Bone morphogenetic protein use in spine surgery in the United States: how have we responded to the warnings?

Abstract

Background ContextRecombinant human bone morphogenetic protein-2 (rhBMP-2) has been widely adopted as a fusion adjunct in spine surgery since its approval in 2002. A number of concerns regarding adverse effects and potentially devastating complications of rhBMP-2 use led to a Food and Drug Administration (FDA) advisory issued in 2008 cautioning its use, and a separate warning about its potential complications was published by The Spine Journal in 2011. PurposeTo compare trends of rhBMP-2 use in spine surgery after the FDA advisory in 2008 and The Spine Journal warning in 2011. Study DesignRetrospective cross-sectional study using a national database. Patient SampleAll patients from 2002 to 2013 who underwent spinal fusion surgery at an institution participating in the Nationwide Inpatient Sample (NIS). Outcome MeasuresProportion of spinal fusion surgeries using rhBMP-2. MethodsWe queried the NIS from 2002 to 2013 and used International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) procedure codes to identify spinal fusion procedures and those that used rhBMP-2. Procedures were subdivided into primary and revision fusions, and by region of the spine. Cervical and lumbosacral fusions were further stratified into anterior and posterior approaches. The percentage of cases using BMP was plotted across time. A linear regression was fit to the data from quarter 3 of 2008 (FDA advisory) through quarter 1 of 2011, and a separate regression was fit to the data from quarter 2 of 2011 (The Spine Journal warning) onward. The slopes of these regression lines were statistically compared to determine differences in trends. No funding was received to conduct this study, and no authors had any relevant conflicts of interest. ResultsA total of 4,167,079 patients in the NIS underwent spinal fusion between 2002 and 2013. We found a greater decrease in rhBMP-2 use after The Spine Journal warning compared with the FDA advisory for all fusion procedures (p=.006), primary fusions (p=.006), and revision fusions (p=.004). Lumbosacral procedures also experienced a larger decline in rhBMP-2 use after The Spine Journal article as compared with the FDA warning (p=.0008). This pattern was observed for both anterior and posterior lumbosacral fusions (p≤.0001 for both). Anterior cervical fusion was the only procedure that demonstrated a decline in rhBMP-2 use after the FDA advisory that was statistically greater than after The Spine Journal article (p=.02). ConclusionsWarnings sanctioned through the spine literature may have a greater influence on practice of the spine surgery community as compared with advisories issued by the FDA.Comprehensive guidelines regarding safe and effective use of rhBMP-2 must be established.