Bone marrow transplantation compared with peripheral blood stem cell transplantation in treatment of patients with hematologic malignancies: A meta-analysis based on randomized controlled trials

Objective To compare the efficacy of peripheral blood stem cell transplantation (PBSCT) and bone marrow transplantation (BMT) in treatment of malignant hemopathy. Methods A total of 329 literatures published from 1999 to 2015 were obtained from PubMed, EMBASE, CENTRAL, etc. Two reviewers independently identified the eligible studies and assessed the methodology of included trials. Twelve RCTs were selected, including 921 patients treated with BMT, and 884 patients treated with PBSCT. Overall survival, disease-free survival, graft-versus-host disease, relapse rate, relapse mortality, non-relapse mortality were analyzed by RevMan 5.2 software. Results There was no significant difference in overall survival, relapse mortality and non-relapse mortality between two groups (44.95% vs 42.87%, RR: 1.05, 95% CI: 0.95-1.17, P=0.09; 15.08% vs 12.92%, RR: 1.18, 95% CI: 0.93-1.48, P=0.17; 22.34% vs 22.58%, RR: 0.99, 95% CI: 0.83-1.18, P=0.90). The overall survival of more than 3 years, the incidence of acute graft-versus-host disease and chronic graft-versus-host disease in BMT group were significantly lower than PBSCT group (42.38% vs 51.02%, RR: 0.83, 95% CI: 0.70-0.99, P=0.04; 42.20% vs 48.13%, RR: 0.87, 95% CI: 0.77-0.99, P=0.003; 46.95% vs 61.89%, RR: 0.52, 95% CI: 0.43-0.63, P< 0.000 01), however, the disease-free survival and rates of relapse in BMT group were significantly higher than PBSCT group (47.89% vs 29.82%, RR: 1.61, 95% CI: 1.23-2.11, P=0.0006; 18.90% vs 12.58%, RR: 1.49, 95% CI: 1.12-2.00, P=0.007. Conclusion The overall survival of more than 3 years and incidence of graft-versus-host disease in patients treated with BMT are lower than patients treated with PBSCT, but the disease-free survival and rates of relapse in patients treated with BMT are higher than patients treated with PBSCT. There is no significant difference in relapse mortality and non-relapse mortality between two methods.