progressively challenged by loading the spine via lower extremity movements. Grading is as follows. B, Grade 1: The patient is able to maintain a neutral spine while extending one leg by dragging the heel along the table; the other leg remains in the starting position. C, Grade 2: The patient is able to maintain a neutral spine while holding both legs flexed 90 degrees at the hip and 90 degrees at the knee, and touching one foot to the mat and then the other. D, Grade 3: The patient is able to maintain a neutral spine while extending one leg by dragging the heel along the table. The other leg is off the mat and flexed 90 degrees at the hip and 90 degrees at the knee. E, Grade 4: The patient is able to maintain a neutral spine while extending one leg hovered an inch or two above the table, while the other leg is off the mat and flexed 90 degrees at the hip and 90 degrees at the knee. Grade 5: The patient is able to extend both legs a few inches off the mat and back again while maintaining the spine in neutral.

Besides determining the strength of the abdominals, strength testing of the back muscles and pelvic stabilizers, such as the hip abductors, can be useful. Assessing for areas of relative inflexibility is also important. Commonly performed tests are hip flexor flexibility, hamstring flexibility, other hip extensors’ length, and gastrocnemius/soleus length. Balance challenges, such as the ability to maintain single-footed stance, the ability to lunge or squat, and other functional tests are also helpful to determine a patient’s baseline status.

Imaging Studies

Imaging of the lumbar spine should be used in the evaluation of low back pain if specific pathology needs to be confirmed after a thorough history and physical examination.

Plain Radiography

Conventional radiographs are indicated in trauma to evaluate for fracture and to look for bony lesions such as tumor when red flags are present in the history. As an initial screening tool for lumbar spine pathology, however, they have very low sensitivity and specificity.⁷³ Anterior–posterior and lateral views are the two commonly obtained views. Oblique views can be obtained to examine for a spondylolysis by visualizing the pars interarticularis and the “Scottie dog” appearance of the lumbar spine (Figure 40-19). Lateral flexion–extension views are obtained to check for dynamic instability, although the literature does not support their usefulness.⁵⁵ They are potentially most helpful from a surgical screening perspective when evaluating a spondylolisthesis. They are commonly obtained in posttrauma and postsurgical patients.

FIGURE 40-19 Oblique drawing of the lumbosacral junction, outlining the “Scottie dog” and the area of spondylolysis.

Magnetic Resonance Imaging

MRI is the preeminent imaging method for evaluating degenerative disk disease, disk herniations, and radiculopathy (Figure 40-20) (see also Chapter 7). On T₂-weighted imaging, the annulus can be differentiated from the internal nucleus, and annular tears can be seen as high-intensity zones. These zones are of unclear clinical significance but are thought to be potential pain generators.

FIGURE 40-20 Disk extrusion in a 48-year-old woman with back and left leg pain. A and B, Sagittal T₂– and T₁-weighted magnetic resonance image (MRI) showing L5–S1 disk extrusion with caudal extension. C, Axial T₂-weighted MRI showing the extrusion is left paracentral in the lateral recess, occupying the space where the S1 root resides.

Adding gadolinium contrast enhancement helps to identify structures with increased vascularity. Contrast is always indicated in evaluating for tumor or infection or to determine scar tissue (vascular) versus recurrent disk herniation (avascular) in postsurgical patients with recurrent radicular symptoms.

The downside of MRI is that, although it is a very sensitive test, it is not very specific in determining a definite source of pain. It is well established that many people without back pain have degenerative changes, disk bulges, and protrusions on MRI. Boden et al.²¹ demonstrated that one third of 67 asymptomatic subjects were found to have a “substantial abnormality” on MRI of the lumbar spine. Of the subjects younger than 60 years, 20% had a disk herniation, and 36% of those older than 60 had a disk herniation and 21% had spinal stenosis. Bulging and degenerative disks were even more commonly found. In another study of lumbar MRI findings in people without back pain, Jensen et al.⁹⁷ demonstrated that only 36% of 98 patients had normal disks. They found that bulges and protrusions were very common in asymptomatic subjects, but that extrusions were not. In a more recent study in 2001, Jarvik confirmed these findings.⁹⁶

Myelography

In myelography, contrast dye is injected into the dural sac and plain radiographs are performed to produce images of the borders and contents of the dural sac (Figure 40-21). CT images can also be obtained after contrast injection to produce axial cross-sectional images of the spine that enhance the distinction between the dural sac and its surrounding structures. This is typically reserved as a potential presurgical screening tool but has been used less with the advancement of MRI.

FIGURE 40-21 Anteroposterior (A) and lateral (B) myelograms of a 59-year-old woman with severe L4–L5 central stenosis caused by a large left L4–L5 zygapophyseal joint synovial cyst. Note the obvious filling defect at the L4–L5 level. She had symptoms of cauda equina syndrome and regained full neurologic function after decompression surgery.

Laboratory Studies

Blood tests are rarely used in isolation as a diagnostic strategy for low back pain. Some blood tests are helpful as an adjunct in diagnosing inflammatory disease of the spine (with such markers of inflammation as sedimentation rates and C-reactive proteins), as well as some neoplastic disorders, such as multiple myeloma with a serum protein electrophoresis and urine protein electrophoresis.

Mechanical Low Back Pain

Nearly 85% of those who seek medical care for low back pain do not receive a specific diagnosis.⁵⁰ The majority of these patients most probably have a multifactorial cause for back pain, which includes deconditioning; poor muscle recruitment; emotional stress; and changes associated with aging and injury such as disk degeneration, arthritis, and ligamentous hypertrophy. This type of back pain can be given many names; simple backache, nonspecific low back pain, lumbar strain, and spinal degeneration are a few of the common names for this condition. The name given to a condition sends certain messages to the patient who receives the diagnosis. The term simple backache might cause a patient to think that the physician misunderstands because, from the patient’s perspective, the pain is not simple if it has not resolved in a few days. The label nonspecific low back pain can cause the patient to continue to seek care from multiple providers to receive a specific diagnosis. Lumbar strain suggests that the condition was caused by overactivity, which is often not the case, and that further physical activity would cause it to recur, which is not true. Spinal degeneration sends the message that the changes are permanent and will probably worsen.²³⁵ The term mechanical low back pain is perhaps the best term for this multifactorial axial backache. It suggests the mechanism of injury better than terms such as strain or sprain, and it does not imply permanence.

The biomechanics of the spine are not unlike the biomechanics of other systems, in that longevity of the components and efficiency of the system depend on precise movements of each segment. In the spine, this means both an alignment in sustained postures and movement patterns that reduce tissue strain and allow for efficient muscle action without trauma to the joints or soft tissue.¹⁸⁹ Clinicians and researchers alike theorize that when alignment and movement patterns deviate from the ideal, degeneration and tissue overload is more likely. This is analogous to the abnormal tire wear that occurs on a car when the wheels are out of alignment. Unlike machinery, the body can adapt over the course of time to stress on the segments. This adaptation can be the healthy response of tissue to loading (as is seen with exercise), such as muscle hypertrophy or increased bone density. It can also, however, begin a cycle of microtrauma that can lead to macrotrauma.^125,140,189 The theoretic model for this approach is strong, and research is beginning to validate many of these concepts, although this is not easy given the complex nature of the system.

Other Causes of Back Pain Greater Than Leg Pain

Disk Herniation

The terminology used to describe disk material that extends beyond the intervertebral disk space is confusing. Herniated disk, herniated nucleus pulposus, disk protrusion, disk bulge, ruptured disk, and prolapsed disk are all commonly used terms, and sometimes are incorrectly used synonymously. Displaced disk material can be initially classified as a bulge (disk material is displaced >50% of its circumference) or as a herniation (<50% of its circumference) (Figure 41-22).⁵⁷ Disk herniations can then be subclassified into protrusions or extrusions. A disk protrusion is defined as a herniation with the distance of the edges of the herniated material less than the distance of the edges at its base. A disk extrusion occurs when the distance of the edges of the herniated material is greater than the distance of the edges at its base. A disk extrusion can be further subclassified as sequestrated if the extruded disk material has no continuity with the disk of origin. Disk herniations can also be described as contained or uncontained depending on the integrity of the outer annular fibers. If the outer annular fibers are still intact, it is described as a contained disk herniation. This classification has no relevance to the integrity of the posterior longitudinal ligament.

FIGURE 40-22 Disk herniation, protrusion, and extrusion.

(Modified from Maus TP: Imaging of the spine and nerve roots. In Kraft GH, editor: Physical medicine and rehabilitation clinics of North America, Philadelphia, 2002, Saunders.)

More than 95% of lumbar disk herniations occur at the L4–L5 and L5–S1 levels.^49,204 Next most common is L3–L4, followed by L2–L3. The most common lumbosacral radiculopathies are consequently L5 and S1. Posterolateral disk herniations are most common because the annulus fibrosus is weakest posterolaterally. Posterolateral disks can affect the nerve root as it descends in the lateral recess or just before it enters the neural foramen. Far lateral or extraforaminal herniations can affect the nerve root as it exits the neural foramen, and central disk herniations can affect any part of the cauda equina, depending on the level.

Disk herniations can cause an inflammatory response that can affect the nerve root, or there can be mechanical compression, both of which can cause radicular symptoms. Disk herniations, however, can also cause solely axial pain. Diagnosing diskogenic low back pain is a challenge because we know asymptomatic subjects can have disk herniations present on MRI.^21,96,97 Discography is a controversial diagnostic tool for diskogenic pain (see Chapter 25). It is typically used as a presurgical screening tool.

Spondylolysis

Spondylolysis is a defect of the pars interarticularis and is a common cause of back pain in children and adolescents. The most common hypothesized mechanism of injury is repetitive hyperextension loading in the immature spine, and is commonly reported in adolescent gymnasts and football linemen.^60,94 Acute fracture from a severe hyperextension injury is also possible but less commonly reported.⁵⁹ Pars defects have been reported in nonathletic individuals as well. In growing children, the defect is rarely seen before walking begins and most commonly occurs at age 7 to 8 years.¹⁵³ An increase in incidence occurs during the adolescent growth spurt between ages 11 and 15 years. The pars defect appears to result from a combination of hereditary dysplasia of the pars and repetitive stressing of the spine by walking and extension loading.⁵¹ Unilateral or bilateral defects can occur, but bilateral involvement might result in spondylolisthesis. Ninety percent of these lesions occur at the L5–S1 level.

Patients typically present with low back pain that is exacerbated by extension and alleviated by rest or activity limitation. Physical examination can demonstrate focal tenderness, pain with lumbar extension, and hamstring tightness (Figure 40-23). The neurologic examination is usually normal. If a spondylolisthesis is present, a palpable step-off with examination of the spinous processes might be evident.

FIGURE 40-23 The standing one-leg hyperextension test.

The ideal radiographic assessment of a suspected pars injury is debated. Plain radiographs are of limited use in diagnosing a symptomatic spondylolysis. Oblique views can show a pars defect well; however, the sensitivity of detection on plain films does not increase much, and the radiation exposure is significantly greater.¹⁷⁴ Standing anteroposterior and lateral radiographs can be obtained initially to identify a spondylolisthesis or any gross bony abnormalities.²⁰⁷ SPECT is more sensitive than plain radiography and planar bone scans. A positive bone scan or SPECT scan correlates with a painful pars lesion.⁸⁶ In most retrospective analyses, however, increased uptake is present for approximately 1 year after the occurrence of the fracture.¹⁷⁴ If the SPECT study is consistent with an active pars injury, a thin-slice CT through the abnormal level can be helpful to confirm the diagnosis and stage the lesion. Staging the lesion according to chronicity is helpful to predict healing. The Tokushima classification grades defects by CT as acute, progressive, or terminal (Box 40-6).⁶⁶ A recent metaanalysis demonstrated by using this classification system 68% of acute defects healed, 28% of progressive lesions healed, and no terminal lesions healed.¹⁰⁹ Determining chronicity can help the clinician decide how quickly the athlete should progress through treatment. The use of MRI in diagnosing spondylolysis has garnered much attention in recent years. Standard magnetic resonance sequences identify only 80% of pars lesion seen on SPECT and should be used to look for other pathology.¹³³ Campbell et al.,³⁸ however, showed that nonstandard magnetic resonance sequences (oblique sagittal images) found 39 of 40 pars defects seen on CT and SPECT, but still MRI only correctly staged the lesion.²⁹

A number of successful management strategies for spondylolysis have been used. Conservative management is most common, typically beginning with relative rest and avoidance of activities that increase pain (especially repetitive extension). Patients typically are advised to rest for 3 months.²⁰⁷ This is the shortest amount of time over which healing of a pars lesion has been identified on serial imaging.²⁵³ Bracing, although a very commonly cited treatment, is not absolutely necessary. A recent metaanalysis demonstrated 84% of 665 patients with a spondylolysis treated nonoperatively were able to return to pain-free unrestricted activity by 1 year. No significant difference in clinical outcomes was seen between the braced and not-braced groups.¹⁰⁹ A rigid brace might be considered after 2 weeks of rest if symptoms are not resolving.²⁰⁷ A radiographically stable bony union is the obvious goal but is not an absolute necessity, given many young athletes can become symptom-free in sports and everyday life even with a persistent pars defect. That same metaanalysis studied 847 defects and showed a 28% radiographic healing rate, with 71% of the unilateral defects healed and only 18% of the bilateral defects healed.¹⁰⁹

In any event, braced or not, the young athlete is at risk for deconditioning. When pain allows, the patient should be encouraged to begin aerobic conditioning and eventually enter a spinal rehabilitation program before return to a sport. Once the athlete has mastered a basic core stabilization program, functional progression back to the specific sport is appropriate, with a focus on neuromuscular proprioceptive control and sport-specific drills before full return to play. For the patient with chronic low back pain and spondylolysis, O’Sullivan et al.¹⁵⁸ demonstrated that a specific exercise program focused on training the lumbar multifidi and deep abdominals can be very effective. Surgical treatment is rarely indicated for the patient with spondylolysis alone but is more common in the setting of spondylolisthesis and/or radiculopathy.

The natural history of spondylolysis and low-grade (<2) spondylolisthesis is benign, that is, it is rare to have progressive slippage. Saraste¹⁹⁰ demonstrated this in a study of 225 patients with a 20-year follow-up period. Most cases of progressive slippage occur during the adolescent growth spurt, however, so very young athletes should be monitored with lateral flexion–extension plain films. Besides the adolescent growth spurt, a listhesis greater than 50% is considered a risk factor for progressive slip.

Spondylolisthesis

Lumbar spondylolisthesis or anterior slippage of one vertebra on another can result from many causes. Spondylolisthesis can be grouped into six different categories by etiology. The most common is the isthmic spondylolisthesis (Figure 40-24). The isthmic slip occurs as a result of a spondylolysis or “stress fracture” of the pars interarticularis (as described above). The dysplastic spondylolisthesis is a congenital slip and is caused by dysplasia of the facet joints of the upper sacrum, leading to an inability to resist shear stresses and forward slippage. Degenerative spondylolisthesis is seen in the older spine and is related to long-standing intersegmental instability from degenerative facet or disk disease. The most common level affected in a degenerative slip is the L4–L5 level. Traumatic spondylolisthesis is rare and is caused by acute fracture secondary to trauma. Pathologic spondylolisthesis is due to medical causes of generalized or local bone disease that can cause decreased bony strength. This form can present as an isthmic defect or an elongated, intact pars. The final category is postsurgical and is due to resulting instability from an extensive decompression, which is uncommon now because of the amount of hardware used for fusions after extensive decompressions.

FIGURE 40-24 Grade 1 isthmic spondylolisthesis.

The patient with spondylolisthesis typically presents with low back pain. Sometimes there is a complaint of intermittent radicular symptoms related to a dynamic radiculitis, that is, nerve root irritation caused by subtle instability at the listhetic segment. Physical examination is not different from that seen in spondylolysis. When imaging a patient with suspected spondylolisthesis, lateral flexion–extension views are helpful for presurgical screening. With lateral plain films, the degree of slip is graded 1 through 5 (Table 40-6).

Table 40-6 Meyerding’s Grading System for Spondylolisthesis^∗

∗ Meyerding divided the anterior–posterior diameter of the superior surface of the first sacral vertebral into quarters and assigned the grade accordingly.²⁴⁹

The natural history of spondylolisthesis is spontaneous stabilization. It is generally accepted that significant slip progression rarely occurs in adults.^190,196 Some controversy exists regarding slip progression in adolescents. Harris and Weinstein⁸⁴ studied youths with grade 3 or 4 slips in a long-term follow-up study and noted that there was a higher incidence of progression of the slip until skeletal maturity was reached. Saraste¹⁹⁰ and Seitsalo¹⁹⁶ had similarly large long-term observational studies that demonstrated that slip progression in youths and adults was small overall. Possible factors positively correlating with slip progression include degree of slip, degenerative disk disease at the level of slip, adolescent age, and ligamentous laxity that manifests as hypermobility on imaging (i.e., motion on flexion–extension views).

Treatment for an isthmic spondylolisthesis in a young patient is similar to that for the athlete with spondylolysis, as described in the previous section. Fusion surgery is generally considered in adolescents if the slip is grade 3 or greater. For degenerative spondylolisthesis, nonoperative management with a rehabilitation program similar to that described in the section on degenerative zygapophyseal joint and disk disease is appropriate because both are typical findings with a degenerative slip. Operative intervention with fusion is generally considered only for recalcitrant pain after an appropriate rehabilitation program, persistent radiculopathy, or progressive instability.

Other Spinal Fractures

Many other types of spinal fracture can occur, the most common of which are briefly discussed below. Many are secondary to trauma. Evidence-based guidelines have not yet been developed for the treatment of traumatic spine fractures. Current literature in this area is mainly of retrospective case series. Outcomes appear to be most dependent on the amount of neurologic injury at the time of injury, and on the time elapsed between injury and surgery if a neurologic injury exists.²³⁰

The three-column structural concept of Denis is the most common way to classify spinal fractures. This concept divides the spine into anterior, middle, and posterior columns. The anterior column is made up of the anterior longitudinal ligament and the anterior half of the vertebral body and disk. The middle column is made up of the posterior longitudinal ligament and the posterior half of the vertebral body and disk. The posterior column is made up of the rest of the bony and soft tissues of the spine. If two of the three columns are intact, the spine is stable and treatment with pain management and rehabilitation is generally indicated.

Cancer and Low Back Pain

Cancer is the second leading cause of death in the United States, and two thirds of patients with cancer develop metastases. The spine is the most common site for bony metastases, and vertebral body metastases are found in more than one third of cancer patients. The most common cancers that involve the spine are lung, breast, prostate, and renal cell.¹⁶⁷

Back pain is by far the most common symptom of metastatic disease. It is caused by stretching of the periosteum and tumor mass effect. The thoracic region of the spine is most commonly involved, although the lumbar spine is a more common site for colorectal cancer.¹⁷⁶ The pain can start gradually and increase as the bone is destroyed. It is a constant ache not exacerbated by movement. Sometimes the pain has a more sudden onset because of a pathologic fracture, and this type of pain can be worse with movement, especially if the spine is unstable. Deyo et al.⁵⁰ found that the most specific historical feature for malignancy as a source of back pain is a previous history of cancer (98% specific), and the authors considered it prudent to consider new-onset back pain in a patient with a history of cancer to be malignant disease until proven otherwise. This historical feature has a sensitivity of only 0.31, however, so only about one third of patients with spinal malignant neoplasm have a history of cancer. Consequently, other features suggestive of malignancy must be explored in the history. Back pain unrelieved by bed rest is greater than 90% sensitive, so if the pain is relieved by bed rest, malignancy is unlikely. This is not specific, however, so many patients without malignancy also complain that their pain is not relieved by bed rest. Other historical features related to cancer include unexplained weight loss and failure to improve with conservative care. New onset of back pain after age 50 is suspicious for malignancy because many other common causes of back pain begin at an earlier age. These features can be combined to give the clinician confidence in determining whether cancer should be included in the differential diagnosis.⁵⁰ Neurologic deficits occur in 5% to 10% of patients with spinal metastases either from the mechanical pressure of the tumor or from bone extruded from a collapsed vertebral body.²²⁰ Neurologic deficits often occur several months after the back pain began.⁷²MRI is the imaging modality of choice for a full evaluation of spinal metastases. It is very sensitive and can show early changes in the bone marrow. It also shows both bony destruction and neural compression.^72,176

Spinal Infections

Spinal infections include osteomyelitis, diskitis, pyogenic facet arthropathy, and epidural infections. These structures are often all infected at the same time. The incidence of spinal infections is increasing. Some of the causes of this include the growing numbers of immunocompromised patients who are at high risk, drug resistance of some infections, and recent increases in tuberculosis.²¹² It is important to diagnose and treat spinal infections quickly to prevent increased morbidity and mortality, and to prevent complications such as epidural abscesses that can cause paralysis.¹⁰⁵ However, this is not always easy. In Tali’s review article of spinal infection,²¹² he describes the “rule of 50” to assist in the diagnosis of spinal infections: 50% of the patients are older than 50, 50% will have a fever, and 50% will have a normal white blood count. The urinary tract is the source in 50%; Staphylococcus aureus is the organism in 50%; the lumbar spine is affected in 50%; and symptoms are present for greater than 3 months in 50%.²¹²

Vertebral osteomyelitis can occur from hematogenous spread or be secondary to a contiguous focus of infection. Hematogenous spread occurs via spinal arteries, and infection can quickly spread from the end plate of one vertebral body into the disk and then into the adjacent vertebral body. The most common source is urinary tract infections, often caused by Escherichia coli and other enteric bacilli. Hematogenous spread is also seen from other sources, such as infected intravenous lines or endocarditis. Patients with diabetes, those on hemodialysis, intravenous drug users, and other immunocompromised patients are at increased risk.^126,212 The most common location is the lumbar spine, and the most common symptom is back pain, although 15% of patients also have radicular pain. Symptoms can begin slowly and progress over months. Many patients do not have a fever or elevated white blood count. The erythrocyte sedimentation rate, however, is usually elevated. The infection can spread to surrounding tissues, and epidural, paraspinal, and psoas abscesses can also be present.

Plain films are usually normal the first 2 weeks, and then the first sign is a periosteal reaction. As the infection progresses, plain films show irregular erosions in the end plates of adjacent vertebral bodies and narrowing of the disk space. This appearance is nearly pathognomonic for infection because tumors and other causes of irregular erosions rarely cross the disk space. Bone scan shows changes as soon as 24 hours after symptoms begin but are not specific. MRI is as sensitive as bone scan and can give important anatomic information; therefore it is generally the imaging technique that should be used.^105,108

Treatment for spinal infections is usually a 4- to 6-week course of intravenous antibiotics. Sensitivity can often be determined by blood cultures, but if these are negative, samples from a bone biopsy might be necessary. Following the erythrocyte sedimentation rate is helpful to determine the effectiveness of treatment. Surgery is generally necessary only if the spine has become unstable, there are progressive neurologic deficits, or medical treatment fails. Spontaneous fusion of the infected segments often occurs after treatment.¹⁰⁵

Osteomyelitis secondary to a contiguous focus of infection is seen after surgical procedures and with extension of infection from adjacent soft tissue. The most common organism is S. aureus.^126,212 Risk factors for development of postoperative osteomyelitis include history of smoking, obesity, poor nutrition, uncontrolled diabetes, administration of steroids, history of malignancy, and radiation treatment in the area of surgery.¹⁰⁵ These infections usually present about 14 to 30 days after surgery.¹⁰⁵ Diagnosis is sometimes difficult because symptoms such as pain or fever can be attributed to soft tissue infection or the surgical procedure. Imaging studies can also be less conclusive because of surgical or soft tissue changes. The erythrocyte sedimentation rate is usually elevated after surgery, so it is not as useful in making the diagnosis in the first weeks after surgery.¹⁰⁵ Treatment for these types of infection usually requires surgical debridement and then a course of antibiotics.^126,212

Diskitis can occur from contiguous spread of infection or iatrogenically from procedures such as diskectomy and diskography. The incidence of these types of infection is low, as studies report a 0% to 3% incidence with procedures, but morbidity if infection occurs is significant. One study found that 55% to 87% of patients were unable to return to their normal work after diskitis. One reason for this poor outcome is the difficulty of using antibiotics to treat the infection because of the relative avascularity of disks.³¹

Spondyloarthropathies

Spondyloarthropathies are a group of diseases associated with the HLA-B27 allele. They include ankylosing spondylitis, Reiter syndrome, reactive arthritis, psoriatic arthritis, enteropathic arthritis, and undifferentiated spondyloarthropathy. It is hypothesized that, in genetically susceptible individuals, an interplay of environmental and immunologic factors leads to clinical manifestations. Although the diseases are grouped together, each has unique features on clinical presentation.¹⁰²

Lumbosacral Radiculopathy

Radicular symptoms can be the result of overt mechanical compression of a nerve root or a chemically mediated inflammatory process. The most common compressing lesion by far is a disk protrusion. Fewer than 1% of patients who present with radicular symptoms have other causes, including infection, malignancy, or fracture.⁴⁷ Rare presentations of radiculopathy, that is, those with fever, weight loss, night pain, cancer history, or osteoporosis risk factors, certainly warrant special attention to evaluate for the less common but potentially more catastrophic causes of radiculopathy.

The most common levels of disk herniation are L4–L5 and L5–S1, with L5 and S1 being the most common nerve roots involved in radiculopathy. Multiple nerve roots can be affected by a single disk herniation, given the organization of the cauda equina. Central disk herniations can also affect nerve rootlets that are descending in the cauda equina. The affected nerve root level might not correlate with the level of the disk herniation. For example, a central L3–L4 disk herniation could impact the L5 or S1 rootlets as they descend through the thecal sac before exiting out of their expected neural foramen. True cauda equina syndrome occurs when the lowest sacral rootlets are affected, resulting in bowel, bladder, and sexual dysfunction. Up to 1% of all disk herniations present as cauda equina syndrome.¹⁹⁸ In the appropriate clinical setting, a large postvoid residual of urine is a good predictor of cauda equina syndrome.⁴⁷ Cauda equina syndrome is a surgical emergency. Recovery of neurologic deficits, including bowel and bladder dysfunction, is greatest if decompressive surgery is performed within 48 hours.²⁴⁰

The natural history of lumbosacral radiculopathy resulting from disk herniation tends to favor spontaneous resolution of symptoms over time.³⁶ Multiple reports have shown that disk protrusions and extrusions can regress without surgical treatment.²¹⁵ Conservative treatment is best used to decrease pain and improve the patient’s level of functioning during acute management of radiculopathy. Even with some neurologic injury, conservative management should be considered because various studies have documented the same neurologic recovery in groups treated surgically and nonsurgically.²⁴⁴

The specifics of conservative management of lumbosacral radiculopathies, however, are still debatable. NSAIDs have not been found to be effective in patients with radiculopathy.^43,227 No definite support exists for oral steroids in the treatment of acute radiculopathy.⁴² The neuropathic pain agents (anticonvulsants and tricyclic antidepressants) are often used for radicular pain.⁴² Small studies have found gabapentin and topiramate to be associated with small improvements in pain scores.⁴³

Although exercise therapy has not specifically been demonstrated to alter the course of acute radiculopathy, there is probably a role for exercise.²²³ Saal et al.¹⁸⁶ reported very favorable outcomes using aggressive nonoperative care (an active exercise program potentially with epidural steroid injections) in the treatment of lumbar disk herniation with radiculopathy. Their protocol is the basis for many exercise programs used in the treatment of lumbar disk herniations with radiculopathy today.

Lumbar epidural steroid injections have become a common adjuvant for the treatment of lumbosacral radiculopathy. The more recent literature supports the use of fluoroscopically guided transforaminal epidural injections for early pain relief and potentially a more rapid recovery and reduced need for surgical intervention.^{101,120,179,221} They are best used in combination with an active rehabilitation program and are commonly used to facilitate active therapy by decreasing pain and inflammation.

Surgical management of lumbosacral radiculopathy is best reserved for those patients who have significant persistent radicular symptoms despite 6 to 8 weeks of maximized conservative management, neurologic progression, or cauda equina syndrome. Common decompressive procedures with favorable outcomes include lumbar hemilaminotomy with diskectomy, and lumbar hemilaminectomy.²⁰⁸ The highly anticipated Spine Patient Outcome Research Trial was a randomized trial evaluating the effect of surgery over individualized nonoperative treatment involving 13 spine centers across the United States with 501 surgical candidates with lumbar radicular symptoms for at least 6 weeks caused by a disk herniation.²⁴⁷ Subjects randomized to either surgery or nonoperative treatment improved substantially over the 2-year period of the study, and the primary outcomes were not statistically different between groups. A secondary outcome of sciatica severity, however, did show statistical significance in favor of surgery. Patients need to be counseled regarding appropriate expectations after surgery for lumbar disk herniation with radiculopathy. Surgery might slightly accelerate the resolution of neurologic deficits for the typical radiculopathy; however, the major benefit of surgical intervention is pain relief.⁴⁹ Relief of leg pain should be expected; however, back pain relief is more difficult to predict. Patients should be counseled that they are likely to have recurrent back difficulties even after a successful decompressive surgery. Figure 40-25 shows a useful algorithmic approach to the management of acute lumbosacral radiculopathy.

FIGURE 40-25 Algorithmic approach to acute lumbosacral radiculopathy (without cauda equina).

(Modified from Chiodo A, Haig AJ: Lumbosacral radiculopathies: Conservative approaches to management, Phys Med Rehabil Clin N Am 13:vii, 609-621, 2002.)

Lumbar Spinal Stenosis

The symptoms of spinal stenosis result from a complex series of changes within the lumbar spine.⁷¹ These changes are generally related to aging. The narrowing of the spinal canal that occurs in stenosis results from the degenerative changes described by Kirkaldy-Willis et al.¹⁰⁸ Not all patients with significant narrowing, however, have symptoms. Vascular and biochemical factors are probably involved that add to the mechanical compression (resulting from narrowing of the canal), which ultimately leads to symptomatic spinal stenosis. These have been described earlier in this chapter. Box 40-7 gives a classification schema for spinal stenosis, and Table 40-7 outlines a radiologic grading scale. Electrodiagnostic studies are often performed in patients with spinal stenosis. In addition to ruling out other reasons for leg symptoms (such as peripheral neuropathy), they can be helpful in fully characterizing the stenosis. In one recent study, H reflexes and F waves were shown to correlate with anatomic changes on MRI of spinal stenosis, whereas limb electromyography did not.⁴² This is consistent with the clinical observation that MRI findings and radicular complaints often do not correlate.

Table 40-7 Grading Lumbar Stenosis on Magnetic Resonance Imaging

The variable symptoms of spinal stenosis are due to the fact that a single nerve or multiple nerve roots can be affected at one or multiple locations within the lumbar spine. Mechanical compression of the nerves can occur as a result of central canal narrowing, lateral recess narrowing, and intervertebral foraminal narrowing. This results in variable symptoms, from a monoradiculopathy to polyradiculopathy to the hallmark symptoms of neurogenic claudication.

Neurologic claudication is the most common presenting symptom of lumbar stenosis and results from central canal narrowing. It is classically described as bilateral leg pain initiated by walking, prolonged standing, and walking downhill (relative lumbar extension). It is typically relieved by sitting or bending forward. If foraminal or lateral recess stenosis is the primary pathologic issue, then patients can present with more standard radicular pain in the typical dermatomal distribution. Most patients default to a forward-flexed posture to widen the central canal, subsequently decreasing mechanical compression of the nerve roots. This can lead to significant hip flexion contractures.

The natural history of lumbar spinal stenosis is fairly favorable overall. Johnsson et al.⁹⁸ followed patients with lumbar stenosis over a 4-year period with conservative treatment. Based on subjective patient reports, 70% remained unchanged, 15% improved, and 15% worsened. Walking capacity improved in 42% of patients, remained unchanged in 32%, and decreased in 26%. Amundsen et al.⁵ reported on a 10-year study of patients randomly assigned to surgical or nonsurgical treatment. Nonsurgical treatment consisted of bracing for 1 month followed by physical therapy. They demonstrated that neurologic deterioration was rare, that delaying surgery (with conservative management) had no effect on postoperative outcomes, and that, at 4 years, half the conservative treatment group and four fifths of the surgical group had favorable outcomes. During the final 6 years of study, clinical deterioration of symptoms was rare. In general, most patients with symptomatic stenosis remain unchanged, whereas some improved and others worsened. It is impossible to predict which patients will fall into each of these categories. It is useful information that a diagnosis of lumbar stenosis does not mean rapid neurologic deterioration, and that conservative management for those with mild to moderate symptoms is warranted.

The primary goals of conservative management are pain control and reducing the functional limitations that result from the lessened activity and the pain of stenosis. Multiple facets are used in this management program, including oral medications, epidural steroids, and a comprehensive functional rehabilitation program. The oral medications are no different than what have been described previously for treatment of radiculopathy. Even more attention needs to be placed on side effects, however, because most patients with lumbar stenosis are elderly and potentially have multiple medical problems that already require multiple medications.

Botwin et al.²⁸ have demonstrated that there probably is a role for epidural steroid injections in the nonoperative treatment of symptomatic lumbar stenosis. They performed transforaminal epidural steroid injections in stenosis patients who were deemed surgical candidates. Patients also received oral medications and physical therapy. Even at 1-year follow-up evaluation, 64% of the patients felt subjectively better. Only 17% of patients went on to surgery within the 1-year follow-up period.

Although there is a paucity of studies examining specific rehabilitation protocols, there is certainly a role for a therapeutic exercise program in the management of lumbar spinal stenosis. The basis of any protocol should be flexion-based lumbar stabilization exercises. This includes strengthening the abdominals and pelvic girdle stabilizers, including the gluteals. Improving hip mobility through stretching, especially of the anterior muscles (iliopsoas and rectus femoris), is also key. Aerobic conditioning is the final component of a comprehensive rehabilitation program for stenosis. Bracing with an abdominal corset might be beneficial for overweight patients with a protuberant abdomen to lessen the forces creating an exaggerated lordotic posture. However, there is a lack of clear data supporting the role of lumbosacral orthoses in spinal stenosis.

Surgical consideration for lumbar stenosis should be given to patients with intractable pain resistant to nonoperative management, profound or progressive neurologic deficit, or lifestyle impairment. A recent study comparing surgical management of spinal stenosis versus “usual care” showed that by 3 months postsurgery, the surgical group had better pain control than the conservative care group and that this improvement in pain persisted for years. No improvement, however, was seen in function in the surgical group compared with the conservative care group.²⁴⁸ Of note, this was not a placebo-controlled study, and other studies involving surgery have shown profound placebo effects with surgery.¹⁸² Age is not a contraindication to surgery, although the patient’s general health status must be considered.^10,88 Laminectomies are the most common decompressive procedures.¹⁸² When spinal stenosis is associated with instability, degenerative spondylolisthesis, deformity, or recurrent stenosis, fusion is often performed. Instrumentation often improves the fusion rate but does not influence the clinical outcome.¹⁹⁷ If selectively chosen, most patients with neurogenic claudication do well with surgical management. If the chief symptomatic complaint is axial low back pain, however, the surgical outcome is generally poorer.¹⁰³

Nonlumbar Spine Causes of “Radicular” Leg Symptoms

A number of nonspinal disorders mimic lumbar radiculopathy because they generate pain referral patterns similar to lumbosacral dermatomes. Their etiology is diverse and includes joint, soft tissue, vascular, and peripheral nerve sources. A thorough history and physical examination can typically help differentiate these disorders from lumbosacral radiculopathy; however, other diagnostic studies might be necessary.

Low Back Pain in Pregnancy

Low back pain is a common problem in pregnancy. Patients are generally divided into two categories: those with low back pain and those with pelvic girdle pain (pain below the iliac crest, such as sacroiliac joint-related pain). Multiple studies have estimated the prevalence of low back pain in pregnancy at 49% to 76%.^{58,112,162,163,242} Risk factors include a history of previous back pain, previous pregnancy-related back pain, and low back pain during menses.^33,242 Low back pain can begin at any time during the pregnancy and generally reaches a peak at 36 weeks.^112,163,242 Pain decreases after this point and in most patients is substantially improved by 3 months postpartum.¹⁶³

A small group of patients will have persistent pain even after postpartum recovery. Risk factors for persistent back pain after pregnancy include having both low back pain and pelvic girdle pain, pain occurring in early pregnancy, weakness of back extensors, older patients, and those with work dissatisfaction.⁷⁹

The etiology of low back pain in the pregnant woman is hypothesized to be due to increased biomechanical strain or to an altered hormonal influence. The biomechanical alterations are due to changes in spine posture related to the anterior movement of the pregnant woman’s center of gravity. An argument against purely biomechanical factors as the primary cause, however, is that the back pain often starts before significant weight gain by the mother, and the incidence does not parallel the weight gain.³⁴ A hormonal influence probably exists in the etiology of low back pain in pregnancy. This etiology’s hypothesis is that the hormonal changes during pregnancy alter the lumbopelvic ligaments, which influences the stability of the lumbosacral spine and makes it more vulnerable to loading.¹¹² A direct correlation between circulating levels of the hormone relaxin and pelvic and back pain, however, is controversial.^4,82,113,123 The prevalence of disk abnormalities on MRI is the same for pregnant and nonpregnant women, so this might be a source of pain for some pregnant women.²⁴⁶ Only a few high-quality studies have evaluated therapeutic interventions in pregnancy-related low back pain, and there is not much evidence on which to base recommendations for management.²⁰⁹ Individualized physical therapy, water aerobics, acupuncture, and massage therapy can be recommended to decrease pain.^{61,107,156,245} Instruction on a home exercise program, use of a sacroiliac belt, and back school have not been shown to significantly decrease pain intensity.^{54,129,142,155} No data support the use of lumbar–abdominal orthoses that are designed to support the pregnant woman’s abdomen. A full discussion regarding the use of medication during pregnancy is beyond the scope of this chapter. In general, any medication use should be discussed with the patient’s obstetrician. Even medications generally thought of as safe and well tolerated can have unexpected consequences during pregnancy. For example, the use of NSAIDs in late pregnancy can cause premature closure of the ductus arteriosus and neonatal renal failure.¹⁶⁹ Most antidepressants have not been approved for use during pregnancy, and for most antiseizure medications, such as gabapentin, there is evidence for increased incidence of birth defects in animals, and they have not been well studied in humans.¹⁰⁰

Pediatric Low Back Pain

In the past back pain in the pediatric population has traditionally been considered to be relatively rare, and when present raised a concern for serious pathology. This belief is now known to be untrue. In a subset of studies involving more than 300 children each, the prevalence of back pain was cited between 30% and 51%.^15,74 Severe back pain, which is either relapsing or permanent, was reported in 3% to 15%.¹⁵ An increase in back pain prevalence is noted as the child ages. In a Finnish cohort study, back pain prevalence was reported as 1% in children 7 years old, 6% in those 10 years old, and 18% when 14 years old.²¹¹ Another reported the prevalence at 12% for 11-year-old children and 50% for those 15 to 18 years old, which approaches the adult prevalence.³⁵ The same study reported that the pain is often recurrent but that the experience of back pain is frequently forgotten. Other studies demonstrate that the prevalence of low back pain has the greatest increase during puberty and the time of the maximum growth spurt.^114,232 Risk factors for nonspecific low back pain in the pediatric population include increase in age, female gender, parents with low back pain, hyperlordotic posture, history of spinal trauma, participation in competitive sports, a high level of physical activity, and depression.¹⁵ The literature does not support the following as risk factors for pediatric back pain: being overweight, hamstring tightness, a low level of physical activity, and poor school performance.¹⁵ Sitting appears to be the main exacerbating factor of low back pain in the pediatric population.¹⁵ There also appears to be a positive correlation between low back pain in adolescence and the presence of pain as an adult.⁸³

There has been more recent attention focused on the role of backpack use in the development of pediatric low back pain; however, there is still not a proven correlation. Carrying a backpack greater than 7.5% to 15% of the wearer’s body weight increases the metabolic demands over what is required to move a person’s body weight alone.^122,180 The general recommendation for a child’s backpack weight is limited to 10% of body weight.¹²² This limit is based on concerns of increasing metabolic costs and not on the risk of back pain development (there are conflicting reports in the literature regarding backpack weight and back pain).^{80,122,219,243} Many new backpack designs improve ergonomic fit; however, there are no studies demonstrating their effectiveness in reducing back pain.¹²²

Some of the specific causes of low back pain in the pediatric population are listed in Box 40-8. Spondylolysis and isthmic spondylolisthesis often present in young athletes with back pain and have been reported as the most common underlying cause of persistent low back pain among children and adolescents.¹⁴³ Most believe the etiology is from overuse, particularly during the growth spurt. The presence of isthmic defects in children in the Western world is between 2% and 7%, and as high as 30% in elite athletes.¹⁶¹

Scheuermann disease typically presents in the adolescent as painless exaggerated thoracic kyphosis. From a postural standpoint, the teenager typically presents with excessive thoracic kyphosis (which is demonstrated to be fixed in attempted hyperextension), with a compensatory lumbar hyperlordosis. Radiographic criteria for the diagnosis of Scheuermann disease include anterior wedging of at least three adjacent vertebrae, end-plate irregularities, Schmorl’s nodes, and disk space narrowing.⁷⁸ These findings are present equally in the adolescent population without back pain, but there is a higher prevalence of concomitant degenerative disk changes in those with pain.²¹⁶

The etiology of Scheuermann disease is uncertain. Some believe that it is due to repetitive loading of the immature spine that might have some preexisting abnormality of the cartilaginous end plate.⁹³ There does appear to be a familial link.¹⁴¹ Scheuermann disease can have a benign course, although some untreated patients develop progressive structural kyphosis. Brace wearing is recommended until skeletal maturity is reached to help prevent the progressive kyphosis.

Idiopathic scoliosis is not generally painful. If associated with pain, often there is more serious underlying pathology such as tumor, infection, or spondylolisthesis. The curve direction in idiopathic scoliosis is typically right thoracic and left lumbar. If an atypical curve is encountered, further evaluation beyond plain films is generally warranted.

Neoplastic disease of the pediatric spine is fortunately rare. Most pediatric spinal tumors are primary (not metastatic) benign bone tumors arising from the vertebrae.⁹³ The most common tumors of the pediatric spine include osteoid osteoma, osteoblastoma, and aneurysmal bone cysts. The classic pain of osteoid osteoma is nocturnal pain that responds to aspirin. The most frequent malignant lesion affecting the pediatric spine is Ewing’s sarcoma.