Factor VIII Gene Variants and Inhibitor Risk in African American Hemophilia a Patients

African American hemophilia A (HA) patients experience a higher incidence of neutralizing antibodies (“inhibitors”) against factor VIII (FVIII) vis-à-vis Caucasian patients. Non-HA-causing non-synonymous single-nucleotide polymorphisms (ns-SNPs) in the F8 gene encoding FVIII-484H, FVIII-1241E and FVIII-2238V are more prevalent in African Americans than in Caucasians. Currently available recombinant FVIII products have FVIII-R484, FVIII-D1241 or E1241, and FVIII-M2238. Therefore a number of HA patients are infused with FVIII products (plasma-derived or recombinant) having amino acid sequences that differ from the sequence encoded by their hemophilic F8 gene. This study tested the hypothesis that immune responses to sequences encoded by these ns-SNPs are a significant risk factor for inhibitor development in the African American HA community.

As part of the Personalized Approaches to Therapies for Hemophilia (PATH) Study, blood samples were obtained from 175 African American and 198 Caucasian severe HA subjects, including several subjects with an intron-22 inversion mutation but moderate bleeding disorder. Questionnaires indicated which subjects had an inhibitor history, defined as ≥1 inhibitor titer measurements of >0.6 Bethesda units/ml. The F8 gene sequences at the three ns-SNPs and HLA-DRB1 genotypes were determined and PBMCs isolated from blood samples.

To determine which HLA-DRB1 proteins are capable of presenting the sequences encoded by these F8 ns-SNPs, synthetic 20-mer peptides with these polymorphic sequences were synthesized and quantitative peptide-HLA-DRB1 binding assays carried out. Binding to the HLA-DRB1 peptide-binding groove is a prerequisite for presentation of potentially immunogenic sequences on antigen-presenting cells. These assays identified several HLA-DRB1 that bound with moderate to high affinity to each of the polymorphic sequences, however none of the peptides showed promiscuous HLA-DRB1 binding. Most of the 11 HLA-DRB1 proteins tested, which together comprise ~40% of HLA-DRB1 alleles in the African American population, did not bind to the polymorphic FVIII sequences.

CD4 T cells from 1 (the only), 7 and 20 HA subjects who had been infused with FVIII products that were mismatched with sequences encoded by their endogenous F8 gene at FVIII-484, -1241 or -2238, respectively, and whose HLA-DRB1 proteins bound these sequences, were cultured with FVIII peptides containing both versions of these polymorphic sequences, in an attempt to identify HLA-restricted T-cell responses to “sequence-mismatched” FVIII. Positive control cultures consisted of parallel cultures of CD4 T cells with peptides corresponding to validated HLA-restricted epitopes in tetanus toxoid (TT) protein. Flow cytometry experiments then quantified the number of CD4 T cells that bound fluorescent HLA-DRB1 tetramers loaded with FVIII or TT peptides. Tetramer-positive cells were then sorted and cultured in attempts to isolate T-cell clones and polyclonal lines recognizing these sequences, as these would constitute unambiguous confirmation of T-cell responses to the peptide sequences. TT-specific clones were obtained routinely, as expected. The few cells positive for FVIII peptide-loaded tetramers did not produce T-cell clones or polyclonal T-cell lines, indicating the marginal tetramer-positive staining had been nonspecific.

Among the 373 HA subjects, inhibitor risks associated with the most common F8 ns-SNPs in African Americans, FVIII-E1241 and FVIII-V2238, were: E1241/D1241, OR=1.3(0.8-2.0, p=0.3); V2238/M2238, OR=1.1(0.5-2.4, p=0.84); 95%CI). Interestingly, African Americans with a F8 intron-22 inversion mutation showed a significantly higher inhibitor incidence compared to Caucasians with the same mutation, OR=2.9(2.4 – 5.8, p=0.004), but this did not correlate with any of the 3 ns-SNPs.

We conclude that immune responses to FVIII products mismatched with the FVIII sequences encoded by these three ns-SNPs are unlikely to contribute appreciably to the overall inhibitor incidence in African Americans with severe HA, but we emphasize that the at-risk population should be better defined through additional immunological investigations. The higher inhibitor incidence in African Americans with an intron-22 inversion mutation clearly warrants further investigation.

On behalf of the PATH Study Investigators (NIH 1RC2 HL101851)