OBJECTIVE: Susceptibility to autoimmune hepatitis (AIH) type 1 has been associated with DRB1*03, DRB1*04, and DRB3 alleles in European and North-American whites, with DRB1*04 in Japan, and with DRB1*04 and DRB1*13 in Latin America. Very few studies have been performed on AIH type 2. The aim of the present study was to evaluate the association of AIH types 1 and 2 with HLA-DR and DQ loci. METHODS: We performed HLA-DRB and -DQB1 typing by polymerase chain reaction amplification with sequence-specific primers (PCR-SSP) in 139 AIH patients. Most had AIH type 1 associated with circulating anti-smooth muscle antibody with F-actin specificity or antinuclear antibody. Twenty-eight patients presented AIH type 2 with anti-liver/kidney microsome type 1 or anti-liver cytosol type 1 antibodies. RESULTS: We observed a significant increase of DRB1*13 (70% vs 26% of controls, p < 0.00001) and DRB3 (93% vs 69% of controls, p < 0.00001) in AIH type 1 patients. Analysis of patients without DRB1*13 disclosed a secondary association with DRB1*03 (70% vs 30% of controls, p = 0.0001) and either the DRB1*13 or the DRB1*03 alleles were present in the majority of these patients (91% vs 48% of controls, p = 0.001). Comparison of DRB1*13- and DRB1*03-positive subjects revealed that the former alleles conferred susceptibility to younger patients with AIH type 1. DQB1 typing showed a significant increase in DQB1*06 (68% vs 41% of controls, p = 0.00007) in strong linkage disequilibrium with DRB1*13, and a decrease in DQB1*0301 (8% vs 47% of controls, p(c) = 0.0003). On the other hand, HLA typing of patients with AIH type 2 disclosed a significant increase in the DRB1*07 (68% vs 20% of controls, p(c) < 0.00014), DRB4 (79% vs 43% of controls, p(c) = 0.004), and DQB1*02 (86% vs 42%, p = 0.00002) alleles. After exclusion of DRB1*07, a secondary association with HLA-DRB1*03 was further observed in these patients (78% vs 30%, p = 0.007) and most of them had either DRB1*07 or DRB1*03 (93% vs 44% of controls, p(c) < 0.0001). CONCLUSIONS: Our data indicate that predisposition to AIH types 1 and 2 is associated, respectively, with the DRB1*13 or DRB1*03 and DRB1*07 or DRB1*03 alleles, and suggest that protection against type 1 disease may be conferred by DQB1*0301. In addition, the cluster of DRB1*13 in children with AIH type 1 also supports the concept that different HLA alleles might influence the onset of the disease.
OBJECTIVE: Susceptibility to autoimmune hepatitis (AIH) type 1 has been associated with DRB1*03, DRB1*04, and DRB3 alleles in European and North-American whites, with DRB1*04 in Japan, and with DRB1*04 and DRB1*13 in Latin America. Very few studies have been performed on AIH type 2. The aim of the present study was to evaluate the association of AIH types 1 and 2 with HLA-DR and DQ loci. METHODS: We performed HLA-DRB and -DQB1 typing by polymerase chain reaction amplification with sequence-specific primers (PCR-SSP) in 139 AIH patients. Most had AIH type 1 associated with circulating anti-smooth muscle antibody with F-actin specificity or antinuclear antibody. Twenty-eight patients presented AIH type 2 with anti-liver/kidney microsome type 1 or anti-liver cytosol type 1 antibodies. RESULTS: We observed a significant increase of DRB1*13 (70% vs 26% of controls, p < 0.00001) and DRB3 (93% vs 69% of controls, p < 0.00001) in AIH type 1 patients. Analysis of patients without DRB1*13 disclosed a secondary association with DRB1*03 (70% vs 30% of controls, p = 0.0001) and either the DRB1*13 or the DRB1*03 alleles were present in the majority of these patients (91% vs 48% of controls, p = 0.001). Comparison of DRB1*13- and DRB1*03-positive subjects revealed that the former alleles conferred susceptibility to younger patients with AIH type 1. DQB1 typing showed a significant increase in DQB1*06 (68% vs 41% of controls, p = 0.00007) in strong linkage disequilibrium with DRB1*13, and a decrease in DQB1*0301 (8% vs 47% of controls, p(c) = 0.0003). On the other hand, HLA typing of patients with AIH type 2 disclosed a significant increase in the DRB1*07 (68% vs 20% of controls, p(c) < 0.00014), DRB4 (79% vs 43% of controls, p(c) = 0.004), and DQB1*02 (86% vs 42%, p = 0.00002) alleles. After exclusion of DRB1*07, a secondary association with HLA-DRB1*03 was further observed in these patients (78% vs 30%, p = 0.007) and most of them had either DRB1*07 or DRB1*03 (93% vs 44% of controls, p(c) < 0.0001). CONCLUSIONS: Our data indicate that predisposition to AIH types 1 and 2 is associated, respectively, with the DRB1*13 or DRB1*03 and DRB1*07 or DRB1*03 alleles, and suggest that protection against type 1 disease may be conferred by DQB1*0301. In addition, the cluster of DRB1*13 in children with AIH type 1 also supports the concept that different HLA alleles might influence the onset of the disease.