Cervical Spondylotic Myelopathy

Pathophysiology

Cervical spondylosis refers to degenerative changes affecting the vertebrae, intervertebral disks, facets, and associated ligaments. Although these changes are most frequently seen with increasing age, other factors associated with an increased risk of spondylotic changes include repeated occupational trauma (e.g., carrying axial loads), genetic predisposition, smoking, and Down syndrome. Despite the widespread presence of cervical spondylosis in the aging spine, most patients do not develop myelopathy because a certain amount of narrowing is tolerated before cervical spinal cord compression becomes a risk. In an anatomic cadaveric study, Arnold noted the strong correlation between sagittal diameter of the spinal canal in cervical spondylosis and the production of myelopathy. The normal sagittal diameter of the spinal canal from C3 to C7 is 17 to 18 mm, whereas the diameter of the cervical spinal cord measures approximately 10 mm. This leaves a compensatory zone of approximately 7 to 8 mm in the anteroposterior dimension that buffers the spinal cord from moderate spondylotic narrowing. A sagittal canal diameter of less than 12 mm has been shown to be a risk factor for myelopathy in patients with cervical spondylosis. Similarly, patients with congenital spinal canal narrowing tolerate far less spondylotic narrowing before spinal cord compression occurs.

The cascade of cervical spine degeneration begins with the loss of integrity of the intervertebral disk. As disks age, the nucleus pulposus fragments, loses water, and collapses. As a result, the central annular lamellae buckle inward, and the external concentric bands of the annulus fibrosis bulge outward. This process leads to disk bulging and loss of disk height, which biomechanically translate to reduced load-bearing capability. In a large cadaveric study, Christe and colleagues described the histologic changes in aging cervical disks by illustrating variable degrees of cystic degeneration, lamellar disorganization, and radiating tears through the annulus fibrosis. These changes result in increased mechanical stresses at the end plates of the adjacent vertebral body.

Abnormal motions and forces may then lead to subperiosteal bone formation, which creates osteophytic bars that extend along the ventral aspect of the spinal canal. Osteophytic bars may have a biomechanically compensatory effect by stabilizing adjacent vertebrae by increasing the weight-bearing surface of the end plates, which are hypermobile as a result of lost disk material. However, osteophytic bars can also encroach on nervous tissue, thus causing compression of the spinal cord or nerve roots. Severe anterior spinal cord compression can also occur with ossification of the posterior longitudinal ligament (OPLL), observed predominantly in certain Asian populations. In addition, osteophytes may form at the uncovertebral joints and facets and may cause bony hypertrophy that frequently breaches the ventrolateral portion of the intervertebral foramina. Posteriorly in the spinal canal, the ligamentum flavum thickens as it loses tension and folds into the spinal canal as the intervertebral disks lose height. These static mechanical factors ultimately lead to circumferential narrowing of the spinal canal and potential compromise of the spinal cord, with resulting cervical myelopathy ( Fig. 14-1 ).

In addition to these static mechanical factors, myelopathy results from dynamic mechanical factors. Dynamic factors relate to the fact that normal flexion and extension of the neck can exacerbate existing spinal cord or nerve root irritation, especially in the presence of advanced cervical spondylosis, OPLL, or a congenitally narrow spinal canal. With neck flexion, the spinal cord lengthens and is stretched over ventral osteophytic spurs protruding into the spinal canal. During extension, the ligamentum flavum may buckle into the spinal cord and cause compression of the cord between the ligamentum flavum and the posterior margin of the vertebral body. Using a computerized simulation model based on the mechanical properties of white and gray matter, Ichihara and colleagues measured the stresses on the spinal cord in the presence of dynamic compressive factors. These investigators showed that, in patients with severe canal narrowing, repeated hyperextension of the cervical spine subjects the posterior spinal cord and, to a lesser degree, the anterior spinal cord to high shear stress comparable to that occurring in acute spinal cord injury. These episodic events may partly account for the neuropathologic changes seen in CSM.

Although the primary mechanism of CSM may be compression of nervous tissue by static and dynamic mechanical factors, growing evidence indicates that spinal cord ischemia may also play a role in CSM. Multiple studies have shown that the histopathologic changes seen in CSM are comparable to those observed in isolated spinal cord ischemia. Investigators have established that oligodendroglia are particularly susceptible to the effects of ischemia that contribute to the early demyelination of the corticospinal tracts. In an autopsy series, patients with varying degrees of spinal cord compression also showed damage to the lateral corticospinal tracts similar to that seen in spinal cord ischemia. Hakuda and Wilson published a study on the effects of anterior spinal cord compression and ischemia on the canine cervical spinal cord and found that the effects of vascular insufficiency and compression of the anterior spinal cord are additive, a finding that partially explains the clinical signs in patients with CSM. Most likely, ischemia occurs at the level of the impaired microcirculation and is probably caused by reduced flow in the pial plexuses, as well as by venous congestion and compression of larger vessels, such as the anterior spinal artery. Other possible factors in the pathophysiology of CSM include impairment of intracellular energy metabolism, free radical–mediated injury, apoptosis, and cation-mediated cell injury.

Natural History

Despite the high prevalence of CSM, few studies have addressed its natural history. Furthermore, consensus on the natural history of CSM remains elusive even with these published reports. Clark and Robinson published the first natural history study of CSM in 1956. Their review of 120 affected patients demonstrated that CSM caused slow, progressive motor deterioration that occurred with acute exacerbations rather than a steady, unrelenting neurologic decline. In their subset of patients who had no treatment, complete remission never occurred, and spontaneous regression of neurologic deficits was uncommon. Most of those affected were left with permanent but variable degrees of disability. Similarly, Lees and Turner reviewed 44 patients with clinical and radiographic evidence of CSM. More than half the patients were followed for more than 10 years. These investigators concluded that CSM follows a prolonged clinical course, in which an initial phase of deterioration is followed by lengthy periods of stable symptoms.

In the 1970s, Nurick confirmed these earlier reports with his retrospective review of 37 patients treated conservatively. In this study, patients were classified according to 6 grades of disability, based on the degree of difficulty encountered in walking ( Table 14-1 ). Nurick considered CSM a benign disorder in which old age was the only risk factor for progression of symptoms, and he believed that treatment should be reserved for this small subset of patients. Nurick concluded that no significant difference in outcome could be found in the majority of patients treated with surgical decompression as opposed to nonoperative care. In contrast, Symon and Lavendar argued that CSM is not a “benign” condition. In their surgical series, 67% of patients demonstrated relentless progression of neurologic deterioration without any static clinical period.

Like the series by Symon and Lavendar, some more recent studies have maintained that patients treated medically show continual progressive neurologic deterioration. These arguments have been used in support of early intervention for even mildly symptomatic patients. Along these lines, Sadasivian and associates conducted a retrospective evaluation of 22 patients, classified according to the Nurick grades; their findings suggested that the natural history of CSM is one of progressive deterioration with no stabilization of symptoms. These investigators noticed a significant mean delay from the onset of symptoms to diagnosis of 6.3 years, during which patients deteriorated on average by two Nurick grades. Bednarik and colleagues developed a predictive model to determine the risk factors for progression to myelopathy specifically in presymptomatic patients with spinal cord compression. These investigators concluded that electrophysiologic abnormalities of cervical spinal cord dysfunction, together with clinical signs of cervical radiculopathy and magnetic resonance imaging (MRI) hyperintensity, are useful predictors of early progression into CSM. In their retrospective analysis of 45 affected patients, their multivariate model predicted early progression into SCM in 81.4% of cases. Although these studies show merit in elucidating the natural history of CSM, they are nevertheless retrospective studies subject to observation bias and lack standardized outcome measures.

Prospective studies on the effectiveness of surgery in altering the natural history of CSM are few and inconclusive, thus highlighting the difficulty in performing controlled studies in these patients. In their Cochrane review of the role of surgery for CSM, Fouyas and colleagues conducted a meta-analysis to determine whether surgical treatment of cervical radiculopathy or myelopathy is associated with improved outcome as compared with conservative management. Most notably, these investigators reviewed a trial of 49 patients who had mild functional deficits associated with cervical myelopathy and in whom the effects of surgery were compared with those of conservative treatment. Although functional scores were better in the conservatively treated group at 6 months, at 2 years no significant differences were observed between the groups. Similarly, Kadanka and colleagues performed a 3-year prospective randomized study of surgical versus nonsurgical treatment of patients with mild and moderate degrees of CSM. The results showed no discernible difference between the two groups over a 3-year follow-up period. In their discussion, these investigators stated that the degree of success in the treatment of CSM with nonoperative approaches was similar to that observed with surgical intervention.

In contrast, excellent results for surgical management of CSM have been demonstrated in many studies. Sampath and colleagues published the results of a multicenter, prospective trial of 503 nonrandomized patients that found significant improvements in the functional status of patients undergoing surgical intervention in comparison with patients who were conservatively treated. Although the medically treated group was less symptomatic before assignment to treatment, surgically treated patients had better outcomes. Numerous uncontrolled reports have documented substantial improvement after surgical treatment, but well-conducted controlled studies that document the long-term benefit of surgical decompression are still lacking.

Signs and Symptoms

The clinical symptoms of CSM are often subtle and can be accompanied by signs that are difficult to elicit, thereby making its early diagnosis challenging. Furthermore, isolated myelopathy is relatively rare in patients with cervical spondylosis. Typically, the presentation includes nerve root impingement and radiculopathy (myeloradiculopathy). Besides neurogenic pain, patients may also complain of nonspecific neck stiffness or regional pain secondary to the presence of advanced spondylosis. The hallmark symptoms of CSM are gait abnormalities and weakness or stiffness of the legs. Additionally, patients may present with loss of manual dexterity and abnormal sensations manifesting with problems buttoning clothes or using a zipper or with poor penmanship. Loss of sphincter control and urinary incontinence are rare, but some patients may complain of urgency, frequency, and hesitancy. Confirmation of clinical suspicion is made by a thorough motor, sensory, and reflex examination of every patient, with special attention to signs suggestive of upper motor neuron dysfunction.

The most typical physical examination findings suggest upper motor neuron dysfunction. These include hyperactive deep tendon reflexes, ankle-patellar clonus, spasticity, the Babinski sign, and the Hoffman sign. The Hoffman sign is considered present when sudden extension of the distal interphalangeal joint of the middle finger causes reflexive flexion of the thumb or index finger. An inverted radial reflex occurs when tapping the distal brachioradialis tendon induces reflexive flexion of one or more of the fingers and diminishing of normal wrist extension; this finding is reportedly seen with C5 to C6 level spinal cord compression. The pectoralis muscle reflex can be elicited by tapping the pectoralis tendon in the deltopectoral groove, thus causing adduction and internal rotation of the shoulder if hyperactivity is present. This reflex suggests compression of the upper cervical spine (C2 to C4).

In the lower extremities, the Babinski response, described as the reflexive extension and abduction of the toes after gentle, sharp stimulation of the lateral aspect of the sole of the foot, also suggests upper motor neuron disease. Similarly, more than two beats of clonus after a rapid Achilles stretch is often pathologic as well. These pathologic reflexes, which result from a loss of the inhibitory function of the upper motor neurons, usually describe the spasticity associated with myelopathy. In patients with generalized spasticity, an increased jaw jerk reflex may differentiate an intracranial or metabolic disorder from suspected cervical myelopathy. Therefore, the presence of spasticity is not specific for myelopathy, although it helps corroborate the diagnosis of CSM and warrants further investigation.

Gait disturbance may be extremely subtle and is often the first physical symptom of CSM. Progression often occurs gradually and slowly over time. Early in the disease, the patient may have a subjective sensation of imbalance or subtle incoordination during turns or walking around corners. Other patients may complain of unsteadiness on uneven terrain or an inability to walk distances. Patients with more advanced CSM may demonstrate a stiff or spastic gait. On examination, myelopathic gait is often described as broad based and hesitant. Loss of motor coordination is further demonstrated by difficulty in maintaining balance with toe walking and heel walking. Measuring of walking times and of the number or steps taken over 30 minutes may be an objective, reproducible, and quantitative method of assessing the severity of gait dysfunction. Furthermore, a positive Romberg test is demonstrated by a loss of balance while the patient is standing with the eyes closed and the arms elevated in front of the body. Qualitatively, these findings can suggest the presence and severity of myelopathy.

In terms of motor examination in the upper extremities, patients with CSM most commonly exhibit triceps or hand intrinsic muscle weakness. Severe myelopathy can also result in atrophy of the intrinsic hand musculature. The finger escape sign may suggest weakness in the hand intrinsic muscles. To test for this, patients are asked to hold their fingers extended and adducted. If the ulnar digits drift into abduction and flexion, these patients have a positive finger escape sign. The patients’ hand function should also be assessed, in addition to the upper extremity strength assessment. A useful maneuver involves having the patient make and release a fist more than 20 times in 10 seconds. Impairment or clumsiness during this maneuver, the grip release sign, may suggest cervical cord dysfunction. In the lower extremities, distal strength is rarely affected without weakness in the more proximal muscle groups. Weakness is most frequently detected in the iliopsoas, followed by the quadriceps femoris. The finding of lower extremity weakness and hyperreflexia without upper extremity symptoms and signs should prompt an investigation of the thoracic spinal cord for other pathologic processes.

Sensory abnormalities in CSM have a variable pattern. Typically, symptoms start in the fingertips, are confined to the hand, and occur in a nonradicular distribution. Loss of vibratory sensation or proprioception in the extremities, particularly the feet, is sometimes the result of damage to the posterior columns and can be one of the early signs of myelopathy. This condition can be detected by tuning fork examination of the great toes or by testing proprioception of the toes or ankles. The Lhermitte sign, which consists of electrical shock–like sensations running down the back and shooting into the limbs during flexion of the neck, may also suggest dysfunction of the posterior columns. This sign is not specific for CSM, however. Spinothalamic sensory loss, when present, usually is asymmetric. The examiner must be aware of other medical comorbidities that can confound the sensory examination, such as diabetes mellitus or other metabolic causes of peripheral neuropathy.

Patients with CSM may also present acutely with a central spinal cord syndrome. This typically occurs when a patient experiences an acute hyperextension injury with preexisting stenosis (congenital or acquired) or myelopathy, resulting in acute spinal cord compression. The syndrome consists of greater upper extremity weakness than lower extremity weakness, varying degrees of sensory disturbance below the lesion, and myelopathic findings, such as spasticity and urinary retention.

Numerous scales are available for objectively grading the severity of CSM. Nurick developed a widely used grading scale based mainly on the degree of patients’ difficulty in ambulating (see Table 14-1 ). More recently, the Japanese Orthopedic Association (JOA) assessment scale gained popularity because it combines extremity motor function with other CSM symptoms including those related to sensory and bladder function. This scale is most widely used as the modified JOA (mJOA) score, as modified by Benzel and colleagues ( Table 14-2 ). The mJOA score also has been demonstrated to have a high interobserver and intraobserver reliability. Other scales used to measure outcomes in patients with CSM include the Cooper scale, Harsh scale, and Prolo scale. Additionally, general instruments that measure pain (e.g., visual analog scale) and functional status (e.g., the Medical Outcomes Study Short Form-36) are used. Most often, these scales and scores are used in research settings, rather than in everyday clinical practice.