Laminoplasty versus laminectomy with posterior spinal fusion for multilevel cervical spondylotic myelopathy: influence of cervical alignment on outcomes

OBJECTIVE Cervical curvature is an important factor when deciding between laminoplasty and laminectomy with posterior spinal fusion (LPSF) for cervical spondylotic myelopathy (CSM). This study compares outcomes following laminoplasty and LPSF in patients with matched postoperative cervical lordosis. METHODS Adults undergoing laminoplasty or LPSF for cervical CSM from 2011 to 2014 were identified. Matched cohorts were obtained by excluding LPSF patients with postoperative cervical Cobb angles outside the range of laminoplasty patients. Clinical outcomes and radiographic results were compared. A subgroup analysis of patients with and without preoperative pain was performed, and the effects of cervical curvature on pain outcomes were examined. RESULTS A total of 145 patients were included: 101 who underwent laminoplasty and 44 who underwent LPSF. Preoperative Nurick scale score, pain incidence, and visual analog scale (VAS) neck pain scores were similar between the two groups. Patients who underwent LPSF had significantly less preoperative cervical lordosis (5.8 degrees vs 10.9 degrees, p = 0.018). Preoperative and postoperative C2-7 sagittal vertical axis (SVA) and T-1 slope were similar between the two groups. Laminoplasty cases were associated with less blood loss (196.6 vs 325.0 ml, p < 0.001) and trended toward shorter hospital stays (3.5 vs 4.3 days, p = 0.054). The perioperative complication rate was 8.3%; there was no significant difference between the groups. LPSF was associated with a higher long-term complication rate (11.6% vs 2.2%, p = 0.036), with pseudarthrosis accounting for 3 of 5 complications in the LPSF group. Follow-up cervical Cobb angle was similar between the groups (8.8 degrees vs 7.1 degrees, p = 0.454). At final follow-up, LPSF had a significantly lower mean Nurick score (0.9 vs 1.4, p = 0.014). Among patients with preoperative neck pain, pain incidence (36.4% vs 31.3%, p = 0.629) and VAS neck pain (2.1 vs 1.8, p = 0.731) were similar between the groups. Similarly, in patients without preoperative pain, there was no significant difference in pain incidence (19.4% vs 18.2%, p = 0.926) and VAS neck pain (1.0 vs 1.1, p = 0.908). For laminoplasty, there was a significant trend for lower pain incidence (p = 0.010) and VAS neck pain (p = 0.004) with greater cervical lordosis, especially when greater than 20 degrees (p = 0.011 and p = 0.018). Mean follow-up was 17.3 months. CONCLUSIONS For patients with CSM, LPSF was associated with slightly greater blood loss and a higher long-term complication rate, but offered greater neurological improvement than laminoplasty. In cohorts of matched follow-up cervical sagittal alignment, pain outcomes were similar between laminoplasty and LPSF patients. However, among laminoplasty patients, greater cervical lordosis was associated with better pain outcomes, especially for lordosis greater than 20 degrees. Cervical curvature (lordosis) should be considered as an important factor in pain outcomes following posterior decompression for multilevel CSM.