Impact of High-Resolution Typing for HLA-a, -B, -C and -DRB1 on Single-Unit Cord Blood Transplantation in Pediatric Patients

BACKGROUND:

Current practice in choosing potential donors for umbilical cord blood transplantation (UCBT) involves matching at the antigen level for HLA-A and HLA-B by low-resolution (LR) typing and at the allele level for HLA-DRB1 by high-resolution (HR) typing.

OBJECTIVES:

To determine the significance of HR HLA matching on UCBT outcome in 62 eligible pediatric patients <=21 years-of-age.  We further explored the presence of anti-HLA antibodies (AHA) in our patient population.

DESIGN/METHOD:

Through IRB approved retrospective chart review, we identified 62 UCBT recipients (>=2.0X10^7TNC/kg) for malignant (53%) and non-malignant (47%) disorders.  We compared original HLA typing (LR HLA-A-B; HR HLA-DRB1) to HR HLA typing (HR HLA-A-B-C-DRB1).  We determined incidence of engraftment, relapse, acute and chronic graft-versus-host disease (GVHD), early transplant-related infections and mortality (from any cause and non-relapse/non-disease progression mortality) and analyzed each of these variables against subgroups of HR and original HLA matching using the Fisher exact test.  Lastly, we determined the presence of pre-UCBT AHA.

RESULTS:

Retrospective determination of HR HLA matching showed 16% of pairs were matched at all HR loci; 6% were mismatched at 1, 14% at 2, 29% at 3, 27% at 4, 3% at 5, and 2% at both 6 and 7 alleles.  Ten of 13 (77%) transplants originally matched at LR HLA-A-B and HR HLA-DRB1 remained matched at the allelic level (HR HLA-A-B-C-DRB1).  Comparison of 3 HR subgroups (<4/8 vs >=4/8, <5/8 vs >=5/8, <6/8 vs >=6/8 allelic matches) showed significant reduction in mortality from any cause and non-relapse/non-disease progression mortality in transplants matched at >=6/8 versus <6/8 alleles (p = 0.02 and 0.01, respectively).  Although decreased in incidence, there was no significant association for graft failure, GVHD or infections.  In subgroup comparison using original HLA typing (<3/6 vs >=3/6, <4/6 vs >=4/6, <5/6 vs >=5/6 matches), a significant reduction in mortality from any cause and non-relapse/non-disease progression mortality was seen in those matched at >=5/6 versus <5/6 loci (p=0.05 and 0.01, respectively).  Screening for class I and II AHA was available in 27 patients and was positive in 8/14 patients with primary graft failure and 3/13 who engrafted.  Results were not available for 30/43 engrafters and 5/19 non-engrafters.

CONCLUSION:

HR typing showed an increase in HLA recipient cord disparity in comparison to standard typing.  We validated the current standard for UCB selection as >=5/6 grafts were associated with significantly improved overall survival (see Figure 1), which was shown in our >=6/8 HR matches as well (see Figure 2).  The role of AHA in graft failure needs to be further tested.  In choosing donors for UCBT, we recommend choosing >=6/8 HR match especially when >=5/6 UCB grafts are available and if able, taking into account AHA.