Comparable Results of HLA-Mismatched Vs HLA-Matched Unrelated Donors for Allogeneic Stem Cell Transplantation Using Pre-Transplant ATG

For patients lacking an HLA-identical sibling, more than ten million unrelated donor (UD) are available for donation of hematopoietic stem cells. Despite this increasing number of UD, up to 30 % of the patients will not find a completely-HLA-matched donor. HLAmismatched unrelated stem cell transplantation (SCT) is associated with an increased risk of acute and chronic graft-versus-host disease and a higher transplant-related mortality resulting in a lower survival after allogeneic SCT. We analyzed on outcome of allele- and antigen-mismatch SCT from UD using pre-transplantation anti-thymocyte globulin (ATGFresenius ®) at a median dose of 60 mg/kg as part of the conditioning regimen. Between 08/1996 and 12/2004, 369 patients received SCT from UD in our institution. In 268, complete molecular typing (4-digit) of HLA-A, -B, -C, and DRB1, and DQB1 was available for donor as well as for recipient. According to high-resolution HLA-typing, 110 of the patients were completely matched for 10 alleles, whereas 91 patients had at least one allele-mismatch (9/10), 67 patients were mismatched for 2–4 alleles (6–8/10). The median age of patients was 41 years (range, 1–68). Diagnoses were acute leukemia or MDS in 135, chronic myeloproliferative disease (CML or PMF) (n=53), lymphoid malignancy (CLL, NHL, Hodgkin’s disease, MM) (n=60), or others (n=20). Stem cell source was either bone marrow (n=103) or peripheral blood stem cells (n=165). The matched and the mismatched group did not differ significantly between age, sex, risk category, graft source, CMV-status, conditioning, or CD34+ transplanted cells. Two patients (1 %) experienced primary graft-failure and were transplanted from 10/10-matched donors while in none of the mismatch-transplanted patients primary graftfailure occurred. The incidence of grade II–IV acute GvHD was 33 % in the 10/10-matched group, 41 % in the 9/10-mismatch-group, and 40 % in the 6–8/10-mismatch-group (p=0.1). The overall rate of chronic GvHD was 42 % in the 10/10-matched group, 46 % in the 9/10-mismatch group, and 46 % in the 6–8/10-mismatch group (p=0.8). Non-relapse mortality (NRM) in the 10/10-matched group was 27 %, in the 9/10-mismatch group 31 %, and in the 6–8/10-mismatch group 32 % (p=0.2). In a multivariate analysis, only age as continuous variable (HR 1.023) (p