Purpose     To quantitatively synthesize the available best evidence for general complications, heterotopic ossification (HO), retrograde ejaculation, cervical swelling, and cancer rates with the use of rhBMP-2 in lumbar and cervical spine fusion.

Methods     We conducted an online search for relevant controlled trials and extracted data on the abovementioned endpoints. Studies were eligible for inclusion if they reported on spinal fusion with rhBMP-2 in humans. Publication bias and heterogeneity were assessed mathematically. These data were synthesized in a meta-analysis using DerSimonian–Laird random effects modeling to calculate pooled odds ratios.

Results     We identified 26 studies reporting on a total of 184,324 patients (28,815 experimental, 155,509 controls) with a mean age of 51.1 ± 1.8 years. There was a significantly higher risk of general complications with rhBMP-2 compared to iliac crest bone graft (ICBG) with an odds ratio (OR) of 1.78 (95 %CI 1.20–2.63), (p = 0.004). The odds ratio for HO was 5.57 (95 %CI 1.90–16.36), (p = 0.002), for retrograde ejaculation 3.31 (95 %CI 1.20–9.09), (p = 0.020), and for cervical swelling 4.72 (95 %CI 1.42–15.67), (p = 0.011), all significantly higher in the rhBMP-2 group. The pooled odds ratio for new onset of tumor was 1.35 (95 %CI 0.93–1.96), which represents no statistically significant difference between the groups (p = 0.111).

Conclusion     rhBMP-2 is associated with a higher rate of general complications as well as retrograde ejaculation, HO, and cervical tissue swelling in spine fusion. There is a slightly increased risk of new onset of tumors, however, without statistical significance.