Cervical Myelopathy

Nurick Scale - Functional Impairment (Most Commonly Used)

Nurick grades the legs. The scale is functional and focuses on gait and employment. It is the most widely used grading system for myelopathic severity and it correlates with surgical outcomes. Grades 2 and 3 are clear surgical indications. Grade 1, cord signs with a normal gait, is the controversial one: some advocate early surgery, others observe. Grades 4 and 5 have poorer outcomes, but surgery still halts progression.

Modified Japanese Orthopaedic Association (mJOA) Score

The mJOA scores more than walking. It is an 18-point scale: motor function in the upper and lower limbs, 0-5 points each; sensory function in the upper limb, lower limb and trunk, 0-2 points each; and sphincter function, 0-3 points. A full 18 is normal, 15-17 is mild myelopathy, 12-14 moderate, and under 12 severe, which is how the AOSpine cohorts and the guidelines stratify severity.

Recovery rate is what the score is really for: (postoperative mJOA minus preoperative mJOA) divided by (18 minus preoperative mJOA), times 100. A recovery rate over 50% is considered good.

Why the trials use it. It assesses every neurological domain, it is quantitative enough to yield that recovery rate, and it is sensitive to subtle change, which is why it is the outcome measure in the AOSpine CSM studies. Against that it is complex and time-consuming beside Nurick, inter-observer variability can be high, and it is not intuitive to recall.

Conservative Management - Limited Role

What conservative care can and cannot do. These measures run in parallel and are aimed at symptoms, not at the disease: myelopathy is caused by mechanical compression of the spinal cord, and nothing in this list decompresses anything. The realistic goals are comfort, function and surveillance that detects deterioration early enough to still operate usefully, because cord recovery depends on how long it has been compressed, and the damage that accrues while a patient is managed conservatively is not recovered by later surgery.

Neck immobilisation is short-term only. A soft collar relieves symptoms during an acute exacerbation. Avoid prolonged use: it causes muscle atrophy and stiffness, and the deconditioned neck is then worse at protecting itself.

Medication treats symptoms only. Gabapentin or pregabalin for the neuropathic pain of a coexisting radiculopathy, and baclofen or tizanidine for upper motor neuron spasticity. There is no role for NSAIDs or steroids in chronic degenerative myelopathy: unlike acute cord injury there is no inflammatory oedema to suppress, only chronic mechanical compression.

Physiotherapy maintains function without touching the cord. Range of motion, strengthening of the supporting musculature and gait training are all worth having, and none of them halts progression of cord compression. Avoid manipulation and traction outright - forcing movement in a stenotic canal risks worsening the compression, and this is the one active harm on the list.

Monitoring is the part that actually matters. Assess clinically every 3-6 months: Nurick grade, mJOA score, gait and upper motor neurone signs, so that deterioration is measured rather than remembered, and repeat the MRI if the examination worsens. Converting to surgery as soon as symptoms progress is the plan rather than a fallback, and the monitoring exists solely to catch the moment it applies.

Where the line sits. Observation is reasonable at Nurick 0-1 when the MRI shows mild compression without T2 signal change, when the patient is a high surgical risk with severe cardiac or pulmonary disease, and when consent has covered the natural history. Operate even at Nurick 1 when T2 signal change is present, when progression is documented over months, or when the patient is young with a long life expectancy.

Surgical Decompression - Definitive Treatment

The goal is to halt neurological decline, decompress the spinal cord and prevent further damage.

Anterior Cervical Discectomy and Fusion (ACDF)

When. One or two levels of anterior compression from disc herniation or osteophyte, with focal cord compression at those levels. It is the most common surgical treatment for cervical myelopathy.

What makes the decompression adequate. Complete PLL removal is the step that decompresses the cord posteriorly, and leaving a posterior osteophyte behind causes persistent compression. Avoid over-distraction, which distracts the facet joints and causes postoperative pain, and check on fluoroscopy that the plate does not cross an adjacent disc space.

What goes wrong. Recurrent laryngeal nerve injury is minimised by the left-sided approach and gentle retraction, dysphagia is worse after multilevel surgery, and adjacent segment disease follows the fusion, with a live debate over how much of that is simply natural history.

Anterior Cervical Corpectomy and Fusion (ACCF)

When. Multilevel anterior compression over 2-3 levels, retro-vertebral osteophytes or OPLL that cannot be reached through a discectomy, and kyphotic deformity needing the anterior column reconstructed.

What it buys. Removing the entire vertebral body gives a wide decompression across several levels through a single construct, direct access to retro-vertebral osteophytes, and correction of kyphosis by lengthening the anterior column.

What it costs. Pseudarthrosis runs at 15-20% for a two-level and 30% for a three-level corpectomy, the graft can subside, particularly in osteoporotic bone, and there is more blood loss and a longer operation. Beyond two levels the construct may need posterior supplementation.

Corpectomy or multilevel ACDF. A multilevel ACDF preserves the vertebral bodies but needs more dissection, and the plate spans more levels, which increases the stress on it. The current trend is to prefer multilevel ACDF for two-level disease and to reserve corpectomy for retro-vertebral compression that an ACDF cannot reach.

Cervical Laminoplasty - Motion-Preserving Decompression

When. Multilevel stenosis of 3 or more levels, typically C3-7; circumferential compression from front and back; congenital stenosis with acquired spondylosis; and OPLL, where it is preferred to an anterior approach because of the CSF leak risk.

The principle. An open-door laminoplasty hinges the laminae on one side, opens them on the other, and props the door with a spacer. The canal enlarges and the cord drifts posteriorly, away from anterior compression, which is how a posterior operation treats a ventral problem.

What it offers. Motion is preserved rather than fused, multiple levels are decompressed through a single posterior approach, there is no pseudarthrosis to worry about because nothing is fused, and the infection risk is lower than with the anterior approach.

What it costs. Aching axial neck pain, C5 palsy, and a 30-50% loss of flexion and extension despite the label of motion preservation. It cannot correct kyphosis, so it needs a lordotic or neutral alignment to begin with.

C5 palsy. Deltoid and biceps weakness in the C5 myotome, often with shoulder pain and patchy deltoid numbness, in a patient who woke from surgery intact. Onset is delayed, at a pooled mean of 3 days (95% CI 2.6-3.6) across 748 cases, which is why it is missed on the day-zero check.

The incidence depends on the operation, and the spread is threefold. ACDF 3.3%, laminoplasty alone 5.1%, laminoplasty plus another posterior procedure 6.5%, anterior corpectomy and fusion 7.5%, and laminectomy with fusion 11.0%, against a pooled 5.3% across 13,621 patients.

The mechanism is contested, and it is worth knowing what the evidence shows. The traditional explanation is tethering: the cord drifts posteriorly after decompression and drags on a short C5 root. Two observations sit badly with it. First, when 107 studies are stratified by approach and diagnosis, anterior decompression for radiculopathy produces essentially no C5 palsy (95% CI 0.00-0.01), while for myelopathy the rates are 4% anterior and 7% posterior with no significant difference between the approaches - yet anterior surgery causes no posterior drift. Second, the risk factors that survive multivariate analysis are duration of symptoms over 12 months and preoperative T2 cord signal change at C4-5, which are properties of the cord rather than of the foramen. Together these favour reperfusion injury of the cord grey matter, with tethering at most contributory.

Prophylactic foraminotomy is widely performed and is not established. In 70 laminoplasties all done with bilateral C4-5 foraminotomy, 5.7% still developed a C5 palsy, no better than the pooled rate, and the authors concluded plainly that prophylactic foraminotomy does not eliminate it. In a prospective series using maximal expansion the foraminotomy subgroups fared worse, not better: 33% versus 20% after laminoplasty and 50% versus 20% after posterior decompression and fusion. What is defensible is avoiding unnecessarily wide decompression and accepting that a chronically compressed, signal-changed cord carries irreducible risk.

Management, and honest consent. Exclude a compressive cause - haematoma or a malpositioned construct - with urgent imaging if the weakness is sudden or severe, then treat expectantly with physiotherapy and shoulder protection. At one year the pooled recovery rates were 100% after ACDF but only 52.9% after laminoplasty and 50% after posterior decompression and fusion, so after the posterior operations described here roughly half of affected patients are still not fully recovered at twelve months, and consent should not promise resolution.

Laminectomy and Posterior Fusion

When. Multilevel stenosis with instability or kyphosis; posterior column pathology such as ligamentum flavum hypertrophy or congenital stenosis; and the situations in which laminoplasty is contraindicated, namely kyphotic alignment or a previous laminectomy.

The principle. Remove the posterior elements, laminae and ligamentum flavum, then fuse with lateral mass or pedicle screws to prevent postoperative kyphosis.

What it offers and what it costs. Fusion prevents the postoperative kyphosis that follows a bare laminectomy and can correct an existing kyphosis through the rod contour; the construct is stable enough for early mobilisation, and the posterior compression is relieved effectively. The trade is motion, since fusion restricts range. It also takes longer and bleeds more than a laminoplasty, carries a pseudarthrosis rate of 5-10%, and screws can be malpositioned, particularly in the C2 pedicle.