Cervical spondylosis is a progressive degenerative process resulting in pathologic changes in the intervertebral discs and surrounding structures ¹. Such changes include intervertebral disc protrusion, osteophyte formation, and hypertrophy of the lamina, ligaments, and hypophyseal joints.^2,³ Clinical onset can begin as early as the third decade, with progression continuing into the eighth decade.¹ Radiographic signs of spondylosis are evident in 50% of the population by the fifth decade, while prevalence is estimated at 98% for people over 70.⁴ Secondary myelopathy is considered to be the most common cause of spinal cord dysfunction in patients over 55.⁵

The effects of spondylosis can range from subclinical symptoms to primary root impingement (radiculopathy) and myelopathy. Proper diagnosis depends on meticulous history-taking, thorough physical examination, and the use of appropriate imaging, neurophysiologic, and laboratory tests. The natural history of cervical spondylosis is not well understood ⁶, and treatment of spondylosis includes both conservative medical management and surgical intervention. A variety of surgical procedures have been well described including anterior and posterior approaches with or without fusion. This chapter reviews the relevant anatomy, pathophysiology, symptomatology, diagnosis, natural history, and management of cervical spondylosis.

Regional Anatomy of the Cervical Spine

Normal anatomy of the cervical spine consists of vertebrae, intervertebral discs, ligaments and joints, neural elements, and surrounding soft tissue and vascular structures.

Osseous Components

The bony cervical spine is composed of seven vertebrae (Figure 24-1A-C). The lower five segments (C3-C7) are similar in morphology, while the first two segments (C1 and C2) are anatomically distinct. The first cervical segment, the atlas (C1), is ring-shaped and articulates primarily with the occipital condyles above and the superior facets of the second cervical segment below. The second cervical segment, the axis (C2), has a cone-shaped projection (the odontoid process) that articulates with the anterior arch of C1. The remaining cervical vertebrae (C3-C7) share similar architecture; the vertebral bodies are roughly cylindrical in shape and increase in size from rostral to caudal. From each body projects an uncinate process superiorly, which indents the posterolateral margins of its respective intervertebral disc. The transverse processes project anterolaterally and house the foramen through which the vertebral arteries pass. The spinal cord runs through the spinal canal, which is formed by the posterior elements including the pedicles, the facet joints, the lamina, and the spinous process. At each level, nerve roots exit the canal between vertically adjacent pedicles (Figure 24-1C,D)

FIGURE 24-1 A, A lateral view of the cervical spine demonstrates the relationship of the cervical facets with the angled superior and inferior articulating facets and their encasing articular capsules. The anterior intervertebral discs and bodies are also seen. The spinous processes and dorsal arches are significantly stabilized by the dorsal interspinous and supraspinous ligaments. B, A posterior view of the cervical spine demonstrates the complex arrangement of the dorsal stabilizing musculoligamentous complex and its attendant musculature. C, An anterior view of the cervical spine demonstrates the anterior longitudinal ligament connecting the vertebral bodies at the level of the discs as well as the intertransverse ligaments between the transverse processes. The relationship of the vertebral arteries coursing rostrocaudally through the foramina transversaria is also shown. D, An axial view of a cervical segment shows the spinal canal containing the neural elements which are stabilized and held in place within the subdural space by the dentate ligaments. The nerve roots with their dorsal root, ventral roots, and spinal ganglia are shown as well. The dorsal arch of the ligamentum flavum lies directly below the neural elements. Again shown are the vertebral arteries within the foramina transversaria with their relationship to the nerve root and the uncinate joint.

The unique morphology of C1 and C2, as well as the increasing vertebral body size of the lower segments, may contribute to the pathogenesis of spondylosis. The anteroposterior diameter of the canal is generally larger at C1 and C2 when compared to the lower vertebrae; thus the spinal cord is estimated to occupy only one third of the atlantal ring at C1, while it occupies up to three fourths of the canal in the lower segments. This variability may account for the predisposition for spinal stenosis with symptomatology at the C4-C7 levels.⁷

Intervertebral Discs

The vertebrae are linked together by intervertebral discs, which provide a stable yet flexible architecture. The discs normally account for 22% of the height of the spinal column and function to spread axial loading forces. Each disc is composed of four elements: the nucleus pulposus, the anulus fibrosus, and the two cartilaginous endplates. The nucleus pulposus accounts for 40% of the cross-sectional area of the intervertebral disc and is located toward the posterior aspect of the disc. It consists of primarily of Type II collagen fibers and mucopolysaccharide ground substance. The anulus fibrosus is made of fibrocartilage that surrounds the nucleus pulposus in a concentric lamellar fashion. It is attached directly to the vertebral bodies via Sharpey fibers, which project from the outer lamella to the epiphyses. The anulus also attaches to the anterior and posterior longitudinal ligaments, but the posterior attachment weakens with age, which may lead to a predisposition toward central posterior disc protrusions. The endplates are cartilaginous structures situated between the nucleus pulposus and the trabecular bone of the vertebral bodies. Perforating fibers anchor the endplates to the nucleus and serve as a conduit for nutrients to enter the disc.

Ligaments and Joints

The cervical spine allows for greater mobility than the thoracic or lumbar regions. Various ligaments reinforce the cervical spine during flexion, extension, and rotational movement. The anterior and posterior longitudinal ligaments extend the entire length of the spine and provide stability to the intervertebral joints. The anterior longitudinal ligament attaches to the ventral aspect of the vertebral column and is apposed to the intervertebral discs, while the posterior longitudinal ligament courses along the dorsal aspect of the vertebral column and merges fibers with the annulus as well as the adjacent endplates (see Figure 24-1C). The ligamentum flavum attaches to the anterior surface of each vertebral arch and to the superior edge of each lamina, and covers each facet joint. Its function is to stabilize the neck during flexion (see Figure 24-1D). Due to its considerable elastic capacity, the ligamentum flavum can normally stretch during flexion without compromising the integrity of the spinal canal. Additional ligaments such as the supraspinous and interspinous ligaments serve see further stabilize the spinal column (see Figure 24-1B).

The superior and inferior facets comprise the articular pillars (see Figure 24-1A). In the cervix, these joints are angled obliquely and inferiorly and are oriented perpendicular to the vertebral bodies. Each superior facet articulates anteriorly to its inferior process, forming synovial joints, each with a fibrous capsule.

Vascular Supply

The cervical cord’s vascularity is provided primarily by the paired vertebral arteries, which arise from the subclavian arteries in the thorax and course superiorly to form the basilar artery at the base of the pons. Branches from the vertebral arteries form the anterior spinal artery, which courses along the ventral median fissure of the spinal cord, and is responsible for supplying the anterior two thirds of the cord. The vertebral arteries also give rise to the paired, posterior spinal arteries, which travel along the dorsal aspect of the cord (see Figure 24-1D).⁸ Cervical medullary arteries branching from the vertebral arteries create variable anastomoses within the canal and contribute additional vascular supplies.

Pathophysiology of Cervical Spondylosis

Cervical spondylosis is a multifaceted degenerative process that can affect all components of the spine including the intervertebral discs, facet joints, ligaments, spinal soft tissues, and bony elements. The initial pathological changes in spondylosis originate in the disc space. The proposed mechanism of disc degeneration occurs secondary to an alteration in the protein composition of the disc matrix. With aging, the molecular weight of the glycoproteins in the disc decreases along with the chondroitin sulfate content. The net effect is a change in the osmotic properties of the disc, with decreased inflow of fluid. Dehydration of the disc leads to loss of height, as well as loss of expansion capability under axial loading. As the disc progressively loses the ability to distribute normal loads of pressure, the nucleus pulposus becomes predisposed to fragmentation. Fragmentation of the nucleus combined with increasing weakness in the annulus with aging can result in herniation of disc material into the spinal canal (Figure 24-2A).

FIGURE 24-2 A, A posterolateral disc herniation is shown in this spondylytic cervical spine segment, causing compression of the nerve root at the level of the foramen without compromise of the cord itself. When symptomatic, such lesions are typically associated primarily with radicular complaints. B, Multilevel spondylytic osteophytes combined with chronic disc herniations contribute to a severely narrowed central canal that can be associated with central hemorrhage and myelomalacia. Clinical syndromes include central cord syndrome and cervical myelopathy. C, A T2-weighted axial MR image of a cervical posterolateral disc herniation provides an example of root compression in cervical spondylosis. The cord, although displaced, is not significantly compressed. D, A T2-weighted sagittal MR image demonstrates multilevel spondylytic changes with disc herniations, ultimately causing significant canal stenosis and spinal cord compression. T2-signal changes within the parenchyma of the spinal cord demonstrate injury to the substance of the neural elements.

Loss of disc height not only contributes to disc herniation, but also results in osteophyte formation. As the annulus bulges under the impaired function of the damaged disc, the periosteum of the adjacent vertebral bodies undergoes reactive processes. Hyperostosis of the subperiosteal bone generates a spondylotic ridge or osteophyte, which can impinge on the ventral canal to cause cord compression. Spondylosis also leads to hyperostosis of the posterior elements. In the dorsolateral spinal column, decreased disc height causes pathological changes in the facets with destruction of the joints and resultant hypertrophy. Abnormal mobility of the spine contributes to osteophyte formation in the neural foramina, leading to peripheral nerve compression and radiculopathy. As spondylosis progresses, the ligamentum flavum becomes hypertrophic and loses elasticity. During hyperextension especially, the ligament tends to buckle inward, contributing additionally to canal compromise.¹⁰

The mechanism by which spondylosis leads to cord injury is not entirely understood; the pathophysiology is complicated by the common absence of symptoms in patients with radiographic evidence of significant disease.¹¹ Spondylosis with disc herniation, osteophyte formation, and ligament hypertrophy clearly reduces the anterior-posterior diameter of the spinal canal. As expected, patients with congenitally narrow canals are therefore more prone to be symptomatic. Studies have suggested that the normal canal diameter is 17 to 18 mm (C3 -C7) and that a reduction in axial diameter to 11 to 13 mm is more likely to lead to myelopathy.¹⁰ Abnormal cervical motion and instability following degenerative changes may exacerbate cord injury. With neck flexion, the cord may move against the ventral spondylotic ridges causing cord damage. Extension may lead to cord strangulation between the folded ligamentum flavum posteriorly and osteophytes or herniated disc material anteriorly (Figure 24-2B).

There is considerable debate as to whether cord injury is due to direct compression of neural structures or secondary to extrinsic compromise of vascular supply. The vascular ischemia theory was first proposed in 1954 by Brain.² Breig noted later that in cervical flexion, mechanical flattening of the spinal cord occurred, with consequent decreased patency of the anterior sulcal and transverse arteries.¹² Other authors have noted that anterior-posterior compression of the spinal cord in both pathologic and experimental studies resulted in stretching of the transverse vessels and terminations of the anterior spinal artery with ischemia of the anterior two thirds of the cord.^8,^13,^14,¹⁵ Clinically, Allen noted that the cervical spinal cord blanched during flexion in patients with spondylosis.¹⁶

Human pathologic studies of cervical spondylosis demonstrate that canal compromise with cord compression results in characteristic histological changes. Ono et al. found that compression of the cord is associated with extensive destruction of both gray and white matter, with consequent demyelinization.¹⁷ Interestingly, the areas of the cord most encroached upon tend to display histopathological evidence of severe infarction. Ogino et al. demonstrated an association between localized infarction of the gray matter and a decrease in the anterior-posterior canal ratio to below 20%.¹⁸ As cord injury secondary to spondylosis progresses, tissue destruction leads to gliosis, scarring, cystic degeneration, and neuronal cell loss.

In conclusion, the pathophysiology of spondylosis appears to be the result of several degenerative processes that occur in conjunction. The decrease in disc height leads to herniation, reactive hyperostosis with osteophyte formation, and hypertrophy of the ligaments. These events, along with abnormal cervical motion, compromise the cord within the canal. The mechanism of cord injury is still not well understood but appears to be related to impaired vascular supply leading to neuronal ischemia. As spondylosis becomes more severe, pathological changes of the cord become evident: demyelination, gliosis, cystic degeneration, and neuronal cell loss.

Clinical Presentation of Cervical Spondylosis

Cervical spondylosis can present with a variety of clinical syndromes. Pain may be localized to the neck or display a radicular pattern. Weakness can occur as a mixture of upper and lower motor neuron findings. Lower motor signs generally predominate at the level of the lesion, while upper motor findings are present at segments below. Atrophy and diminished reflexes are common in the involved upper extremity. Lower segmental involvement presents as hyperactive reflexes, increased tone, clonus, or (most commonly) abnormal gait. Sensory impairment is highly variable, with patchy sensory loss in both the upper and lower extremities occurring along three neural pathways. Pain and temperature sense are often affected contralateral to the lesion, due to spinothalamic tract fibers crossing at levels near their entrance into the canal. The posterior columns, which convey position and vibration sense, decussate in the brain stem and therefore are often affected ipsilateral to the lesion. Spondylosis can also affect the dorsal root as it enters the canal, resulting in impaired dermatomal sensation.

Myelopathy is a common and severe manifestation of cervical spondylosis.⁵ Symptoms may be slowly progressive and associated with intermittent periods of remission and exacerbation.²⁰ Clinical presentation generally consists of lower motor neuron involvement at the level of the lesion, with upper motor neuron signs at segments below. Upper extremity involvement is often unilateral, while that of the lower extremity is bilateral. Lower motor neuron findings include weakness and atrophy with progressive loss of dexterity, particularly at the level of the lesion. The lower extremities may demonstrate spasticity, clonus, hyperreflexia, or abnormal gait, with a positive Babinski sign. Sensory disturbances are poorly localized, generally affecting the lower extremities and trunk, while rarely involving cervical levels.²⁰ Bowel and bladder impairment are rare, but indicate poor prognosis.

Due to the complex symptomatology of cervical degenerative disease, Crandall et al. described five clinical syndromes to aid in clustering various findings.²¹

1. Transverse myelopathy: involvement of the corticospinal, spinothalamic, and dorsal column tracts, as described earlier.

2. Principally motor symptoms: corticospinal tract involvement with minimal sensory deficit.

3. B rown-Séquard syndrome: hemi-cord disease affecting ipsilateral motor strength, ipsilateral proprioception and vibration sense, and contralateral pain and temperature sense. Generally reflects an asymmetric narrowing of the canal.

4. Central cord syndrome: motor and sensory deficit predominant in the upper extremity.

5. Radiculopathy: direct root compression secondary to a herniated disc or spondylotic change.²² Patients typically present with sensory disturbances in a radicular pattern, specific motor group weakness, and decreased specific reflex. With chronic disease, profound weakness and atrophy may be present.

Among other possible findings is the “numb, clumsy hand.”²³ This condition involves a glove-like distribution of primary sensory loss combined with motor loss. Tandem spinal stenosis simultaneously affects the cervical and lumbar regions, presenting with a trio of symptoms: neurogenic claudication, gait abnormality, and mixed upper and lower motor neuron signs.²⁴ Vertebral artery insufficiency can present with dizziness and unsteadiness when the head is rotated.²⁵ Rarely, large osteophytes can cause dysphagia due to direct compression of the esophagus.⁵

It is important to consider other neurological conditions that may exhibit symptoms mimicking cervical spondylosis. Any mass lesion within the spinal canal that compresses the cord or nerve roots can manifest with such findings. Extradural, intradural, and osseous tumors of the spine, as well as infectious processes such as epidural abscesses, can compromise canal integrity. Fortunately, these conditions can generally be distinguished from cervical spondylosis via effective MRI. Multiple sclerosis is another condition commonly confused with cervical spondylosis, and a mixed picture of upper and lower motor neuron signs is a hallmark finding in amyotrophic lateral sclerosis. Correctly diagnosing cervical spondylosis depends on detailed history-taking, complete neurological examination, and diagnostic measures including imaging, neurophysiology, and laboratory tests.

Diagnostic Modalities

Neuroradiology

Plain film radiographs of the cervical spine are traditionally performed with a series of anteroposterior, lateral, and oblique films. Relevant findings on lateral films include the height of disc spaces and evidence of osteophytes protruding into the spinal canal. Also of import is the anteroposterior diameter of the canal, as it is highly indicative of disease severity in patients with symptomatic spondylosis.²² The diameter is determined to be the shortest distance from the dorsal aspect of the vertebral body (including any posteriorly projecting discs or spurs) to the spino laminar line, with 12 mm in the lower cervical region being the lowest normal value.²⁶ This calculation, however, is manipulated by the magnification of the film — an obstacle circumvented by an alternative method described as Pavlov’s ratio.²⁷ This number represents the ratio of the anteroposterior diameter of the spinal canal divided by the anteroposterior diameter of the corresponding vertebral body. A normal value is approximately 1, with values of 0.8 or less suggesting compression. This method allows quick appraisal of the integrity of the canal without being influenced by magnification.

Computed tomography allows for better assessment of the spinal canal than plain radiography.²⁶ CT axial plane images have been shown to provide an accurate estimate of the canal diameter while also differentiating laterally projecting osteophytes and midline calcifications (i.e., as seen in OPLL).²⁸ CT scans alone, however, poorly visualize the soft tissue structures within the spinal canal. With the addition of intrathecal contrast, CT myelography can allow for quantification of cord compression at every level. CT myelography has been effective in correlating symptomatic disease with cross-sectional area of the canal.²⁹

Magnetic resonance imaging is the most recent advancement in the radiographic evaluation of cervical spondylosis, offering the advantages of imaging in multiple planes and improved definition of neural and ligamentous elements. Disc herniations are readily demonstrated and often have associated signal changes (see Figure 24-2C). MRI also distinguishes cervical spondylosis from disease processes that mimic it clinically, such as tumors, epidural masses, demyelination, and syrinx. The complete neuraxis can also be easily imaged if necessary. In comparison to CT myelography, MRI is a safer, less invasive procedure, making MRI the procedure of choice for initial evaluation of radiculopathy or myelopathy.^28,^30,³¹

Unlike conventional x-ray technology, MRI allows for the demonstration of pathological processes within the spinal cord parenchyma. Intramedullary signal intensity changes have been noted at segments adjacent to areas of spondylotic compression ³² (see Figure 24-2D). In experimental models, histological confirmation of cord injury is found at levels demonstrating MRI signal change with maximum mechanical compression.³³ The cause of signal change is attributed to myelomalacia, gliosis, and edema.^32,^34,³⁵ Clinical data correlating the degree of signal change with outcome are confusing at best, but high-intensity lesions are thought to suggest poor prognosis.^34,^36,^37,³⁸

MRI still poses problems in diagnosing certain degenerative changes. Small, lateral osteophytes can be difficult to distinguish from lateral disc herniations.²⁸ Also, midline calcifications seen in ossification of the posterior longitudinal ligament (OPLL) may be poorly visualized. In addition to these limitations, the high incidence of degenerative abnormalities imaged in asymptomatic individuals also proves problematic. Teresi et al. found disc protrusions in 57% and spinal cord impingement in 26% of patients over 65 years of age when clinical evidence of cervical spondylosis was absent.³⁹ So although imaging techniques allow direct visualization of disease progression, determination of patient prognosis and indication for surgical intervention is not made by such modalities alone.

Neurophysiology

Neurophysiological evaluation of cervical spondylosis may prove a valuable supplement to other findings. Recent interest in these functional diagnostic modalities stems from the difficulty in interpreting common radiographic abnormalities in asymptomatic patients. Neurophysiological testing may also assist in predicting prognosis and measuring response to treatment. In evaluating cervical spondylosis, electromyography (EMG) allows differentiation of radiculopathy from neuropathy, and peripheral from central nerve entrapment.⁴⁰ EMG also helps localize affected nerve roots by demonstrating conduction abnormalities in muscles innervated by adjacent cervical segments. This technique may assist in preoperative determination of levels requiring decompression. Somatosensory evoked potentials (SSEPs) involve electrical stimulation of peripheral sensory nerves while recording evoked activity from either the spinal cord or sensory cortex. Clinical studies showing spondylotic involvement of the posterior columns suggest that SSEPs may help appraise the functional status of the sensory system. Leblhuber et al. found that dermatomal SSEPs were altered at levels corresponding to cervical segments with degenerative changes.⁴⁴ Yet these neurophysiologic and radiographic abnormalities were also found in asymptomatic patients. Although experimental models have found a temporal relationship between changes in SSEPs and the onset of neurological deficit,³³ the diagnostic value of SSEPs has been challenged through studies that found median and ulnar nerve abnormalities in only a small percentage of patients with symptomatic cervical spondylotic myelopathy.^42,⁴³

Cortical motor evoked potential (MEP) recording has been suggested as a more sensitive test of spinal cord dysfunction than SSEPs,⁴⁴ due to the predominance of motor findings in patients with cervical spondylosis.⁴⁴ MEP abnormalities may also be detected in patients before the onset of clinical symptoms.⁴³^–⁴⁵ In comparison, MEP can detect abnormalities in 84% of patients with radiographic cord compression, while SSEPs show dysfunction in only 25%.⁴⁶

The role of electrophysiological studies in diagnosing cervical spondylosis or predicting outcome is not entirely clear at this time. Cusick suggests that combining MEP and SSEP recordings allows for evaluation of long tract function of both ascending and descending white matter. These two tests provide insight into the integrity of two spinal cord areas often affected by spondylosis. By integrating electrophysiological studies and radiographs, patient vulnerability to neurological deficit may be estimated, as well as optimal timing of surgical intervention for patients with subclinical disease.⁴⁴

Natural History of Cervical Radiculopathy

The natural history of cervical spondylosis is not well described. Since early descriptions of the condition, surgery was widely accepted as the treatment of choice, and no studies were made to determine long-term progression. Only a few investigators have attempted to formulate a likely picture using patients treated with a collar.

Lees and Aldren-Turner classified cervical spondylosis as a relatively benign condition.¹⁹ The common course experienced by their patients was characterized by long periods of stable symptomatology interspersed with short bouts of deterioration; chronic, gradual deterioration was rare. They also found that myelopathy did not develop in a group of patients presenting solely with radiculopathy. In following studies, Nurick agreed with the findings of Lees and Aldren-Turner, and also found that patients who had undergone surgical laminectomy had no significant improvement over those with no treatment.³ Thirdly, he realized that age of onset served to significantly determine the prognosis for later deterioration.

When comparing 48 patients who underwent surgery with those in Lees and Aldren-Turner’s study, the other authors found that 70% of patients enjoyed improved conditions from cervical laminectomy. Their conclusion was that patients with moderate or severe symptoms due to cervical spondylosis benefit greatly from surgery, whereas those with mild disability are not likely to be significantly helped. Doubt was later cast upon Lees and Aldren-Turner’s theories in that they were thought to have bias toward milder cases.

Scoville found that the best outcome of surgery was in patients treated within one year of onset of symptoms.⁴⁸ He further elaborated on Lees and Aldren-Turner’s model in claiming that patients should be treated surgically, shortly after mild disability was noticed, and before further progression. He did admit that mild cases were adequately treated conservatively.

Smith and Robinson more recently described the course of cervical spondylosis that is most accepted today.⁴⁷ They found that motor complaints tended to be more permanent than neck, bladder, and sensory symptoms. Motor findings were also predominant in the lower extremities, and sensory in the upper. Although most of their patients followed an episodic but unpredictable pathway, one third of cases were found to be nonprogressive between acute episodes, while two thirds experienced a gradual increase in symptoms between intermittent, acute episodes of worsening. A minority of patients had a constant worsening in condition, and very few people enjoyed spontaneous improvement. Their conclusion was that although progression of the disease is usually slow, prognosis is poor, and improvement rare. They hypothesized that patients reporting improvement may simply be coping better, or may simply be reporting a slowing of progression.

Although many agree that there is a need to compare the outcome of different surgical treatments for cervical spondylosis with the natural history, such a study would be unethical. As surgery is widely accepted as incontrovertibly beneficial, it would prove difficult to randomize patients with severe or progressive disability to nontreatment.