Genetics Are Not Likely to Offer Clinically Useful Predictions for Elevated Liver Enzyme Levels in Patients Using Low Dose Methotrexate

Objective: To examine genetic influence on the risk of elevations in liver function tests (AST and ALT) among patients using low-dose methotrexate (LD-MTX).

Methods: We examined data from the LD-MTX arm of a randomized double-blind placebo-controlled trial conducted among subjects without rheumatic disease. Genome wide association studies (GWAS) were performed in subjects of European ancestry to test the association between single nucleotide polymorphisms (SNPs) and the log transformed maximum values of AST, ALT, and dichotomized outcome with AST or ALT > 2 times upper limit of normal (ULN). The association between variants in MTX metabolism candidate genes and the outcomes was also tested. Furthermore, associations between a drug induced liver injury (DILI) weighted genetic risk score (wGRS) and the outcomes were tested, combining 10 SNPs and 11 classical HLA alleles associated with DILI.

Results: In genome-wide genetic analyses among 1,429 subjects of European ancestry who were randomized to receive LD-MTX, two SNPs reached genome wide significance for association with log transformed maximum ALT. We observed associations between established candidate genes in MTX pharmacogenetics and log transformed maximum AST and ALT, as well as in dichotomized outcome with AST or ALT > 2 x ULN. There was no association between DILI wGRS or candidate variants and AST, ALT, or DILI response.

Conclusions: Modest evidence was observed that common variants affected AST and ALT levels in subjects of European ancestry on LD-MTX, but this genetic effect is not useful as a clinical predictor of MTX toxicity.