Reoperation rates for symptomatic nonunions in posterior cervical (subaxial) fusions with and without bone morphogenetic protein in a cohort of 1158 patients.

Objective: Bone morphogenetic protein (BMP) was first approved in 2002 for use in single-level anterior lumbar fusions as an alternative to iliac crest grafts. Subsequent studies have concluded that BMP provides superior fusions rates and therefore reduces reoperations for nonunions. The purpose of this study was to determine the reoperation rates for symptomatic nonunions in posterior cervical (subaxial) spinal fusions with and without the use of BMP and to determine if the nonunion rates are statistically significantly different between the two groups.
Methods: Between January 2009 and September 2013, the authors identified 1158 posterior cervical spinal fusion cases in the subaxial spine (C2-7) from a large spine registry (Kaiser Permanente). Patient characteristics, diagnoses, operative times, lengths of stay, and reoperations were extracted from the registry. Reoperations for symptomatic nonunions were adjudicated via chart review. Logistic regression was conducted to produce estimates of odds ratios (OR) and 95% confidence intervals (CIs). Kaplan-Meier curves for the non-BMP and BMP groups were generated and compared using the log-rank test.
Results: In this cohort there were 1158 patients (19.3% with BMP) with a median follow up of 1.7 years (interquartile range [IQR] 0.7-2.9 years) and median duration to operative nonunion of 0.63 years (IQR 0.44-1.57 years). Kaplan-Meier curves showed no significant difference in reoperation rates for nonunions using the log-rank test (p = 0.179). In a subset of patients with more than 1 year of follow-up, 788 patients were identified (22.5% with BMP) with a median follow-up duration of 2.5 years (IQR 1.7-3.4 years) and a median time to operative nonunion of 0.73 years (IQR 0.44-1.57 years). There was no statistically significant difference in the symptomatic operative nonunion rates for posterior cervical (subaxial) fusions with BMP compared with non-BMP (1.1% vs. 0.7%; crude OR 1.73, 95% CI 0.32-9.55, p = 0.527) for more than 1 year of follow-up.
Conclusions: This study presents the largest series of patients using BMP in posterior cervical (subaxial) spinal fusions. Reoperation rates for symptomatic nonunions with more than 1 year of follow-up were found to be 1.1% with BMP and 0.7% without BMP. There was no significant difference in the reoperation rates for symptomatic nonunions with or without BMP.