Acute and chronic graft-versus-host disease after allogeneic peripheral-blood stem-cell and bone marrow transplantation: A meta-analysis

Purpose: Controversy exists as to whether the incidence of graft-versus-hos t disease (GVHD) is increased after peripheral-blood stem-cell transplantat ion (PBSCT) when compared with bone marrow transplantation (BMT). We perfor med a meta-analysis of all trials comparing the incidence of acute and chro nic GVHD after PBSCT and BMT reported as of June, 2000. Secondary analyses examined relapse rates after the two procedures. Methods: An extensive MEDLINE search of the literature was undertaken. Prim ary authors were contacted for clarification and completion of missing info rmation. A review of cited references was also undertaken. Sixteen studies (five randomized controlled trials and 11 cohort studies) were included in this analysis. Data was extracted by two pairs of reviewers and analyzed fo r the outcomes of interest. Meta-analyses, regression analyses, and assessm ents of publication bias were performed. Results: Using a random effects model, the pooled relative risk (RR) for ac ute GVHD after PBSCT was 1.16 (95% confidence interval [CI], 1.04 to 1.28; P = .006) when compared with traditional BMT. The pooled RR for chronic GVH D after PBSCT was 1.53 (95% CI, 1.25 to 1.88; P < .001) when compared with BMT. The RR of developing clinically extensive chronic GVHD was 1.66 (95% C I, 1.35 to 2.05; P < .001). The excess risk of chronic GVHD was explained b y differences in the T-cell dose delivered with the graft in a meta-regress ion model that did not reach statistical significance. There was a trend to wards a decrease in the rate of relapse after PBSCT (RR = 0.81; 95% CI, 0.6 2 to 1.05). Conclusion: Both acute and chronic GVHD are more common after PBSCT than BM T, and this may be associated with lower rates of malignant relapse. The ma gnitude of the transfused T-cell load may explain the differences in chroni c GVHD risk. J Clin Oncol 19:3685-3697. (C) 2001 by American Society of Clinical Oncolog y.