Journal of Neurosurgery: Spine, Dec 2009, Vol. 11, No. 6, Pages 729-740

Osteoinductive bone graft substitutes for lumbar fusion: a systematic review

Rajender Agarwal, M.D., M.P.H.1, Kendal Williams, M.D., M.P.H.1,2,3, Craig A. Umscheid, M.D., M.S.C.E.1,2,3, and William C. Welch, M.D.4
1Center for Evidence-Based Practice, University of Pennsylvania Health System; 2Department of Medicine; 3Center for Clinical Epidemiology and Biostatistics; and 4Department of Neurosurgery, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
Abbreviations used in this paper: AGF = activated or autologous growth factor; AIBG = autologous iliac crest bone graft; ALIF = anterior lumbar interbody fusion; BMP = bone morphogenetic protein; DBM = demineralized bone matrix; DJD = degenerative joint disease; ODI = Oswestry Disability Index; PLF = posterolateral lumbar fusion; PLIF = posterior lumbar interbody fusion; RCT = randomized controlled trial; rhBMP = recombinant human BMP; RR = relative risk. ©1944-2009 by the American Association of Neurosurgeons

Object
Autograft and allograft, the standard approaches for lumbar fusion procedures, have important disadvantages. Bone graft substitutes such as recombinant human bone morphogenetic proteins (rhBMP-2 and rhBMP-7) have emerged as viable alternatives. The authors conducted a systematic review to compare the efficacy and safety of osteoinductive bone graft substitutes using autografts and allografts in lumbar fusion.

Methods
A search for prospective controlled trials was conducted on MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials databases. Data were extracted for key outcomes including radiographically demonstrated nonunion, Oswestry Disability Index, operating time, blood loss, and length of hospital stay. The quality of randomized controlled trials was assessed using the Jadad scale. Meta-analyses were performed when feasible, and heterogeneity was assessed using the Q statistic and the I2 statistic.

Results
Seventeen of 732 potential studies met the inclusion criteria, with 9 examining rhBMP-2, 3 examining rhBMP-7, 3 examining demineralized bone matrix, and 2 examining autologous growth factor. Recombinant human BMP-2 significantly decreased radiographic nonunion when compared with autologous iliac crest bone graft (AIBG) in a meta-analysis (relative risk 0.27, 95% CI 0.16–0.46). Stratification of meta-analyses by the type of surgical procedure performed yielded similar results. Funnel plots suggested publication bias. Trials of rhBMP-2 suggested reductions in the operating time and surgical blood loss, with less effect on the length of hospital stay. There was no difference in radiographic nonunion with the use of rhBMP-7 when compared with AIBG (relative risk 1.02, 95% CI 0.52–1.98). Neither rhBMP-2 nor rhBMP-7 demonstrated a significant improvement on the Oswestry Disability Index when compared with AIBG. The limited data on demineralized bone matrix and autologous growth factor showed no significant improvement in radiographic outcomes.

Conclusions
Recombinant human BMP-2 may be an effective alternative to AIBG in lumbar fusion. Data are limited for other bone graft substitutes.

bone morphogenetic protein; bone substitute; lumbar fusion; meta-analysis; systematic review