Cervical spondylosis describes the degenerative change that affects the five articulations of the cervical segment, including the intervertebral discs, the bilateral zygapophyseal joints, and the uncovertebral joints of Luschka (30). The degenerative cascade is thought to begin with disc desiccation and loss of disc height, which is followed by an approximation of the uncovertebral joints and a disruption of the normal zygapophyseal joint biomechanics. Uncovertebral and zygapophyseal joint hypertrophy, osteophyte formation, anular disruption, and ligamentum flavum hypertrophy can comprise the ensuing phases of degenerative change (31, 32, 33). Radiographic evidence of degenerative change is appreciated in 10% of individuals by the age of 25, 35% by age 40, and in up to 95% by the age of 65 (34, 35, 36). Plain film imaging in asymptomatic subjects reveals degenerative change affecting the cervical spine in 70% of women and 95% of men between the ages of 60 and 65 (37). It is essential to remember that although there is typical radiographic evidence of degenerative disease, including loss of disc space height and age-dependent osteophytes, in individuals presenting with cervical pain, not
all individuals with degenerative findings are symptomatic (3, 37). The incidence of neck pain similarly increases with age, progressing in a fairly linear fashion from age 20 through 60 (3, 38). Approximately 95% of individuals will experience cervical pain by the age of 65 (35). Severe episodes of neck discomfort will affect 10% of the population at some point in their lives. The annual incidence of significant cervical pain has been estimated at 12.3% (36).

Referred pain is pain perceived in a region innervated by nerves other than those innervating the true source of the pain. Pain referral patterns arise as the brain is unable to decipher the true pain origin secondary to convergence at the level of the spinal cord and thalamus. Referred pain is typically experienced as deep, diffuse, and poorly localized pain (19). Any of the innervated structures of the cervical spine can contribute to local pain as well as symptom referral. Although injury to the cervical musculature and ligamentous structures can result in local and regional pain (39, 40), the intervertebral disc and zygapophyseal joints have been more extensively investigated as pain generators.

Degenerative disc disease is most commonly observed at C5-6, with C6-7 involvement the next most prevalent. These intervertebral discs might be more commonly affected because of increased segmental motion at these spinal segments (41, 42). An injured disc, resulting from degenerative change or more acute trauma, can produce local and referred symptoms (17). The outer third of the annulus fibrosus houses nerve endings that can be stimulated during injury. It has been suggested that degenerative or traumatic alteration of the internal architecture of the annulus can result in pain production through stimulation of local mechanoreceptors and nociceptors (17, 41). The ability of the annulus to mediate pain has been demonstrated during surgery through mechanical and electrical stimulation of cervical intervertebral discs (42). In addition to this mechanical component of disc injury, discogenic pain might also be biochemical in origin. An annular defect could allow for the migration of nuclear material, which might then stimulate the outer annulus, dura mater, posterior longitudinal ligament, dorsal root ganglion, or spinal nerve (43). Increased levels of inflammatory mediators have been identified in degenerative and herniated discs when compared with asymptomatic controls (44).

Using cervical discography, characteristic discogenic pain patterns have been described (42, 45, 46). The correlation between cervical disc magnetic resonance imaging (MRI) abnormalities and pain-generating potential observed through discography has been shown to be poor (46). In a study of 807 disc injections, 404 concordant pain responses were used to describe pain referral patterns. Pain referral to the scapular region was observed to arise from disc stimulation at the C3-4 through C7-T1 levels. The C5-6 and more caudal segments were noted to refer symptoms to the upper limb, and C6-7 stimulation was unique in that pain was referred to the anterior chest wall (45) (Fig. 32-5).

The zygapophyseal joints can also become active pain generators in the setting of degenerative change or following trauma.
These posterior elements are particularly vulnerable to injury during a whiplash event. Although often not observed radiographically, zygapophyseal joint fractures, intra-articular hemorrhage, and capsular tears have been observed in pathologic studies (47, 48, 49, 50, 51, 52, 53). The prevalence of whiplash-induced zygapophyseal joint pain has been described (54, 55, 56). Using diagnostic blocks, chronic cervical zygapophyseal joint pain has an estimated frequency of 54% to 60% (55, 56). In this chronic neck pain population, the C2-3 joint has been observed to be most frequently symptomatic, followed by C5-6 (55, 56). Of those patients with chronic neck pain, 58% to 88% describe significant headaches (54, 55, 56). The prevalence of C2-3 jointmediated headaches has been estimated at 50% to 53% in those individuals with a chief symptom of headache following whiplash (54, 56). The Oc-C1 joints (54, 57, 58, 59) and C1-2 joints (57, 58) have similarly been described as active nociceptors in cervicogenic headaches. Studies in a limited number of asymptomatic and symptomatic volunteers have suggested particular resultant pain patterns from intraarticular zygapophyseal joint stimulation and anesthetization (60, 61, 62). Pain arising from C2-3 stimulation was described as referred to the upper neck and occipital region, whereas C3-4 and C4-5 joint pain was observed to affect the mid- and lower cervical region, extending toward the superior scapular border. Pain from C5-6 stimulation was observed to extend toward the shoulder, whereas C6-7 pain radiated over the more caudal scapular region (Fig. 32-6). These degenerative or posttraumatic pain referral patterns arising from primary axial zygapophyseal joint or discogenic pain generators can be difficult to distinguish clinically from one another and from radicular pain presenting without a more distal component or neurologic correlate.

A randomized controlled study investigating the therapeutic benefit of intra-articular zygapophyseal joint injections using corticosteroid has suggested that such treatment is ineffective (63). A retrospective uncontrolled study of intra-articular C2-3 joint injection in the treatment of post-traumatic headaches has suggested a significant therapeutic response. A randomized placebo controlled trial of radiofrequency neurotomy in patients with unremitting cervical zygapophyseal joint pain has demonstrated a significant treatment effect (64). Pain is expected to return following neurotomy, as distal axons innervating the zygapophyseal joint are likely to regenerate.

Cervical radiculopathy can arise from pathologic compressive processes affecting the nerve root, including acute disc herniation, degenerative foraminal stenosis, trauma, or tumor (65). In the 1920s, cervical nerve root compression was described as a cause of anginoid pain (66). Cervical disc herniation resulting in cord compression was initially recognized as a syndrome of spinal cord tumors or chondromas. This syndrome of cord compression was subsequently defined by Mixter and Ayer in 1935 (67) as arising from a ruptured intervertebral disc. In 1936, cervical spondylosis and neural irritation was described as a cause of shoulder girdle and arm pain (68). In the early 1940s, cervical nerve root irritation arising from disc injury was defined (69).

The annual incidence of cervical radicular pain is 5.5/100,000 (70). These injuries represent 5.3% of all nerve root injuries caused by disc pathology. The nerve roots most commonly involved by cervical radiculopathy are C7 and C6, with most studies suggesting C7 syndromes as the most common, followed by radiculopathy of C5 and C8 (65, 71, 72, 73, 74). Patients younger than 55 years are more likely to present with radiculopathy arising from acute disc herniations, whereas those older than 55 years are more likely to demonstrate symptoms arising from degenerative foraminal or central canal stenosis (30).

The cervical nerve root can be compressed in the neural foramen by the intervertebral disc, degenerative change affecting the zygapophyseal or uncovertebral joint, or in a combined fashion. Cervical radiculopathy arising from a herniated disc is more common than radiculopathy arising from the more slowly evolving degenerative foraminal stenosis typically observed in older individuals (75, 76, 77, 78) (Fig. 32-7). Disc herniations have been classified as “soft” and “hard.” “Soft” refers to more acute disc injuries and displacement of nucleus pulposus
material, whereas “hard” disc material describes a calcified and spondylotic ridge (72, 79, 80). With disc degeneration, tears in the annulus can allow for the displacement of the nucleus. Anterior disc herniations are less common as the anterior longitudinal ligament is wider and stronger than the posterior longitudinal ligament (30).

Disc herniations have been characterized based on location (81, 82). The three locations described are intraforaminal, posterolateral, and central. Intraforaminal herniations are the most common and may result in an acute radiculopathy affecting the nerve root exiting through the respective foramen. Therefore, an intraforaminal herniation of disc material at the C4-5 level might result in a C5 radiculopathy. The posterolateral disc herniation occurs because the rhomboidal shape of the posterior longitudinal ligament tends to direct disc material to one side or the other. These herniations are located between the lateral edge of the posterior longitudinal ligament and the posterior aspect of the uncinate process. The central disc herniation passes through the substance of the posterior longitudinal ligament. Such lesions are more likely to result in central canal compromise and spinal cord compression. Central herniations are believed to be more likely to occur in the later stages of cervical degeneration when uncovertebral joint hypertrophy serves as a barrier to the more lateral migration of disc material (83) (Fig. 32-8).

The resultant injury to the nerve roots likely arises from combined mechanical and biochemical pathophysiologic processes. A study in which cervical disc specimens removed surgically from patients with radiculopathy were compared with discs from a traumatic control group found significantly increased levels of matrix metalloproteinase, nitric oxide, prostaglandins, and interleukins in the radiculopathy group (44). These findings, in conjunction with a plethora of literature describing the combined mechanochemical nature of lumbar radicular syndromes, suggest that these biochemical markers are likely involved in the injury process of cervical radiculopathy (84, 85, 86, 87, 88).

Plain films can be used as an initial screening tool but are not necessary in every individual presenting with cervical or radicular pain. In addition to anteroposterior and lateral images, flexion and extension views can prove of value in patients with recent trauma, suspected instability, or in cases of ankylosing spondylitis or RA (65).

MRI is considered the imaging modality of choice in cervical radiculopathy (89). Earlier comparative studies demonstrated a close correlation between findings on cervical MRI, computed tomography (CT) scan, and myelography (90, 91, 92). More recent studies have suggested a higher correlation between radiographic and surgical pathology when MRI was used than with either myelography (93) or CT (94). A high incidence of abnormalities has been observed on cervical MRI imaging of asymptomatic individuals. Boden et al. (95) observed significant abnormalities in 19% of 63 asymptomatic patients.
In those younger than age 40, 14% had a herniated disc or neural foraminal stenosis, and similar abnormalities were observed in 25% of individuals older than age 40 (83). An investigation of cervical myelography has identified nerve root filling defects in 21% of asymptomatic subjects (96).

As there is a significant prevalence of such abnormalities in asymptomatic subjects, advanced cervical spine imaging in symptomatic individuals must be interpreted carefully to determine if a correlation exists between radiographic pathology and clinical findings (97, 98). Also of interest is the natural radiographic history of symptomatic disc pathology in the cervical spine. A study following “soft” disc herniations in 21 patients with radiculopathy found that the largest lesions decreased in size the most, with 16 of 21 lesions reducing in size by 35% to 100% at 15-month follow-up CT (99). In similar studies observing patients with serial MRI, 40% to 92% of patients demonstrated radiographic regression (100, 101). More lateral and extruded discs were observed to be more likely to resolve (Fig. 32-9).

Electromyography (EMG) studies can help to localize the level of a cervical radiculopathy and differentiate such a presentation from a brachial plexopathy, more distal entrapment, or peripheral neuropathic process. Such testing is not necessary in all patients presenting with cervical radicular syndromes. Electrodiagnostic studies might prove particularly useful in individuals with multiple levels of radiographic pathology whose physical examination findings are less conclusive in identifying the segmental pathology of clinical significance. In other cases, a patient might present with clinical findings suggestive of radiculopathy but without a clear radiologic correlate. In these cases, electrodiagnostic studies might similarly prove useful. The earliest abnormality that might be observed in the setting of motor root compromise is reduced voluntary recruitment observed during needle exam. Abnormal activity at rest in the form of positive sharp waves or fibrillation potentials, indicative of axonal loss, might not be observed until at least 18 to 21 days after the onset of a radicular syndrome (102).
EMG abnormalities have been shown to correlate well with myelographic abnormalities and operative pathology (103). Needle examination of seven limb muscles is likely adequate for the identification of segmental pathology (104). The segmental sensitivity and specificity of EMG abnormalities might only approach 67% and 50%, and EMG abnormalities can be appreciated at asymptomatic levels (105). The sensitivity and specificity of EMG abnormalities have not been studied against an agreed-upon gold standard. The false-negative rate of electrodiagnostic studies is likely higher than desired (106).

A diagnosis of cervical radiculopathy is established in a patient when corroborative radiographic findings are observed in association with a suggestive pain distribution and/or a myotomal strength deficit, reflex abnormality, or sensory loss. If the cervical radiographs and clinical findings are less conclusive, electrodiagnostic studies might prove helpful in confirming a level of radicular pathology when a myotomal denervation pattern is observed. If radicular pain remains within the differential and these diagnostic criteria are not satisfied, a diagnostic selective nerve root block might be employed. Although the specificity and sensitivity of lumbar diagnostic injections have been described as ranging from 87% to 100% (107, 108), the utility of diagnostic selective cervical injections has not been as well-defined (109).

Patients with radiculopathy typically present with a chief report of pain, weakness, paresthesias, or a combination of sensorimotor deficits (65). The majority of patients describe cervical and upper-extremity pain that began without trauma or a particular inciting event (72, 73, 110). Most studies suggest that pain affects the upper limb more commonly than the neck, although most often both are involved (73, 79). Coughing, sneezing, or a Valsalva maneuver may lead to symptomatic worsening. The reports of pain might also incorporate the anterior chest wall, resulting in a syndrome of pseudoangina pectoris (67, 111, 112, 113). It is often difficult to determine the symptomatic level based on the patient’s pain description alone. The distribution of symptoms might not follow the classic dermatomal patterns as defined by Keegan and Garrett (114) or Foerster (115). The limitations of these mapping studies include the assignment of dermatomal distribution based predominantly on the loss of sensation or hyposensitivity arising from compressive injuries. Each of these classic studies describes considerable dermatomal overlap. Dermatomal pain patterns, arising from selective cervical spinal nerve stimulation, have recently been described and may be more representative of the pain and paresthesia patterns clinically observed in radiculopathy (116). Using fluoroscopically guided cervical nerve root stimulation, pain patterns were described after the stimulation of 134 roots in 87 subjects. Some of the highlights of this study included pain distribution distal to the elbow in only 14% of C5 stimulations, C6 affecting the ulnar hand in 67%, C7 uniquely resulting in anterior head symptoms and most commonly referring pain to the chest, and C8 affecting the thumb in 14%. No nerve root pain distribution was observed to be confined to the classically defined dermatomes in greater than 50% of stimulations (116) (Fig. 32-10). Some of the discrepancy between cervical dynatome and the more classic dermatomal maps might also be explained by intrathecal cervical dorsal root intersegmental anastomosis, which can be present in as many as 61% of individuals (117, 118, 119) (Fig. 32-11). Additionally, pain in radiculopathy might arise in part from a myotomal or myalgic pain component arising from ventral nerve root irritation (120).

When examining the patient with suspected cervical radiculopathy, it is essential that the clinician remains mindful of extraspinal disorders that can mimic nerve root pathology. A brachial plexopathy or neuralgic amyotrophy with upper-trunk involvement can appear similar to a radiculopathy of C5 or C6 origin (121, 122, 123). More localized disorders affecting the upper limb can similarly result in regional pain, including subacromial
bursitis, lateral epicondylitis, and bicipital tendinitis (65). More distal neural entrapments affecting the median, radial, or ulnar nerve should also be considered. A median neuropathy at the wrist can result in pain extending as proximal as the cervical region (65). When cervical radiculopathy and myelopathy coexist, such as in amyotrophic lateral sclerosis and syringomyelia, the examiner should seek signs of possible combined cord and root compression (65, 124).

Muscle girth should be carefully inspected for any evidence of atrophy that might be consistent with a more chronic radiculopathy. Although C5 and C6 involvements can lead to wasting of the biceps and periscapular musculature, loss of triceps girth can be observed in the setting of a C7 radiculopathy. Atrophy of the hand intrinsics might be observed in patients with C8 or T1 pathology. Active cervical range of motion can be recorded and observed for symptom provocation in a particular plane. In particular, extension and ipsilateral side bending are more likely to result in a reproduction of radicular symptoms. Normal range of motion of the neck, using an inclinometer, has been estimated at 60 degrees of flexion, 75 degrees of extension, and 45 and 80 degrees of bilateral side-bending and rotation (125).

Manual muscle testing has been suggested to be a more sensitive means of isolating a level of segmental pathology than either reflex or sensory testing. Variations in the contributions of the spinal nerves to the cords and branches of the brachial plexus can lead to an altered pattern of muscular innervation in the upper limb and perhaps more perplexing motor testing findings (65, 74). Reflex testing of the upper extremity should include the bicep, tricep, brachioradialis, and Hoffman’s response. Lower-extremity reflex testing should also be included to rule out hyperreflexia or long tract signs. Gait should similarly be observed to rule out ataxia, which might be observed in the setting of cervical cord compromise.

Spurling’s maneuver incorporates rotation and side-bending of the head toward the affected side, and this can be combined with axial compression. The goal of this maneuver is to diminish the foraminal area and reproduce radicular symptoms (30). This neck compression test was initially defined by Spurling and Scoville in the 1940s at a time when cervical radicular syndromes were first recognized (126). This maneuver has been demonstrated to have a high specificity but a low sensitivity (110). Such maneuvers should be performed with caution in patients with a suspected radicular syndrome, as this could lead to further neural irritation. In those patients demonstrating a reproduction of radicular symptoms with active cervical extension or ipsilateral side-bending, this provocative test should be avoided (65). The shoulder abduction relief sign has the patient place the affected hand on the top of the head. This position of upper-extremity abduction reduces stress on the nerve root affected by a disc herniation and relieves pain (127, 128).

A complete exam should also include passive ranging of the shoulder and impingement maneuvers. Tinel’s testing at the supraclavicular fossa, elbow, and wrist and carpal tunnel stress maneuvers, such as Phalen’s testing and carpal compression, might reveal a proximal entrapment neuropathy or carpal tunnel syndrome. Observing for a pain response during palpation of the medial and lateral epicondyle can be helpful in ruling out a contributing focal tendonitis. Thoracic outlet maneuvers should also be considered to rule out a contributing lower-trunk plexopathy or dynamic vascular insufficiency, also referred to as “neurogenic” or “vascular” thoracic outlet syndromes. These regional exam techniques are included in the complete examination of a patient with suspected radiculopathy and might enlighten the examiner to a contributing musculoskeletal disorder or more distal neuropathic process.

Lees and Turner studied 57 patients with a diagnosis of cervical radiculopathy for a period of up to 19 years (35). Patients in this study pursued a variety of treatments, including rest, manipulation, therapeutic exercise, and soft-collar use. Of these 57 patients, 45% described one primary pain episode without a future symptomatic recurrence. Of the remaining patients, 30% described a continuation of mild symptoms, and 25% reported a persistence of significant symptoms or worsening. Of the 57 patients, 32 remained significantly improved, whereas 18 described their condition as static or symptomatic. In their series of patients with radiculopathy, none became myelopathic. Cervical collars appeared to correlate with initial symptomatic relief, but end results were similar among all patients studied.

A review of multiple retrospective and uncontrolled nonsurgical cervical radiculopathy studies revealed satisfactory outcomes in 80% to 90% of patients (73, 129, 130). Several randomized controlled studies in patients with lumbar syndromes have demonstrated an efficacy of nonsteroidal anti-inflammatory agents and muscle relaxants. Studies investigating the efficacy of oral corticosteroids have been inconclusive (131, 132). In a retrospective study of 26 consecutive patients treated nonsurgically, 20 achieved good to excellent results with an average of 2.3 years of follow-up. Oral analgesics, collars, traction, and physical therapy were employed in this treatment group. Twenty-two patients used oral steroids, and nine received epidural injections. Patients with extrusions were noted to fair better than those with contained protrusions (133). In a randomized prospective study, 81 patients were treated with anterior decompressive surgery, physical therapy, or collar use. The surgical group demonstrated significantly greater reductions in pain at 16 weeks, and both the surgical and physical therapy groups demonstrated improved functional abilities as compared with the collar group. No significant difference in outcome as measured by pain and functional profiles was observed between the groups at the final 12-month follow-up evaluation (134). In a prospective study of 155 patients with cervical radiculopathy, one third were treated surgically, whereas two thirds were prescribed medical treatment that included oral analgesics and steroids, traction, and epidural injection. At 1-year follow-up, significant improvements in pain and functional status were observed in both groups, with a greater degree of improvement in the surgical group. Twenty-six percent of the surgical group continued to experience severe pain (135).