Clinical and laboratory evidence support an immune pathogenesis in most cases of idiopathic aplastic anemia (AA) and closely related disorders such as paroxysmal nocturnal hemoglobinuria (PNH). While external triggers are likely necessary, a complex constellation of immunogenetic factors may determine disease susceptibility. Many immunogenetic factors can influence the quality of immune response and affect the propensity to immune-mediated attack on hematopoietic stem cells in AA. Here we investigated whether KIR and KIR-L (HLA-A) genotype and cytokine/receptor gene variants are over-represented in AA and PNH. We studied a cohort of 77 patients with AA (23 AA, 20 AA/PNH and 34 PNH), 10 with hypocellular MDS and 175 healthy controls. The following SNPs in immunoregulatory genes were analyzed: IL-1α (−889 T/C), IL-2 (−330 T/G +166 G/T), IL-4 (−1098 T/G −590 T/C −33 T/C), IL-1R (−1970 C/T), IL-1Rα (mspa111100 T/C), IL-4RA (+ 190 G/A), IL-1β (−511 C/T, +3962 T/C), IL-6 (−174 C/G, nt565 G/A), IL-10 (−1082 G/A, −819 C/T, −592 C/A), IL-12 (−1188 C/A), TGF-β (+10 C/T, +25 G/C), INF-γ (+874 A/T), TNF-α (−308 G/A, −238 G/A) and immunomodulatory receptor genes including CTLA-4 exon 6 (+49 G/A), FcRIIIa (158 F/V) and CD45-exons 6 (+138 A/G), and 4 (+54 A/G, +77 C/G). As binding of KIR to the appropriate HLA ligand (KIR-L) can modulate activation of NK and cytotoxic T cells, we examined the combined impact of KIR/KIR-L genotypes on the risk of AA and PNH syndrome. In AA we found a decreased frequency of inhibitory KIR-2DL3 genes (68% vs. 89%, p=.0002); analysis of the KIR genotype in correlation with the corresponding KIR-L profile, revealed a decreased frequency of stimulatory 2DS1/C2 mismatch resulting in a potentially enhanced cytotoxic activity (14% vs.44%, p=.003). No association was found for most of the SNPs tested. However, when we examined the frequency TGF-β genotypes, increased frequency of GG variant in codon 25 (61% vs. 35% in controls, p=.03), associated with the “high secretor” phenotype, was found in AA. This relationship was also present in hypocellular MDS (82% vs. 32%, p=.007). Additionally, we found a lower incidence of TT genotypes for the IL-1Rα gene (33% vs. 62% p=.02). We confirm that the hypersecretor genotype T/T of INF-γ was over-represented in AA (28% vs. 10% in controls, p=.02). Subgroup analysis revealed that the T/T genotype of IFN-γ (35% vs. 14% p=.01) correlated with presence of a PNH clone. Previously, we have shown the association of HLA-DR15 with responsiveness to immunosuppression. When AA patients were subgrouped according to response to ATG/CsA, therapy refractoriness correlated with the presence of the C2/C2 haplotype (30% vs. 0% p=.02) and inhibitory KIR-2DL3/C1 mismatch (70% vs. 0%, p=.01) which may result in a greater propensity to breach of self-tolerance. In comparison, in the total AA group, C2/C2 haplotype and KIR-2DL3/C1 mismatch were present in 17% vs. 24% and 8% vs. 16% of controls, respectively. An increase in the frequency of 2DL3 and a decrease in 2DS1 mismatch may result in imbalance between cytotoxicity and KIR inhibition. In sum, our findings demonstrate that complex inherited traits involving immunogenetic factors may genetically determine propensity to bone marrow failure syndromes.

Author notes

Disclosure: No relevant conflicts of interest to declare.