Background: Recombinant human bone morphogenetic protein-2 (BMP-2) is Food and Drug Administration-approved for use in acute open tibial shaft fractures. Some surgeons, however, also use BMP-2 in an "off-label" application for other acute fractures and for nonunion care. This retrospective study was performed to assess radiographic outcomes of off-label uses of BMP-2 for acute fractures and nonunions at our institution.
Methods: All eligible off-label BMP-2 applications between 2004 and 2008 for acute fractures or nonunions were reviewed. Univariate and multivariate analyses were completed to identify patient and clinical factors that could predict radiographic success or failure of the procedure.
Results: One hundred sixteen of 145 BMP-2 applications in 104 of 128 patients met inclusion and exclusion criteria. The overall radiographic union rate was 66% (76 of 116). In the univariate analysis, five factors correlated with significantly higher union rate: volume of bone defect <4 cm3, >2 cortices in contact at the index procedure, male gender, body mass index <30, and history of closed fracture pattern. Within the multivariate analysis, factors independently predictive of radiographic union included open versus closed fracture, gender, and volume of bone defect.
Conclusion: Off-label use of BMP-2 in acute fractures and nonunions resulted in a 66% success rate. It remains uncertain whether there is any clinical advantage to this approach, but it appears that female gender, open injury, and higher volumes of bone defect may be important negative prognostic factors for obtaining radiographic union. Appropriately powered prospective randomized trials are needed for further clarification, especially in light of the high cost of this treatment.