HLA-C and HLA-DQB1 compatibility in unrelated cord blood transplants

Umbilical cord blood (UCB) cells have been definitively proved to be a source of hematopoietic stem cells with repopulating capacity when transplanted into pediatric hosts with neoplastic or non-neoplastic disease. Moreover, due to the immaturity of the UCB lymphoid compartment, these transplants are usually associated with a low incidence and severity of GvHD. This clinical observation and the immaturity of the UCB lymphoid compartment justify the acceptance of UCB units which differ from their recipient by 1 or 2 HLA antigens of the six HLA A, B and DRB1 antigens conventionally typed. Whether the number and type of HLA disparities affect clinical outcome of UCB transplants has not, however, been clearly demonstrated yet.

In the present study on 14 pediatric patients with high risk leukemia transplanted with UCB from unrelated donors, evaluation of HLA compatibility was extended to HLA-C and DQB1 genes and correlated to the engraftment rate and occurrence of GvHD. Conditioning regimen and GvHD prophylaxis were identical in all cases. HLA-A and B antigens were typed by serology, whereas DNA based methods were used to define HLA-C gene groups, and HLA-DRB1 and DQB1 alleles.

Conventional HLA-A, B and DRB1 typing demonstrated that 12 recipient/donor pairs differed at one HLA locus, while 2 pairs had 2 HLA disparities. The extended HLA-typing showed that only one out of the six pairs with a different HLA-A locus had additional mismatches at HLA-C and DQB1 loci, whereas all the remaining 8 pairs, which already differed at HLA-B and/or DRB1 loci after conventional typing, had additional HLA-C and/or DQB1 mismatches (p = 0.002). By contrast, engraftment rate and occurrence of GvHD did not significantly correlate with level of HLA-mismatches even after extended HLA-typing.

The present data show that additional mismatched HLA-C and/or DQB1 antigens are significantly more frequent in pairs which after conventional HLA-typing differed at HLA-B and/or DRB1 loci, than in those showing one HLA-A mismatch. This observation provides an additional criterion for selection of UCB donors with the closest HLA-match when more than one unit are available. We did not, however, observe any correlation between engraftment rate, occurrence of GvHD and degree of HLA disparities detected either by standard or extended typing. These data support the notion that certain HLA differences do not affect the clinical outcome of UCB transplants and indicate that the expensive and time consuming molecular typing of HLA-C and DQB1 loci might be avoided for UCB donor selection.