The Impact of Graft/Recipient ABO Compatibility on Outcomes after Umbilical Cord Blood Transplant for Non-Malignant Disease

Background:

Existing literature shows mixed conclusions regarding the impact of ABO incompatibility on outcomes following hematopoietic cell transplantation (HCT). As the future for umbilical cord blood (UCB) expansion technologies is bright, we assessed whether this typically-overlooked graft characteristic impacted various outcomes following UCB transplantation (UCBT) for non-malignant disorders (NMD).

Patients and Methods:

A prospectively maintained institutional BMT program database was queried for all patients undergoing first UCBT for NMD and their demographic, disease and transplant-related characteristics. For each transplant, UCB and recipient ABO compatibility was considered (1) matched, (2) major mismatched (recipient isoagglutinins against UCB antigen), (3) minor mismatched (UCB isoagglutinins against recipient antigen), or (4) bi-directional mismatched. In double UCBT, only the compatibility status of the dominant unit was considered.

The impact of ABO incompatibility was assessed on the following outcomes using Kaplan-Meier and cumulative incidence functions: graft failure, aGvHD (grade II - IV and III - IV), cGvHD, platelet recovery, and overall survival. Cumulative red blood cell transfusion exposure per patient between HCT days 0 and +100 (c-RBC) was obtained from laboratory records. Donor hematopoietic chimerism values on the myeloid fraction of peripheral blood were reviewed at day +100 and most recent time assessed. HLA typing was at the antigen level for HLA-A and -B and allele level at -DRB1; in double UCBT, the matching status of the dominant unit was considered.

Results:

Through December 2014, 270 patients have undergone first UCBT for various NMD (inherited metabolic disorder = 165; marrow failure = 71; other = 34). The mean age at UCBT was 7.9 years (± 9.6). Most patients had a maximal performance score before UCBT (Lansky/Karnofsky of 100, n = 162; 60%) and were male (n = 164; 61%). Most patients received a single UCB unit (n = 225; 83%) and the majority of grafts were unrelated (n = 253; 94%). The average total nucleated cell dose was 8.0 x 10⁷/kg (± 6). With respect to HLA-matching, 79 (29%) received matched, 133 (49%) single-mismatched and 57 (21%) two-mismatched grafts. Most patients received myeloablative conditioning (n = 209; 77%) and GvHD prophylaxis with cyclosporine and mycophenolate mofetil (n = 150; 56%). Negative CMV serostatus in both the graft and recipient was most common (n = 172; 64%).

ABO compatibility for the cohort was as follows: matched, n = 93 (34%); major mismatch, n = 72 (27%); minor mismatch, n = 80 (30%); bi-directional mismatch, n = 23 (9%). ABO compatibility status was indeterminate in 2 transplants. Among all 4 ABO compatibility groups, no significant difference was seen in age, gender, NMD diagnosis category, conditioning intensity, HLA-matching, cell dose, number of UCB units, transplant era, performance score, GvHD prophylaxis, or CMV serostatus. Related UCB unit transplants were more likely to be ABO matched.

Table 1 shows outcomes by ABO compatibility group with a median follow-up time of 4.5 years (range, 0.5 - 20). ABO compatibility status did not appear to impact any outcomes assessed, though a trend toward increased grade III - IV aGvHD was seen in recipients of major mismatched units.

Conclusion:

ABO compatibility status did not seem to impact outcomes following UCBT for our large NMD cohort. Further analysis of the cohort to exclude interference on some outcomes assessed will be needed for a more definitive conclusion.