The HLA system in hematopoietic stem cell transplantation - Immune Biology of Allogeneic Hematopoietic Stem Cell Transplantation - page 25

HLA-C has served as an important model for understanding differential
risks conferred by allele and antigen mismatches [40,43,44,52] . Since its
discovery as a classical transplantation antigen [54] , donor mismatching
for HLA-C has consistently been shown to be a risk factor after myeloabla-
tive, non-myeloablative, unrelated donor, cord blood, marrow and periph-
eral blood stem cell transplantation. In the era of growth factor-mobilized
peripheral blood stem cell transplantation, the potential effects of HLA
disparity on outcome have been reassessed [53] . In a large retrospective
analysis of HLA-A, C, B, DRB1-allele-matched transplants, “8/8” donor
matching was associated with improved 1-year survival compared to any
single mismatch (“7/8”). Clinical differences in risks associated with HLA-C
antigen versus HLA-C allele mismatches were observed. Specifically, donor
HLA-C antigen mismatching was associated with the worst outcomes com-
pared to HLA-A, B or DRB1 mismatching. Whereas HLA-C antigen mis-
matching was associated with increased mortality, lowered disease-free
survival and increased grades III-IV acute GVHD, HLA-C allele mismatches
did not increase risks. HLA-B allele or antigen mismatches were associated
with GVHD; there were no statistically significant associations of HLA-A,
DRB1 or DQB1 mismatches with transplant outcomes. These data suggest
that for peripheral blood stem cell recipients, criteria for the selection of
mismatched donors may not necessarily be the same as that for marrow
sources. Future studies are warranted to fully evaluate the clinical signifi-
cance of the stem cell source on the risks conferred by specific loci and reso-
lution of donor matching.

26

Additive effects of HLA disparity

A step-up of risks associated with increasing numbers of HLA mismatches
is in keeping with a biological model for alloreactivity. Such multi-locus
mismatch effects have been observed in graft failure, GVHD and mortal-
ity [40,42-44,47,52,55-57] . Not only is the total number of HLA mismatches
an important determinant of risks, but the specific combination of mis-
matched loci can also define relative risks [44] . Not surprisingly, HLA-C dis-
parity in the presence of mismatching at any other HLA locus (class I and/
or class II) is associated with significantly increased incidence of grades
II-IV acute GVHD [43] . In the Japanese experience, HLA-C allele mismatch-
ing was an independent risk factor for severe acute GVHD. Lower rates of
relapse among patients with GVHD were observed (the “graft-versus-leu-
kemia effect”); when HLA-C mismatching occurred together with HLA-A, B,
DR or DQ mismatching, survival was significantly lower.

New approaches for defining permissible HLA mismatches

The feasibility of identifying HLA-matched donors depends on the HLA
alleles, antigens and haplotypes of the patient, along with the size and
composition of donor registries [35,38,58-64] . However, since every patient
has a mismatched donor, intense efforts have been made to determine the
characteristics of HLA mismatches that do not increase post-transplant
risks (“permissible” or “tolerable” HLA mismatches). Among the earliest
studies, donor-recipient mismatching for residue 116 of HLA-B was signifi-
cantly associated with increased risk of acute GVHD and transplant-related
mortality [65] . In more recent analyses, donor-recipient mismatching for