The HLA system in hematopoietic stem cell transplantation - Immune Biology of Allogeneic Hematopoietic Stem Cell Transplantation - page 27

for selecting DPB1-mismatched donors whose mismatch may not increase
risks [78] . The study furthermore demonstrates that the presence of permis-
sible DPB1 mismatches in HLA-A, C, B, DRB1 or DQB1-mismatched trans-
plants, may afford comparable outcomes with those observed after HLA
10/10-matched non-permissive DPB1 mismatches. These data collectively
suggest that when HLA 10/10-matched donors are not available, DPB1 typ-
ing and selection of TCE permissible donors may afford these patients simi-
lar favorable outcomes as a traditional HLA 10/10-matched transplant.

Towards a haplotypic view of allogenecity

HLA haplotypes represent a series of HLA genes and other loci that are
physically linked on the same chromosomal strand ( Figure 2.1 ). The HLA
region has served as a model for understanding patterns of LD [5,29,
79-87] . HLA haplotypes are characterized by conserved “blocks” of genes
and sequences of variable lengths [29] . This block-like structure distin-
guishes “ancestral” haplotypes that carry HLA-A, C, B, DRB1 and DQB1
alleles in strong positive LD with one another, and with key markers that
reside in between the classical HLA loci [29] .

28

Since the five classical HLA loci represent a fraction of the total gene content
of the MHC [3] , undetectable haplotype-linked variation could be respon-
sible for increased risks after HLA-matched unrelated donor transplanta-
tion. Early observations pointed to donor-recipient disparity for variation
that resides outside of the classical HLA loci [88,89] . In the DNA typing era,
even with HLA 10/10 allele matching of unrelated donors, patients are at
increased risk of GVHD and mortality. Since HLA-matched patients and
unrelated donors are not identical by descent, it is possible that patient
and donor HLA-A, B and DR haplotypes are not the same. Using a patient
and donor who are both HLA-A1,2/B7,8/DR2,3 matched as an example,
the HLA-A antigen could be linked to different HLA-B/DR antigens: A1/
B8/DR3, A2/B7/DR2, A2/B8/DR3 and A1/B7/DR2. To test the hypothesis
that HLA-A, B, DR-defined haplotypes can serve as markers for undetected
linked genes that confer clinical effects, a novel method for phasing HLA
alleles was developed to physically separate the two HLA haplotypes [90] .
Application of the phasing method to HLA 10/10 allele-matched unrelated
patients and donors has revealed that a subset of identical pairs encode
different HLA-A, B, DR haplotypes [91] . Patients transplanted from HLA-
matched but haplotype-mismatched donors had significantly increased
risk of severe acute GVHD. These observations suggest that there are unde-
tected differences contributed by different haplotypes.

What are the candidates within the MHC that could be responsible for
GVHD risk? Microsatellite markers have been instrumental in probing the
gene-dense MHC region for potential areas that harbour new transplanta-
tion determinants [92] . One of the earliest studies to define susceptibility
genes was conducted in a Japanese population using microsatellite mark-
ers to query the MHC [93] . Among patients who developed clinical acute
GVHD, variation within the TNF complex was associated with lower sur-
vival. The importance of TNF polymorphisms has recently been confirmed
in a study of Caucasian patients and transplant donors [31] . Microsatellites
were used to define TNF alleles in HLA 10/10-matched transplant pairs.