Abstract / Summary
Background and Objectives: Conventional midline expansive open-door laminoplasty (EODL) may disrupt the posterior cervical extensor mechanism. This study compared 10-year outcomes after conventional midline and unilateral para-midline muscle-preserving EODL. Materials and Methods: This retrospective cohort included 100 adults who underwent EODL for multilevel cervical spondylotic myelopathy from 1 January 2011 through 31 December 2015 (42 conventional and 58 muscle-preserving). At 10 years, 94 patients completed clinical assessments and 53 underwent MRI examinations. The primary outcome was the postoperative trajectory of the 17-point Japanese Orthopaedic Association (JOA) score from 3–6 months through 10 years. Models adjusted for age, sex, baseline outcome, and decompressed levels; sensitivity models additionally adjusted for operative year. Results: After operative-year adjustment, group-by-time interactions were significant for JOA (F(4, 385.749) = 9.094, p < 0.001), visual analog scale (VAS; F(4, 384.457) = 5.559, p < 0.001), and Neck Disability Index (NDI; F(4, 396.021) = 5.109, p = 0.001). At 10 years, adjusted means in the muscle-preserving versus conventional groups were 10.121 vs. 8.738 for JOA (difference, 1.383; 95% CI, 0.867–1.899), 3.718 vs. 5.946 for VAS (difference, −2.228; 95% CI, −2.643 to −1.813), and 15.395 vs. 20.592 for NDI (difference, −5.197; 95% CI, −6.952 to −3.442; all p < 0.001). Observed 10-year lordosis was 4.73° vs. −3.04°, sagittal vertical axis was 18.83 vs. 38.22 mm, and range of motion was 14.55° vs. 8.57° (all p < 0.001). Adjusted normalized paraspinal cross-sectional area was 7.840 vs. 5.591 (difference, 2.250; 95% CI, 1.752–2.748; p < 0.001). Operative time, blood loss, and hospital stay were lower with the muscle-preserving approach (88.89 vs. 121.82 min, 169.48 vs. 233.33 mL, and 6.12 vs. 7.19 days; all p < 0.001). Reoperation occurred in 4/58 vs. 6/42 patients (p = 0.314); the study was not powered to establish safety equivalence. Conclusions: The muscle-preserving approach was associated with more favorable intermediate- and long-term clinical, radiographic, and muscle-morphology outcomes. These associations persisted after operative-year adjustment for the clinical and CSA outcomes but do not establish causality because residual temporal and treatment-selection confounding remain possible.