Incidence of Cancer in Spinal Deformity Patients Receiving High-Dose (≥40 mg) Bone Morphogenetic Protein (rhBMP-2)

Study Design. Level III, Retrospective observational study. Objective. To determine if there is an increased risk of developing cancer after exposure to high-dose recombinant human bone morphogenetic protein-2 (rhBMP-2) and if risk is dose and/or exposure-dependent. Summary of Background Data. Concerns have been raised regarding a relationship between rhBMP-2 and cancer. Methods. A total of 642 adult deformity patients from a single institution receiving a cumulative rhBMP-2 dose ≥40 mg from July, 2002 to July, 2009 were identified. Patients with a history of surveillance, epidemiology, and end result (SEER) cancer before rhBMP-2 exposure were excluded. To determine the occurrence of a cancer event, questionnaires were mailed and telephone follow up attempted for nonresponders. Only cancers tracked by the National Cancer Institute (NCI) SEER registry were included. Observed cancer counts were compared to expected cancer counts based on general population incidence rates within 5-year age strata. Cumulative incidence competing risk (CICR) modeling was used to evaluate the association between rhBMP-2 exposure and cancer controlling for potential confounding variables. Results. Forty-nine patients were lost-to-follow up, leaving 593 patients (92.4%; 138 males/455 females) available for analysis. Mean age was 52.8 years at the time of first exposure. Mean cumulative rhBMP-2 dose was 113.5 mg with 85% having one exposure (range: 1-8). Mean follow-up [date of exposure to date of death (regardless of cause) or returned completed questionnaire] was 5.6 ± 1.9 years; median follow up was 5.4 years. A total of 342 patients have greater than 5-year follow up. Minimum follow up was 2.0 years or until occurrence of a SEER cancer. Our total 8-year cumulative incidence of new SEER cancer accounting for the competing risk of death was 7.4% for 30 cancers in 593 patients. Fewer cancers were observed than expected based on general population rates, though the difference was not statistically significant (expected = 34; standardized incidence ratio = 0.88, 95% confidence interval, CI = 0.60-1.26). CICR found neither cumulative rhBMP-2 dose (hazard ratio, HR = 0.995, 95% CI 0.988-1.003; P = 0.249) nor number of exposures (HR = 0.776, 95% CI 0.359-1.677; P = 0.519) increased the risk of developing a postexposure cancer after controlling for known cancer risk factors. Conclusion. The incidence of a SEER cancer after rhBMP-2 exposure was similar to incidence reported by the NCI. There were no significant rhBMP-2 dose or multi-exposure related risks of developing a life-threatening cancer.