Drug-induced idiosyncratic liver disease (DIILD) depends largely on host susceptibility factors. Small studies support the genetic influence of human leukocyte antigen (HLA) class II molecules on the predisposition to DIILD. We sought associations between HLA-DRB and -DQB alleles and DIILD considered collectively or according to the biochemical expression of liver damage. We studied a total of 140 patients with a definitive or probable diagnosis of DIILD, as assessed with the Council for International Organizations of Medical Sciences scale, with 635 volunteer bone marrow and blood donors serving as controls. HLA-DRB1* and -DQB1* genotyping was performed by hybridization with sequencespecific oligonucleotides after genomic amplification. The group with DIILD did not differ from control subjects with regard to the distribution of HLA-DRB and -DQB antigens. The frequencies of alleles DRB1*15 (35.4% vs. 18.6% of controls; P ‫؍‬ .002; odds ratio [OR] 2.31) and DQB1*06 (61.5% vs. 40.8%; P ‫؍‬ .001; OR 2.32) were significantly increased in patients with the cholestatic/mixed type of liver damage in comparison to healthy subjects. By contrast, frequencies of alleles DRB1*07 (16.9% vs. 35.4%; P ‫؍‬ .003; OR 0.37) and DQB1*02 (32.3% vs. 55.8%; P ‫؍‬ .0003; OR 0.39) were significantly decreased. In conclusion, there is no association between any specific HLA allele and the propensity to develop DIILD. However, the genetic influence associated with HLA class II alleles appears to play a role in the biochemical expression of liver injury in cholestatic/mixed hepatotoxicity and may explain why a given drug may cause different patterns of liver damage.

No abstract