Crossing the Cervicothoracic Junction During Posterior Cervical Fusion for Myelopathy Is Associated With Superior Radiographic Parameters But Similar Clinical Outcomes

BACKGROUND: For laminectomy and posterior spinal fusion (LPSF) surgery for cervical spondylotic myelopathy (CSM), the evidence is unclear as to whether fusions should cross the cervicothoracic junction (CTJ). OBJECTIVE: To compare LPSF outcomes between those with and without lower instrumented vertebrae (LIV) crossing the CTJ. METHODS: A consecutive series of adults undergoing LPSF for CSM from 2012 to 2018 with a minimum of 12-mo follow-up were identified. LPSF with subaxial upper instrumented vertebrae and LIV between C6 and T2 were included. Clinical and radiographic outcomes were compared. RESULTS: A total of 79 patients were included: 46 crossed the CTJ (crossed-CTJ) and 33 did not. The mean follow-up was 22.2 mo (minimum: 12 mo). Crossed-CTJ had higher preoperative C2-7 sagittal vertical axis (cSVA) (33.3 16.0 vs 23.8 +/- 12.4 mm, P = .01) but similar preoperative cervical lordosis (CL) and CL minus T1-slope (CL minus T1-slope) (P > .05, both comparisons). The overall reoperation rate was 3.8% (crossed-CTJ: 2.2% vs not-crossed: 6.1%, P = .37). In adjusted analyses, crossed-CTJ was associated with superior cSVA (beta = -9.7; P = .002), CL (beta = 6.2; P = .04), and CL minus T1-slope (beta = -6.6; P = .04), but longer operative times (beta = 46.3; P = .001). Crossed- and not-crossed CTJ achieved similar postoperative patient-reported outcomes [Visual Analog Scale (VAS) neck pain, VAS arm pain, Nurick Grade, Modified Japanese Orthopedic Association Scale, Neck Disability Index, and EuroQol-5D] in adjusted multivariable analyses (adjusted P > .05). For the entire cohort, higher postoperative CL was associated with lower postoperative arm pain (adjusted Pearson's r -0.1, P = .02). No postoperative cervical radiographic parameters were associated with neck pain (P > .05). CONCLUSION: Subaxial LPSF for CSM that crossed the CTJ were associated with superior radiographic outcomes for cSVA, CL, and CL minus T1-slope, but longer operative times. There were no differences in neck pain or reoperation rate.