Cervical Spondylosis

Etiology

Cervical spondylosis originates from a multifactorial degenerative cascade whose dominant driver is age-related degeneration of the intervertebral disc and cervical spinal elements. Age-related loss of proteoglycans dehydrates the nucleus pulposus, lowers disc height, and shifts load transmission toward facet and uncovertebral joints, initiating osteophyte formation and ligamentous infolding. Degenerative changes in the spine structures, including the uncovertebral joints, facet joints, posterior longitudinal ligament, and ligamentum flavum, cause spinal canal and intervertebral foramina narrowing. Consequently, the spinal cord, spinal vasculature, and nerve roots can be compressed, resulting in the 3 clinical syndromes in which cervical spondylosis presents: axial neck pain, cervical myelopathy, and cervical radiculopathy.

Factors that can contribute to an accelerated disease process and early-onset cervical spondylosis include exposure to significant spinal trauma, a congenitally narrow vertebral canal, dystonic cerebral palsy affecting cervical musculature, and specific athletic activities like rugby, soccer, and horse riding.[4][5] Genetic susceptibility exerts a substantial influence on the tempo and pattern of degeneration. Lifestyle factors act mainly as disease accelerants. Cigarette smoking compromises vertebral microcirculation and increases oxidative stress, hastening disc desiccation.[6] 

Further, excess body mass increases compressive forces, while physical inactivity promotes static overloading; both conditions are correlated with faster structural failure of the disc–vertebra complex. Mendelian randomization further links lower educational attainment to higher cervical spondylosis risk through increased prevalence of smoking and obesity, underscoring the contribution of social determinants.[7] Occupational exposure to prolonged neck flexion, frequent overhead arm activity, and substantial loading of the upper extremities is associated with a higher rate of cervical spondylosis requiring surgical treatment among construction workers and other manual laborers.[8]

Epidemiology

Radiographic studies frequently demonstrate spondylotic changes in the cervical spine, yet the majority of affected individuals remain asymptomatic. Epidemiological data indicate these changes are present in approximately 25% of individuals younger than 40, 50% in those older than 40, and up to 85% of individuals older than 60. The most frequently affected levels are C6–C7, followed by C5–C6.

Symptomatic cervical spondylosis most commonly presents as neck pain. In the general population, the point prevalence of neck pain ranges from 0.4% to 41.5%, the 1-year incidence ranges from 4.8% to 79.5%, and lifetime prevalence may be as high as 86.8%. According to the Global Burden of Disease 2015, low back and neck pain remain the leading cause of years lived with disability and the fourth leading cause of disability-adjusted life years.[3][4][9]

Pathophysiology

The pathogenesis of cervical spondylosis involves a degenerative cascade that produces biomechanical changes in the cervical spine, manifesting as secondary compression of neural and vascular structures. An increase in the keratin-chondroitin ratio prompts changes to the proteoglycan matrix, resulting in loss of water, protein, and mucopolysaccharides within the intervertebral disc. Disc desiccation diminishes the elasticity of the nucleus pulposus as it dehydrates and becomes fibrotic. As the nucleus shrinks, axial loads shift to the annulus fibrosus and the uncovertebral and facet joints, prompting osteophyte formation and capsular hypertrophy. The continued collapse of the motion segment leads to buckling and thickening of both the posterior longitudinal ligament and the ligamentum flavum, which encroach upon the spinal canal and neural foramina, resulting in spinal stenosis.

With further dehydration, annular fibers lose mechanical integrity under compressive stress, altering load distribution throughout the cervical spine. The loss of disc space height reverses the normal lordotic curvature into kyphosis. This malalignment causes annular and Sharpey fibers to detach from the vertebral endplates, triggering a reactive bone formation response. Osteophytes subsequently arise along the ventral or dorsal margins of the cervical vertebrae and can project into the spinal canal and intervertebral foramina, worsening stenosis. Moreover, disrupted load balance increases axial stress on uncovertebral and facet joints, leading to joint hypertrophy and accelerated bony spur development within the neural foramina. Together, these degenerative changes reduce cervical lordosis and mobility, while narrowing the diameter of the spinal canal.[10][11]

Dynamic factors amplify injury. Extension narrows the canal, while flexion stretches an already compromised cord, especially in focal or segmental kyphosis. Repetitive shear stresses accelerate micro-motion at degenerated levels, worsening instability and canal encroachment. Radiculopathy arises when uncovertebral or facet osteophytes, disc protrusion, or hypertrophy of the foraminal ligament deform the nerve root. Root traction or compression disrupts axoplasmic flow, provokes venous congestion, and lowers intraradicular perfusion. Sensitising cytokines from annular tears or leaked nucleus pulposus further heighten nociceptor excitability and chronic pain.[12] Myelopathy reflects a combined static and dynamic compromise of the spinal cord. Ventral osteophytes flatten the anterior cord, while ligamentum flavum infolding compresses dorsally, reducing the sagittal diameter of the spinal canal. Repetitive extension causes dynamic compression, leading to further microischemia, demyelination, and axonal loss.[13]

Cervical spondylosis produces a continuum of mechanical deformation, vascular insufficiency, and neuroinflammatory change. The relative dominance of these mechanisms determines whether patients manifest axial pain alone, radiculopathy, or progressive cervical myelopathy. The terms cervical degenerative disc disease, cervical spondylosis, and cervical disc herniation are interrelated: degenerative disc disease is the disc-centric pathology, spondylosis encompasses the wider degenerative remodeling of the cervical spine (including discs, vertebral bodies, facets, and ligaments), and disc herniation represents a focal herniation event that may arise during the degenerative process.[14][15]

Histopathology

Disc herniation can be an early precursor to the development of spondylosis. While spondylotic and herniated discs undergo similar degenerative changes (eg, macrophage infiltration, upregulation of growth factors, and cytokines), immunohistological distinctions exist between the 2 disease processes. In a 2008 study by Kokubo et al, a total of 500 cervical intervertebral discs excised from 198 patients with disc herniation and 166 patients with spondylosis were examined via en bloc histological analysis and immunohistochemical staining.

Chondrocytes taken from both groups were abundant in CD68-positive macrophages, tumor necrosis factor (TNF)-α, matrix metalloproteinase (MMP)-3, basic fibroblast growth factor, and vascular endothelial growth factor. However, herniated discs demonstrated more profound inflammatory reactions involving CD68-positive macrophage infiltration into the outer layer of the annulus fibrosus. Spondylotic discs were observed to have thicker bony endplates with a more diffuse expression of TNF-α and MMP-3 in the inner layer of the annulus fibrosus.[10][11]

History and Physical

The history collection should focus on the timeline of the pain, radiation of pain, aggravating factors, and inciting events. Classically, symptomatic cervical spondylosis presents as 1 or more of the following 3 primary clinical syndromes:

Axial Neck Pain

• Patients commonly complain of stiffness and pain in the cervical spine, which is most severe in the upright position and relieved with bed rest, as removing the axial gravity load from the neck.
• Neck motion, especially in hyperextension and side-bending, typically increases the pain.
• In upper and lower cervical spine disease, patients may report radiating pain into the back of the ear or occiput versus radiating pain into the superior trapezius or periscapular musculature, respectively.
• Occasionally, patients can present with atypical symptoms of cervical angina, such as jaw pain or chest pain.
• Upper cervical radiculopathy can present as axial neck pain.

Cervical Radiculopathy

• Radicular symptoms typically follow a myotomal distribution, depending on the nerve root(s) involved, and can present as unilateral or bilateral neck pain, arm pain, scapular pain, paresthesia, and arm or hand weakness.
• Pain is exacerbated by head tilt toward the affected side or by hyperextension and side-bending toward the affected side.

Cervical Myelopathy

• This condition typically has an insidious onset, but often occurs without neck pain.
• This can initially present with hand weakness and clumsiness, resulting in the inability to complete tasks that require fine motor coordination (eg, buttoning a shirt, tying shoelaces, picking up small objects).
• Patients frequently report gait instability and unexplained falls.
• Urinary symptoms are rare and typically appear late in disease progression.

On first appearance, the patient may appear immobile and stiff in the head and neck region, particularly with increased axial neck pain upon cervical spine movement. Tender “trigger” points are frequently present within the superior trapezius, cervical paraspinal, and/or periscapular muscles. Suppose there is radiating pain down the upper limb with head extension and ipsilateral head rotation to the affected side. In that case, it is considered a positive Spurling test for cervical radiculopathy.

Results from a 2011 study by Shabat et al showed the Spurling test was 95% sensitive and 94% specific for diagnosing nerve root pathology in 257 patients, as confirmed by cervical spine computed tomography and/or magnetic resonance imaging (MRI).[16] In some cases, manual neck distraction may alleviate radicular pain. Electric shock-like sensations radiating down the spine and into the extremities with cervical flexion are a positive Lhermitte sign, which is concerning for degenerative cervical myelopathy (DCM), also known as cervical spondylotic myelopathy. A more specific sign for DCM is the Hoffman sign, which is elicited by flicking the patient’s distal phalanx of the middle finger and observing reflexive flexion of the thumb and/or index finger.

All physical exams should include a meticulous evaluation of bilateral extremities for muscle strength, sensation, and deep tendon reflexes to look for weakness in a myotomal distribution, sensory deficits in a dermatomal pattern, and reflex changes, respectively; all of which can help to identify the compromised nerve root(s) and/or myelopathy. The clinician can evaluate the patient’s gait and balance using a toe-to-heel walk and Romberg tests. In the latter, the patient stands with eyes closed and arms held forward. An increased loss of balance is interpreted as a positive Romberg test and indicates dysfunction involving the spinal cord's dorsal columns.

The presence of upper motor neuron signs (eg, spasticity, hyperreflexia, sustained clonus, extensor Babinski response) should raise the examiner’s clinical suspicion of spinal cord compromise. Another screening test for myelopathy is the grip-and-release test. Typically, a patient can make a fist and release it 20 times in 10 seconds, with decreasing cut-off values with increasing age and lower cut-off values in women compared to men.[17]

Patients with DCM present with a diverse range of complaints. Hand paresthesias, numbness, clumsiness, weakness, and gait imbalance are common.[18] Less conventional symptoms account for about 40% of the total burden. A feeling of heavy legs is the lone early feature that predicts faster diagnosis.[19] Bedside signs may be subtle. The Tromner reflex and general hyperreflexia are most sensitive. Babinski, Tromner, clonus, and the inverted supinator sign are highly specific. Upper-motor-neuron signs may be absent. Clinicians must combine the full symptom history with imaging to confirm the diagnosis.[20]

Evaluation

Imaging used in the assessment of suspected cases of cervical spondylosis includes the following:

Radiograph

Plain radiographs are an appropriate initial imaging study for neck and upper extremity pain without “red flag” symptoms. However, degenerative changes seen on imaging often poorly correlate with neck pain.[21] Common radiographic findings include osteophyte formation, disc space narrowing, endplate sclerosis, degenerative changes in the uncovertebral and facet joints, and calcified or ossified soft tissues.

Anteroposterior, lateral, and oblique views of the spine are adequate to evaluate for foraminal stenosis, sagittal alignment, and the size of the spinal canal. The Torg-Pavlov ratio is obtainable by comparing the spinal canal sagittal diameter to the vertebral body's sagittal diameter. The normal value is 1, with a ratio of less than 0.8 indicating cervical stenosis. Flexion and extension views also merit consideration if there is a concern for ligamentous instability.

Magnetic resonance imaging

Magnetic resonance imaging (MRI) is the gold standard for evaluating neural structures and soft tissues. This imaging enables the proper visualization of the entire cervical spine without exposing the patient to radiation. Sagittal and axial cuts can help quantify the extent of nerve and cord compression and reveal offending pathological changes (eg, herniated discs, bony spurs, ligamentum flavum hypertrophy, or facet joint arthropathy). A hyperintense spinal cord signal on T2-weighted images can indicate edema, inflammation, ischemia, myelomalacia, or gliosis.[22] Despite the high sensitivity of MRI studies for spondylotic changes, they should not be a routine part of the diagnostic workup unless indicated, given the high prevalence of degenerative findings on MRI in asymptomatic individuals.[23]

Computed tomography 

Computed tomography clearly defines bony structures and is more sensitive than plain radiographs in assessing intervertebral foraminal stenosis, particularly in uncovertebral or facet hypertrophy. However, it is less sensitive than MRI for evaluating soft tissues and nerve root compression.

Computed tomography myelogram

Computed tomography myelograms are better able to evaluate the location and amount of neural compression than standard computed tomography. Although they are more invasive than an MRI, they can be considered in patients who have a contraindication to MRI (eg, pacemaker) or have an artifact from the hardware placed in previous surgeries.

Discogram

Discogram is rarely necessary for cervical spondylosis; however, it is useful for the evaluation of patients who are experiencing cervical discogenic pain or have multiple herniations in which surgery is a strong possibility. However, the diagnostic procedure remains controversial as it may accelerate the degeneration of normal discs.[24]

Electromyogram

Electromyograms can supplement neuroimaging findings in the diagnosis of cervical radiculopathy. They are especially valuable in differentiating nerve root compression from other possible concomitant neurologic conditions, including peripheral neuropathies, entrapment neuropathies, brachial plexopathies, myopathies, and motor neuron diseases.

Imaging Grading Systems

Cervical spondylosis encompasses a broad range of degenerative changes affecting the discs, facet joints, overall sagittal balance, and the central canal and neural foramina. Each structure may deteriorate at its own pace. As a result, no single scale captures the entire spectrum of the disease. Clinicians instead select from several complementary grading tools, each designed to quantify the severity of a specific component.

Disc degeneration

The Pfirrmann system, although created for the lumbar spine, is the most widely adopted method for rating cervical discs from grade I to V based on T2 signal, nucleus–annulus distinction, and disc height. In a head-to-head reliability study, Pfirrmann ratings showed fair interobserver agreement and almost perfect intraobserver consistency, demonstrating its practical robustness despite its lumbar origin. The Suzuki scale was developed specifically for the cervical spine and categorizes the assessment into 4 grades, explicitly noting disc bulge. Interobserver agreement is comparable to Pfirrmann, and intraobserver reliability is substantial. Including bulging offers a clearer reflection of morphological changes that may precipitate neural compression; however, the method remains largely qualitative and subject to reader interpretation.

• The Pfirrmann classification is a magnetic resonance imaging scale that rates cervical disc degeneration from grade I to grade V.
  • Grade I represents a homogeneous, bright white disc in which the nucleus and annulus are demarcated and disc height is preserved.
  • In grade II, the nucleus signal becomes inhomogeneous, sometimes showing horizontal bands, yet height is still maintained.
  • Grade III shows a grey and heterogeneous nucleus with loss of the sharp nucleus–annulus boundary and the first signs of height reduction.
  • Grade IV depicts a dark grey to black disc with indistinguishable internal architecture and a moderate decrease in height.
  • Grade V describes a black disc space that has collapsed almost completely, reflecting advanced structural failure.
• The Suzuki classification was developed specifically for the cervical spine; this system recognizes 4 stages of degeneration.
  • Grade 0 denotes a healthy disc that retains high, homogeneous signal and normal height.
  • Grade I shows an inhomogeneous nucleus, but no bulge or loss of height.
  • Grade II is characterised by blurring of the nucleus–annulus border, the appearance of a disc bulge, and less than a 25% reduction in height.
  • Grade III reflects more than 25% height loss, signaling advanced degeneration.[25]
• The disc signal intensity index (DSI²) introduces an objective approach by averaging the T2 signal measured in three disc regions and dividing this value by the cerebrospinal fluid signal on the same midsagittal image. The result is a continuous number that declines with advancing degeneration and shows a linear relationship with Pfirrmann grades. Recent work in 770 cervical discs reported excellent intraobserver and interobserver intraclass correlation coefficients of 0.95 and 0.90, respectively, confirming high reproducibility. The same study's results demonstrated that age, body mass index, and Modic end-plate changes independently influence DSI², highlighting its potential for longitudinal monitoring and risk stratification.[26]

Endplate/Marrow Changes

Modic scheme

Modic changes describe MRI signal alterations in the vertebral endplates and adjacent marrow, interpreted as different stages of a degenerative cascade. On T1- and T2-weighted images, Modic type I appears hypointense on T1 and hyperintense on T2, reflecting inflammatory edema; type II becomes hyperintense on both sequences, indicating fatty marrow replacement; type III turns hypointense on both sequences, corresponding to sclerotic bone. Several authors add grade 0 for a normal endplate and further stratify the spatial extent of each type, but the 3-tier Modic scheme remains the clinical standard.

The prevalence of Modic changes in the cervical spine ranges from approximately 5% to 40%, with type II being the most dominant pattern and the C5–C6 level being the most commonly affected. Variation in reported frequency largely reflects differences in study populations—healthy volunteers sit at the lower end of the spectrum, whereas symptomatic cohorts approach the higher values. Regardless of prevalence, studies consistently show that the lower subaxial segments bear the greatest mechanical load and exhibit Modic conversion most often. Clinically, Modic changes correlate with both structural and symptomatic markers of degeneration. Segments that display Modic signal alterations are more likely to harbor high-grade disc degeneration or frank herniation than segments without such changes.

Conversely, patients with Modic involvement report higher rates of chronic axial neck pain and lower disability scores. Type I lesions, representing active inflammatory marrow, appear to have the strongest association with accelerated disc collapse, although type II lesions dominate cross-sectional studies. Recognition of these marrow changes enriches the radiographic evaluation of cervical spondylosis by linking endplate biology to patient symptoms and flagging levels at risk for rapid deterioration.[27]

The disc–endplate–bone marrow complex (DEBC) classification builds on the traditional Modic scheme by analyzing short tau inversion recovery (STIR) sequences in addition to standard T1- and T2-weighted images, grading the disc, endplate, and adjacent marrow as an interdependent unit. Because STIR suppresses fatty signal, the system reveals inflammatory edema that a conventional Modic read can mistake for fat, and it labels the composite appearance as type A (acute edema), type B (mixed chronic inflammation), type C (latent fatty change), or type D (sclerosis).

Intraobserver and interobserver agreement for assigning these 4 phenotypes is excellent (κ ≈ 0.81), emphasizing the method’s reproducibility. In a 2025 observational cohort of 301 neck-pain patients and 200 trauma controls, DEBC changes were present in about 1 in 5 symptomatic individuals and 1 in 8 controls, with type C predominating at C5–6 and C6–7 in both groups. The STIR component reclassified roughly a quarter of endplates that Modic had called type 2, demonstrating that mixed edema-fat lesions are common and easily overlooked when STIR is omitted. Most importantly, the study showed that surgical risk is driven by the constellation of imaging findings rather than by any single feature.

When a disc herniation coincided with a DEBC lesion, the odds of requiring operative treatment rose nearly 7-fold compared with herniation alone. In contrast, isolated DEBC or end-plate erosions conveyed little incremental risk. These data reinforce the concept that disc, endplate, and marrow act as a single biomechanical and biological complex and that combined pathology at this interface identifies segments prone to clinical failure.[28]

Facet joint degeneration

Kellgren and colleagues developed a plain radiographic grading system to assess cervical facet joint degeneration on lateral views, assigning grades from 0 to 4.

• Grade 0 indicates a normal joint.
• Grade 1 is assigned when marginal osteophytes are faint or questionable.
• Grade 2 reflects definite osteophyte formation with associated subchondral sclerosis.
• Grade 3 includes moderate osteophytes and irregularity of the articular surfaces.
• Grade 4 denotes severe degeneration, characterized by numerous large osteophytes, marked subchondral sclerosis, and pronounced articular surface irregularity, consistent with advanced arthropathy.[29]

Park et al proposed a 4-tier system that relies on axial, sagittal, or coronal computed tomography images.

• Grade I describes a joint that appears normal.
• Grade II captures early degeneration characterised by joint-space narrowing, small osteophytes, cyst formation, and no hypertrophic widening of the facet masses.
• Grade III denotes progressive disease in which large osteophytes produce facet hypertrophy without bony fusion.
• Grade IV is reserved for end-stage arthrosis when a solid osseous bridge obliterates the articular cleft.[30]

Spinal alignment 

A complete radiographic appraisal of cervical alignment begins with an upright lateral film that includes the skull base and the upper thoracic inlet. From this image, a small group of reproducible measurements captures the sagittal profile and has been linked to pain scores, neurologic recovery, and the risk of adjacent-segment disease. The C2–7 Cobb angle is obtained on an upright lateral radiograph by drawing lines along the inferior endplates of C2 and C7, erecting perpendicular lines, and measuring the intersecting angle. The C2–7 Cobb angle reflects the global cervical lordosis. In asymptomatic adults, the mean value of lordosis is approximately 40°. Most surgical reconstructions aim to restore the curve to within about 20° of the patient’s T1 slope, thereby optimizing neurological and functional outcomes.[31] 

The C2–7 sagittal vertical axis represents the horizontal distance from a plumb line dropped from the center of C2 to the posterior-superior corner of C7 vertebral body. Normal alignment averages 1.7 cm, whereas offsets exceeding 4 cm indicate positive sagittal malalignment that correlates with worse health-related quality-of-life scores and a higher risk of adjacent-segment failure after fusion.[32] The T1 slope is the angle between a horizontal reference and the superior endplate of T1. Because higher slopes demand greater lordosis to balance the head, surgeons often calculate the T1 slope minus the cervical lordosis mismatch. Values greater than 20° define deformity, whereas a mismatch below 20° suggests acceptable sagittal balance and is a common postoperative target.[33]

The chin–brow vertical angle (CBVA) gauges horizontal gaze by measuring the angle between a line from the brow to the chin and a true vertical. When the CBVA deviates beyond ±10°, patients often experience difficulty with everyday tasks, such as walking or navigating stairs. As a result, corrective osteotomies typically aim to restore approximately 10° of flexion to optimize functional gaze and mobility.[34]

Canal stenosis

MRI is the reference standard for grading central cervical canal stenosis. Kang and colleagues proposed a 4-point scale that relies solely on sagittal T2-weighted images.

• Grade 0 indicates a normal canal.
• Grade 1 records obliteration of more than half of the subarachnoid space while the spinal cord contour remains intact.
• Grade 2 designates visible cord deformation without alteration of the intramedullary signal.
• Grade 3 identifies a hyperintense signal within the compressed cord segment. The same study documented a clear rise in neurological deterioration and surgical consideration with increasing grade.[35] 

Vaccaro and coauthors later proposed a 5-grade scheme that maintains the cerebrospinal fluid–cord relationship described by Kang but distinguishes pure canal compromise from actual cord deformation.[36] Both the Kang and Vaccaro grading schemes are generated entirely from a single, mid-sagittal T2-weighted magnetic resonance image. While this approach is fast and reproducible, it ignores information in the axial plane. As a result, subtle or asymmetric cord indentations that are visible on axial sections can be under-represented.

For example, residual cerebrospinal fluid clefts on the sagittal view may mask a marked anterior cord flattening that is only visible axially, leading both systems to under-grade the stenosis. Several quantitative axial metrics have been proposed to capture deformation that is missed on sagittal imaging. One of the simplest is spinal cord eccentricity (sometimes described as “loss of roundness”). When the uncompressed cervical cord is modelled as a circle, its eccentricity approaches 0; compression transforms the cross-section into an ellipse, increasing the eccentricity value in proportion to the degree of flattening.[37]

Flexion-extension MRI (also referred to in the literature as dynamic or kinematic MRI) can reveal dynamic stenosis that static scans in a neutral neck position may miss. In 81 patients with degenerative disease, flexion-extension scans showed that the prevalence of cord compromise rose with advancing degeneration. The cervical extension position uncovered new cord impingement in 27% of the patients, whereas flexion revealed new cord impingement in 5%.[38] In a separate surgical cohort, adding flexion-extension views altered the planned decompression level in 46% of cases and yielded superior outcomes (modified Japanese Orthopaedic Association score +3.9 vs +2.4; visual analog scale –4.1 vs –2.7) compared with plans based solely on neutral MRI views.[39]

Diffusion MRI provides quantitative microstructural biomarkers that surpass canal diameter in grading the severity of cervical spondylotic myelopathy. In a prospective study, diffusion basis spectrum imaging differentiated controls, mild, and severe cases with 81% accuracy. These metrics predicted postoperative modified Japanese Orthopaedic Association improvement more accurately than conventional imaging, confirming their prognostic value.[40] A technique with future potential is deep learning-accelerated MRIs, which can shorten spine MRIs by 40% while preserving diagnostic equivalence with standard MRIs.[41] 

Foraminal stenosis

The axial T2-weighted grading system proposed by Kim and colleagues classifies each neural foramen into 3 grades by comparing its narrowest width with the diameter of the adjacent extraforaminal nerve root on the same slice.

• Grade 0 indicates no stenosis when the foraminal width exceeds the nerve-root width.
• Grade 1 describes moderate narrowing in which the foraminal width is equal to or less than, but still more than 50% of, the nerve-root width.
• Grade 2 denotes severe stenosis when the foraminal width is 50% or less of the nerve-root width or when the foramen is almost completely obliterated.[42] 

The oblique sagittal MRI system developed by Park et al grades stenosis from 0 to 3 according to the degree of perineural fat loss and whether the nerve root is deformed.

• Grade 0 shows an intact halo of fat around a normal-shaped root.
• Grade 1 represents mild stenosis with obliteration of less than half of the circumferential fat and no root distortion.
• Grade 2 indicates obliteration of more than half of the fat while the root contour remains intact.
• Grade 3 captures severe disease in which perineural fat is essentially absent and the root is flattened or collapsed.[43]

Treatment / Management

The treatment strategy for cervical spondylosis depends on the severity of a patient’s signs and symptoms. In the absence of “red flag” symptoms or significant myelopathy, the goals of treatment are to relieve pain, improve functional ability in day-to-day activities, and prevent permanent injury to neural structures. Red flags are clinical warning signs—such as severe or rapidly progressing neurological deficits, bowel or bladder dysfunction, or unexplained weight loss—that may indicate serious underlying pathology requiring urgent evaluation. Symptomatic cervical spondylosis should be approached in a stepwise fashion, starting with nonoperative management.

Nonsurgical

The mainstay of nonsurgical treatment is a 4- to 6-week course of physical therapy, including isometric and resistance exercises to strengthen the neck and upper back muscles. A meta-analysis by the Philadelphia Panel found that physical modalities such as cervical traction, heat, cold, therapeutic ultrasound, massage, and transcutaneous electrical nerve stimulation lacked sufficient evidence regarding their efficacy in treating acute or chronic neck pain.[44] However, in patients experiencing radicular pain, cervical traction may be incorporated to alleviate the nerve root compression that occurs with foraminal stenosis.(A1)

Pharmacologic agents, including nonsteroidal anti-inflammatory drugs, oral steroids, muscle relaxants, anticonvulsants, and antidepressants, can be prescribed for pain relief. Therapy can be escalated to opioid analgesics for refractory axial neck pain, but is not recommended as first-line or for long-term use due to their potential adverse effects. Durable medical equipment can be a consideration for symptomatic relief. Short-term use of a soft cervical collar can sometimes alleviate acute neck pain and spasm. Nighttime use of a cervical pillow may help relieve neck pain by maintaining the normal cervical lordosis, improving biomechanical load distribution between discs, and promoting better sleep quality.

Trigger point injections can be employed to treat myofascial trigger points, which can clinically manifest as neck, shoulder, and upper arm pain. More invasive interventional treatment options include epidural steroid injections (ESIs), zygapophysial (facet) joint injections, medial branch blocks, and radiofrequency lesioning. In a 2019 systematic review and meta-analysis by Conger et al, approximately half of the patients with cervical radicular pain experienced at least 50% pain reduction at 1 and 3-month follow-ups after cervical transforaminal ESIs.[45] (A1)

There are long-term reports of success in 40% to 70% of patients who underwent interlaminar or transforaminal ESIs to treat cervical radiculopathy. In a 2015 systematic review by Manchikanti et al, long-term pain relief was observed with cervical radiofrequency lesioning, medial branch blocks, and facet joint injections.[46] However, the limitations of these systematic reviews include a paucity of high-quality studies and, more specifically, the lack of investigations with placebo or sham control comparison groups.(A1)

Surgical

Surgical intervention should be considered in patients with cervical myelopathy, significant, persistent axial neck pain, or cervical radiculopathy following failure of nonoperative measures. These affected individuals must also have a pathological condition, as demonstrated by neuroimaging studies, that corresponds to their clinical features. The surgical approach depends on the clinical syndrome and the pathology site(s).

The anterior approach involves a cervical discectomy or corpectomy followed by autograft, allograft, or artificial intervertebral disc fusion. Anterior plates, metallic cages, and synthetic spacers can be used with bone grafts, resulting in comparable fusion rates; however, the long-term outcomes are still unclear. In patients who experience radicular pain due to central or bilateral disc herniation, an anterior approach is preferable, whereas either an anterior or posterior approach is an option for a lateral disc lesion. Anterior cervical discectomy and fusion are used to treat patients with myelopathy and pathological compression up to 3 levels or when cervical lordosis is lost.

The posterior approach options consist of partial discectomy, laminotomy-foraminotomy, laminoplasty, and laminectomy. Foraminotomy alone is adequate in patients with foraminal stenosis due to bone spur formation and/or lateral disc herniation. Laminectomy allows for multilevel cervical decompression, and it is typically paired with instrumented fusion to prevent postoperative iatrogenic kyphosis. Laminoplasty offers a motion-preserving alternative, but it should be selected only after confirming the absence of significant kyphosis, segmental instability, or substantial ventral compressive pathology, as these findings contraindicate the procedure.

Differential Diagnosis

Many disorders outside the cervical spine can replicate the pain or neurologic deficits seen in cervical spondylosis. Careful history, focused examination, and targeted investigations separate these entities and prevent missed pathology.

Local musculoskeletal sources

Cervical sprain, strain, facet arthropathy, and myofascial trigger points generate axial neck pain without dermatomal radiation and leave neurologic testing normal. Glenohumeral arthritis, rotator cuff disease, and acromioclavicular joint degeneration cause lateral shoulder pain that worsens during overhead activity, while the Spurling maneuver remains negative. Shoulder injection that abolishes pain confirms a shoulder origin.

Peripheral nerve and plexus lesions

Carpal tunnel syndrome yields nocturnal paresthesia in the thumb, index, and middle fingers, preserves neck motion, and shows slowed median conduction across the wrist. Cubital tunnel syndrome produces numbness in the ring and little fingers with intrinsic hand weakness but spares the ulnar forearm skin, a pattern not explained by C8 involvement. Brachial neuritis presents with acute shoulder pain followed by patchy paresis unrelated to a single root.

Thoracic outlet and vascular compression

Positional paresthesia or arm fatigue that improves when the shoulder is lowered suggests neurovascular compromise at the thoracic outlet. Provocative elevation tests reproduce symptoms, whereas cervical traction does not relieve them.

Nondegenerative spinal conditions

Acute fracture, epidural abscess, osteomyelitis, or metastatic tumor must be excluded when pain is constant, worsens at night, or accompanies systemic signs such as fever, weight loss, or a history of malignancy. MRI is essential when red flags appear.

Intrinsic neurologic disorders

Multiple sclerosis, vitamin B12 deficiency, amyotrophic lateral sclerosis, and Guillain-Barré syndrome mimic myelopathy by producing limb weakness, gait imbalance, or sensory loss without concordant imaging compression. Electromyography and appropriate laboratory tests differentiate these entities.

Prognosis

Cervical spondylosis is a slowly progressive, degenerative disease process that deteriorates with age. However, the severity of symptoms may not necessarily correlate with the degree of spondylosis seen on neuroimaging. Patients who present with axial neck pain typically improve over time, but can have a recurrence of pain. One study's results noted that 79% of patients with neck pain improved or became asymptomatic by the 15-year follow-up after the onset of symptoms.[47] Anywhere from 50% to 75% of persons with current neck pain report neck pain again 1 to 5 years later.

In a 2008 study by Carroll et al, psychosocial factors, including psychological health, coping patterns, and the need to socialize, were the strongest prognostic factors of neck pain.[48] Most patients with cervical radiculopathy have an eventual resolution of symptoms over 1 to 2 years without surgical intervention.[49] On the other hand, the long-term prognosis in cervical spondylotic myelopathy is less clear. In patients with mild-to-moderate symptoms, the natural course of cervical spondylotic myelopathy is highly variable.[50] In patients with moderate and severe degenerative cervical myelopathy (DCM), surgery is recommended, while for patients with mild DCM, surgery is considered a valid option.[51][52]

Complications

Untreated cervical spondylosis may culminate in permanent neurologic deficit. In a 2019 cohort study by El-Yahochouchi et al, the overall incidence of immediate and delayed adverse events following an epidural steroid injection was 2.4% and 4.9%, respectively.[53] Complications include:

• Neurologic injury
• Epidural abscess
• Epidural hematoma
• Increased pain
• Vasovagal reactions
• Central steroid response (eg, facial flushing, nonpositional headaches)
• Endocrinologic effects (eg, hyperglycemia, hypothalamic-pituitary axis suppression, decreased bone density)

Complications from anterior and posterior cervical spine surgery include:

• Injury to the spinal cord and nerve roots
• Infection
• Dural tear and cerebrospinal fluid leak
• Recurrent laryngeal, superior laryngeal, and hypoglossal nerve injuries
• Esophageal injury and dysphagia
• Vertebral and carotid artery injuries
• Tracheal injury
• Adjacent segment degeneration
• Pseudoarthrosis
• Postlaminectomy kyphosis [54][55] 

A 2025 meta-analysis, pooling 222 observational cohorts and anterior cervical discectomy and fusion cases, recorded an overall morbidity of 16% and a mortality rate of 0.1%. Prevertebral or neck swelling was most common at 11.3%, followed by pseudarthrosis (10%) and dysphagia (9.5%). Cage subsidence occurred in 9.4%, myelopathy in 7.7%, hoarseness 2.3%, C5 palsy 2.1%, and recurrent laryngeal palsy 2%.[56] In explainable machine learning models, several features emerge associated with complications in surgeries usually performed for cervical spondylosis. These features include single versus multiple-level surgery, age, body mass index, preoperative hematocrit, and American Society of Anesthesiologists physical status.[57][58][59][60]