Combined Anterior-Posterior Decompression and Fusion for Cervical Spondylotic Myelopathy

Operative Technique: Posterior Approach

A midline incision was made through the skin and subcutaneous tissue to the level of the deep cervical fascia. Then, dissection was performed to the tips of the lateral masses. Instrumentation and fusion preceded spinal decompression. This order, chosen to preserve bony landmarks for guidance during instrumentation, did not interfere with subsequent decompression. Segmental spinal instrumentation was placed using lateral mass screw-rod fixation. After the laminae and ligamenta flava were bilaterally mobilized, the entire bony ligamentous complex spanning the area of fusion was removed en masse (most commonly C3–C7) in order to decrease the number of instrument passes near the spinal cord. Next, a modest foraminotomy was performed to extend the opening laterally and ensure adequate decompression of the nerve roots. Autograft harvested from the spinous processes and laminae was used. The posterior portion of the operation contributed significantly to blood loss and postoperative pain during the perioperative period. We recommend performing a very meticulous dissection to minimize these consequences. No patient in this study required a halo orthosis.

Results

Twenty-one patients with CSM were treated with CAPDF between February 2010 and March 2015 (Table 1).

Mean age was 62.1 years (range, 44-79 years). Nine patients were female, and 12 were male. The primary diagnosis was CSM, but several patients presented with associated conditions, including congenital cervical spinal stenosis (3 cases), cervical kyphotic deformity (3 cases), and ossification of posterior longitudinal ligament (OPLL; 4 cases). Two patients previously underwent CS surgery: anterior fusion at C4–C6 (patient 8) and separate anterior fusions (C5–C6, C6–C7) about 11 years apart (patient 16). For these 2 patients, combined anterior-posterior surgery was performed not to revise their fusions but to extend their constructs to address ASD. The 21-patient cohort had high rates of comorbidities: Thirteen patients (61.9%) were obese or morbidly obese, 6 (28.6%) had diabetes mellitus (insulin-dependent in 2 cases), and 13 (61.9%) had a history of tobacco use (6 of these patients used tobacco the previous year).

Table 2 summarizes the operative data. Mean number of levels fused was 2 (range, 1-3) anteriorly and 3 (range, 1-4) posteriorly.

C3–C4 was the most common fusion range anteriorly, and C3–C7 was the most common range posteriorly. Figure 1 shows the frequency of type of fusion performed both anteriorly and posteriorly.

Mean operative time, which included supine-to-prone repositioning at the end of the anterior operation, was 4 hours 55 minutes (range, 3:04-6:22). Mean EBL was 131 mL (range, 55-278 mL), and mean LOS was 5 days (range, 2-10 days).

Of the 21 patients, 9 (42.3%) had at least 1 complication during the perioperative period. Table 3 summarizes all encountered complications. Neither neurologic instability nor mortality was observed after surgery.

With the exception of 1 case of adjacent segment kyphosis, all complications were transient, yielding a long-term complication rate of 4.8%. The most frequently encountered complications were dysphagia (28.6%) and excessive wound discharge (9.5%). Dysphagia is a common complication after anterior cervical surgery, with exposure above C4 being a significant risk factor.¹³ Such exposure was found in 4 (66.7%) of the 6 patients with dysphagia. One patient in the study experienced kyphotic collapse below the level of fusion. Subsequent computed tomography showed no evidence of hardware failure, flexion/extension radiographs showed no progression of the kyphotic deformity, and the patient remained asymptomatic and did not undergo reoperation. The deformity was attributed to low bone quality, not to any abnormality in the surgical construct.

Patient 7 was lost to follow-up. For the other 20 patients, mean time to “3-month follow-up” was 96 days (range, 51-149 days). The most commonly noted improvements in quality of life included resolution of numbness, improvement in gait, and return to previous activities, such as walking and even exercising.

The most common improvements noted on neurologic examination included decreased hyperreflexia, less reproducible beats or complete absence of ankle clonus, gait improvement, and increased motor strength. Neurologic status was subjectively improved for 19 patients and unimproved for 1 patient; no patient’s neurologic status was worse (Table 4).

Mean Nurick grade was 1.9 before surgery and 1.1 after surgery (mean difference, 0.80; P < .001). Table 5 shows the distribution of patients’ Nurick grades before and after surgery.

Representative Case

Patient 15, a 53-year-old man, presented with complaints of dysesthesias of the hands. Focused neurologic evaluation at the time revealed limited CS range of motion on extension. The patient (Figures 2A-2D) was diffusely hyperreflexic and had pathologic spread in the upper extremities.

He underwent C3–C6 anterior discectomy, instrumentation, and fusion followed by C3–C6 posterior laminectomy, medial facetectomy, instrumentation, and fusion.