Impact of ABO-Mismatch on Treatment Outcome and Transfusional Needs in Non-Myeloablative, Allogeneic, Peripheral Blood Stem Cell Transplantations

Background: Suitable stem cell donors are selected by HLA-typing, CMV status, gender and age. The ABO blood groups are not part of the initial search and selection process. Since ABO-antigens are inherited independently from HLA antigens, patient-donor mismatches may occur and are considered acceptable. However, with the increased use of unrelated donors, more ABO-mismatched transplants are expected to take place. The influence of ABO-mismatched transplants on clinical outcomes has been studied earlier, but is still under debate. Comparisons between studies are challenging because of the heterogeneity of patient populations, conditioning regimes, various stem cell sources and different outcome parameters. Only a few studies have evaluated transfusional needs after ABO mismatched stem cell transplantations.

Aim: To gain insight in the impact of ABO-(mis)matching on red blood cell (RBC) transfusion dependency in the first 90 days after transplant and on clinical outcome after 1 year follow-up.

Study design and methods: All patients treated from 2004 till 2014 with a first, allogeneic, peripheral blood derived stem cell transplantation after a reduced-intensity conditioning regimen, were included for analyses of treatment outcomes and transfusional needs.  

Results: In the population for treatment outcome analysis (n=233), 124 patients were ABO-identical, 31 were ABO-major mismatched and 68 were ABO-minor mismatched. Ten patients had a bidirectional ABO-mismatch and were excluded from analyses. Multivariate analysis showed no impact of ABO-mismatch on overall survival, acute graft-versus-host disease (GVHD) incidence, non-relapse related mortality (NRM) and GVHD-related mortality. The minor ABO-mismatch group had more relapse related mortality (RRM) after 1 year (26.5%) compared to the ABO-identical (12.1%, P=0.01) and the ABO-major mismatched group (6.4%, P=0.04). The hazard ratio (HR) for RRM in ABO-minor mismatch patients versus ABO-identical patients was 2.41 (95% confidence interval, 1.14 to 5.08, P=0.021).  In the population for transfusion analysis (n=204), 111 patients were ABO-identical, 27 were ABO-major mismatched and 57 were ABO-minor mismatched. Nine patients had a bidirectional ABO-mismatch and were excluded from analyses. No influence of ABO-mismatch on transfusional needs was seen in the first 30 days after transplant.  30-60 days after transplant, 14 (52%) major ABO-mismatched patients needed RBC transfusions with an overall mean of 2.19 RBC units compared to 24 (22%) patients with an overall mean of 0.68 in the ABO-identical group (P=0.001) and 16 patients (28%) with an overall mean of 1.04 in the ABO-minor mismatched group (P=0.028). More than five transfusions were given to 15% of the major ABO-mismatch group, compared to only 3% in the ABO-identical group (Relative risk (RR) 5.0, 95% CI 1.2-21.0, P=0.028) and 5% in the minor ABO-mismatch group (RR 2.8, 95% CI 0.7-11.7, P=0.155). 60-90 days after transplant, the ABO-major mismatch group was still significantly more transfusion dependent with 13 patients needing transfusions (48%) and an overall mean of 2.33  RBC units per patient, compared to 23 patients (21%) with a mean of 1.04 units in the ABO-identical group (P=0.002) and 9 patients (16%) with an mean of 0.58 in the ABO-minor mismatch group (P=0.002). In this period,19% of ABO-major mismatch patients received more than 5 transfusions, compared to 6% in the ABO-identical group (RR 2.9, 95% CI 1.0-8.5, P=0.048) and 2% in the ABO-minor mismatch group (RR 10.6, 95% CI 1.3-86.0, P=0.028).  Differences in transfusional needs remained significant regardless of pre-transplant hemoglobin levels. Pure red cell aplasia (PRCA), defined as transfusion dependency from day 30 until longer than day 90 after transplant, was seen in 10 patients with an ABO-major mismatch (37%).

Conclusion:  ABO-major mismatched patients needed significantly more RBC transfusions in the second and third month after transplant. More relapse related mortality was seen in the minor mismatch group. However, this is probably an artefact due to relatively small patient numbers.  Other measures for treatment  outcome showed no significant differences. Our results confirm that  ABO-matching is not mandatory in donor selection for stem cell transplantation. Still, whenever possible ABO major mismatching should be preferably avoided  to decrease transfusion burden.