Study design: Prospective, longitudinal cohort study.
Objective: To examine the incidence of bone morphogenetic protein (BMP)-2 antibody formation in lumbar spine applications and to determine the clinical significance of an antibody response.
Summary of background data: Immune responses can affect the safety and efficacy profile of recombinant proteins. Type, incidence, and time course of antibody formation were evaluated in clinical studies investigating recombinant human bone morphogenetic protein (rhBMP)-2 in spinal arthrodesis.
Methods: Analysis of antibody formation to BMP-2, bovine collagen, and human collagen was performed after three prospective clinical studies investigating rhBMP-2 in single-level lumbar spinal arthrodesis. Two studies investigated rhBMP-2 applied to an absorbable collagen sponge at 1.5 mg/cm3 in lumbar interbody fusion (n=449); one study investigated rhBMP-2 applied to a ceramic and collagen compression-resistant matrix at 2.0 mg/cm3 in instrumented posterolateral fusion (n=239). Control patients received iliac crest bone graft (n=360). Two validated enzyme-linked immunosorbent assays were used to test for BMP-2 antibodies. Neutralizing antibodies were assessed using a cell bioassay. The incidence of antibodies to bovine and human collagen was determined. Radiographic and clinical outcome data were assessed to determine whether antibodies were correlated to patient outcomes.
Results: BMP-2 antibody rates ranged from 0.8% to 6.4% in rhBMP-2 patients and from 0% to 2.3% in control patients. Formation of BMP-2 antibodies peaked within the first 3 months and returned toward baseline values by 12 months. No neutralizing antibodies were detected. Bovine collagen antibody rates ranged from 12.7% to 18.8% in the rhBMP-2 patients and from 12.9% to 21.2% in the control patients. No antibodies to human collagen were detected. Adverse event rates were similar in antibody-positive and antibody-negative patients. BMP-2 antibodies did not affect bridging bone rates.
Conclusion: Formation of anti-BMP-2 antibodies was low and transient. No neutralizing antibodies were observed. Formation of antibodies did not affect fusion success or appear to have clinical sequelae.