A Polymorphism In Interferon Gamma Gene Impacts the Extent of Joint Damage In Patients with Severe Hemophilia

Abstract 546

The basis for 10–15% of patients with severe hemophilia having clinically mild disease is not fully understood. We have described two levels of heterogeneity that is clinically obvious – first in the number of bleeds and second is in the amount of joint damage with similar number of bleeds. We hypothesized that differences in inflammatory responses determine the latter phenomenon. A total 114 patients with hemophilia A (n=95) and hemophilia B (n=19) were studied. All these patients were on minimal on-demand treatment (50-150 IU/kg/year). Patients were evaluated for the frequency and site of hemorrhage. Their WFH clinical and radiological (Pettersson) joint scores were documented. Disease causing mutations in factor 8 or factor 9 genes was genotyped. Functional polymorphisms in the inflammatory cytokine genes (tumor necrosis factor alpha; transforming growth factor (TGF) beta; interleukin (IL) 10; IL 6; IL 1beta; IL 1 beta receptor antagonist; tumor necrosis factor beta), immunoregulatory cytokine genes (interferon gamma(IFNg); HLA B27; FC gamma receptor) were genotyped. Patients with extremes of joint damage marked by their clinical and radiological scores, were classified as having severe (n=10, upper 10 percentile, ≥21.8 clinical and ≥36.1 radiological scores) or mild (n=10, lower 10 percentile, ≤2.6 clinical and ≤2.9 radiological scores) disease. The hemostasis parameters, factor VIII/or IX levels and tests of global hemostasis as measured by clot curve analysis (MAX2) were comparable between these groups. The number of reported bleeds between the severe group and mild group of patients within the classification of clinical score (19.7 Vs 12 bleeds, p=0.08) or radiological scores (14.2 Vs 13.4 bleeds, p=0.544) was not significantly different. This reflects our clinical observations that even among those who bleed frequently, the degree of synovial reaction and the extent of damage to joint cartilage can vary. Of the polymorphisms studied, the IL-10 -819T allele (60 Vs 22%, p=0.009) and -592C (60 Vs 22%, p=0.009) allele predicting higher levels of this anti-inflammatory cytokine, and IFNg 874A allele (60 Vs 0%, p= 0.011) predicting lower levels of pro-inflammatory cytokine were associated with a mild phenotype in patients classified using radiological scores. Interestingly, the IFNg 874A allele (60 Vs 0%, p= 0.011) and the inheritance of moderate disease causing mutation (missense mutation, 88 Vs 28%, p=0.041) was also associated with mild phenotype by clinical score classification. This data suggests that a cytokine mileu in target joint comprising of lower levels of IFNg that results in reduced co-stimulation of pro-inflammatory cytokines and higher levels of anti-inflammatory IL-10 may reduce the risk of developing severe arthropathy. This phenomenon needs to be evaluated in larger number of patients receiving on-demand treatment.