To determine whether recipient HLA phenotypes are correlated with an increa
sed or decreased risk of alloantibody sensitization in end-stage renal dise
ase (ESRD) candidates for first or repeat kidney transplantation; we analyz
ed 19 440 kidney allograft recipients consisting of 13 216 Caucasians and 6
224 non-Caucasians transplanted between 10/87 and 11/98 at South-Eastern Or
gan Procurement Foundation (SEOPF) member institutions. Relative risk value
s and 95% confidence limits were obtained using Wolfe's method. Logistic re
gression was used to adjust for covariates that influence sensitization, i.
e. ethnicity, gender, age, pregnancies, transfusions, primary/repeat transp
lant and living versus cadaver donor. Univariate analysis of the entire coh
ort indicated that nine HLA allelotypes (DR 1,4.7; 138,12,40; A1.2,11) were
associated with a significantly reduced risk of sensitization, and five al
lelotypes (B42,B53; A10, 19,36) were associated with an increased risk of P
RA responses. Corrected for the number of statistical comparisons, recipien
ts with DR 1, DR4, A1 or A2 were 15% less likely to be sensitized per allel
otype. Recipients with B42, B53 or A36 were at increased risk of preformed
antibodies. after correction of the P value, for an average of 38'% increas
ed risk per allelotype. In the multivariate analysis, HLA phenotypes identi
fied as independent risk factors associated with protection against sensiti
zation were DR 1,4,7; B12(44,45); and A1,2, with an average reduced risk of
9% per allelotype. The only independent susceptibility allelotype was A36
with an increased risk of -29%. The A10 (25,26,34,66) group reached borderl
ine significance. We also looked for HLA-DR.-B,-A combinations that could p
otentially represent protective or at risk haplotypes/genotypes. Stepwise l
ogistic regression identified five combinations associated with protection:
DR1-B35-A3; DR1-B35-A2; DR1-B44-A2; DR4-B44-A2; DR7-B57-A1 (RR range 0.83-
0.63) with 27% average reduced risk per combination. Phenotype combinations
associated with an increased risk of sensitization were: DR2-B44-A2; DR2-B
53-A2; DR3-B8-A1; DR3-B42-A30; DR6-B42-A30; DR11-B53-A30 (RR range 2.76-1.4
8) with an average increased risk of 70% per combination. This study provid
es strong evidence that HLA-linked genes influence the anti-HLA PRA respons
e. The magnitude of the altered PRA response risk in DR-B-A combinations wa
s approximately twice that of the allelotypes at individual loci. HLA-DR ge
nes seemed to contribute most of the altered risk. The correlations between
DR types and PRA responsiveness are consistent with the DR types previousl
y regarded as predictors of kidney graft survival. The magnitude of increas
ed PRA risk attributable to an allelotype or combination was approximately
twice that associated with a decreased risk. We conclude that some HLA clas
s II-linked genes modulate the PRA response in a clinically significant man
ner. This immune response gene (Ir) regulation probably operates through po
lymorphic HLA molecules in their physiologic roles of antigen processing an
d presentation to helper T cells.