Multicenter analyses demonstrate significant clinical effects of minor histocompatibility antigens on GvHD and GvL after HLA-matched related and unrelated hematopoietic stem cell transplantation.

The effect of minor H antigen mismatching on the occurrence of graft-versus-host disease (GvHD) and graft-versus-leukemia (GvL) after HLA-matched hematopoietic stem cell transplantation (HSCT) has mainly been demonstrated in single-center studies. Yet, the International Histocompatibility and Immunogenetics Workshops (IHIW) provide a collaborative platform to execute crucial large studies. In collaboration with 20 laboratories of the IHIW, the roles of 10 autosomal and 10 Y chromosome-encoded minor H antigens were investigated on GvHD and relapse incidence in 639 HLA-identical related donor (IRD) and 210 HLA-matched unrelated donor (MUD) HSCT recipients. Donor and recipient DNA samples were genotyped for the minor H antigens HA-1, HA-2, HA-3, HA-8, HB-1, ACC-1, ACC-2, SP110, PANE1, UGT2B17, and HY. The correlations with the primary outcomes GvHD (acute or chronic GvHD), survival, and relapse were statistically analyzed. The results of these multicenter analyses show that none of the HLA class I-restricted HY antigens were found to be associated with any of the primary outcomes. Interestingly, of the HLA class II-restricted HY antigens analyzed, HLA-DQ5 positive recipients showed a significantly increased GvHD-free survival in female-to-male HSCT compared with male-to-female HSCT (P = .013). Yet, analysis of the overall gender effect, thus independent of the known HY antigens, between the gender groups demonstrated an increased GvHD incidence in the female-to-male transplantations (P < .005) and a decreased GvHD-free survival in the female-to-male transplantations (P < .001). Of all autosomally encoded minor H antigens, only mismatching for the broadly expressed minor H antigen HA-8 increased the GvHD incidence in IRD HSCT (Hazard ratio [HR] = 5.28, P < .005), but not in MUD HSCT. Most striking was the influence of hematopoietic restricted minor H antigens on GvL as mismatching for hematopoietic minor H antigens correlated with lower relapse rates (P = .078), higher relapse-free survival (P = .029), and higher overall survival (P = .032) in recipients with GvHD, but not in those without GvHD. In conclusion, the significant GvHD effect of the broadly expressed minor H antigen HA-8 favors matching for HA-8 in IRD, but not in MUD, patient/donor pairs. The GvHD-GvL association demonstrating a significant lower relapse in hematopoietic minor H antigen mismatched patient/donor pairs underlines their clinical applicability for adoptive immunotherapy, enhancing the GvL effect in a GvHD controllable manner.

A male-specific increase in the HLA-DRB4 (DR53) frequency in high-risk and relapsed childhood ALL.

Previous studies reported significant HLA-DR associations with various leukemias one of which is with HLA-DRB4 (DR53) family in male patients with childhood ALL. We have HLA-DR-typed 212 high-risk or relapsed patients with childhood (n=114) and adult (n=98) ALL and a total of 250 healthy controls (118 children, 132 adult) by PCR-SSP analysis. The members of the HLA-DRB3 (DR52) family were underrepresented in patients most significantly for HLA-DRB1*12 (P=0.0007) and HLA-DRB1*13 (P=0.0001). In childhood ALL, the protective effect of DRB3 was evident in homozygous form (P=0.001). The DRB4 marker frequency was increased in males with childhood ALL (67.4%) compared to age- and sex-matched controls (42.1%, P=0.003) and female patients (35.7%, P=0.004). Besides being a general marker for increased susceptibility to childhood ALL in males, HLA-DRB4 is over-represented in high-risk patients. These results further suggest that the HLA system is one of the components of genetic susceptibility to leukemia but mainly in childhood and in boys only.

HLA system affects the age-at-onset in chronic myeloid leukemia.

Chronic myeloid leukemia (CML) occurs from childhood to old age. The adult form is characterized by the presence of Philadelphia chromosome resulting from bcr/abl translocation. The BCR-ABL fusion proteins are immunogenic, and the junctional sequences show unique HLA class I and class II restriction patterns in vitro. A previous study in the west of Scotland showed an influence of several HLA genotypes on the age-at-onset of CML. In the present study, we examined the HLA-A, -B, and -DRB1/3/4/5 allele and haplotype distributions in Turkish CML patients diagnosed in a single center where they are routinely HLA-typed by PCR-SSP analysis as a preparation for stem cell transplantation. The patients were 169 subjects of age 17-60 years. The older patients were not HLA typed and missing from the study group. The age-matched control group (n = 213) was healthy blood donors from the same geographical area. HLA-B*37 showed a risk association with CML [P = 0.02; odds ratio (OR) = 5.35]. The DRB1*10 association at similar magnitude was due to its linkage disequilibrium (LD) with B*37. HLA-B*35 and DRB1*11 showed independent protective effects (P = 0.007 and 0.017; OR = 0.54 and 0.60, respectively). The protective association of DRB1*11 may be due to its involvement in the presentation of the common (b3a2) fusion gene. HLA-B*14 and DRB1*01 showed strong LD, and all 5 patients who were positive for the presumed haplotype B*14-DRB1*01 were of age 43 years old or older (P = 0.003), suggesting a delay effect. We also examined the influence of homozygosity for DRB3 (DR52) and DRB4 (DR53) haplotypes on susceptibility. As previously shown in CML and CLL, DRB4 homozygosity was a risk marker (P = 0.01; OR = 3.36), and DRB3 homozygosity was protective (P = 0.007; OR = 0.51). Despite the lack of elderly patients in the study group, the opposite accelerating (DRB4) and delaying (DRB3) effects of homozygous genotypes on the age-at-onset were evident. Besides replicating previously found associations in a different population, this study also suggested new, and probably population-specific associations in CML. The mechanisms by which the HLA system modifies susceptibility to CML are unknown, likely to include immune and nonimmune ones, and worthy of further studies.