PWE-234 Optimising the response to thiopurine therapy: a search for novel explanations for thiopurine hypermethylation

Abstract

Introduction Thiopurines are not effective in up to 1/3 of patients with inflammatory bowel disease (IBD) and 1/5 have to discontinue therapy due to side effects. An important cause of these problems is thiopurine hypermethylation. This is a catabolic process leading to an unfavourable thiopurine metabolite profile (high methyl-mercaptopurine (MeMP) to low thioguanine nucleotide (TGN) ratio; >11:1), which cannot be predicted by measurement of thiopurine-S-methyltransferase (TPMT) activity. Importantly thiopurine hypermethylation can be circumvented with the use allopurinol in combination with a low dose thiopurine. The aim of this study is to establish the mechanism of thiopurine hypermethylation and identify predictive genetic markers to allow early combination therapy. We hypothesised that thiopurine hypermethylation occurs as a result of genetic factors that affect methylation flux and the cellular transport of methylated metabolites. Methods 168 age and dose-matched patients prescribed AZA/6-MP were identified. Genomic DNA was extracted from EDTA blood samples of 76 patients demonstrating thiopurine hypermethylation and 92 patients with normal methylation profiles. Polymorphic sequence variants in genes predicted to affect thiopurine methylation flux and cellular metabolite transport were identified from single nucleotide polymorphism (SNP) databases and genotyped by Taqman assay. Associations were tested using Fisher's Exact test. Results We found a significant association between the haplotype of rs9332377 T and rs4646316 C, which encodes a low-activity synonymous Catechol-O-methyltransferase (COMT) variant, and protection from thiopurine hypermethylation (rs9332377 T, p=0.0178, rs4646316 C, p=0.03). A polymorphism in the nucleo-base transporter, ABCB5, was significantly associated with thiopurine hypermethylation (rs2031641 G/G, p=0.0098). The association was strengthened when patients with MeMP levels >5000 pmol/l vs MeMP <5000 pmol/l were compared (p=0.0065). Conclusion Changes in methylation flux due to the activity of methyltransferases other than TPMT affect the formation of thiopurine methylated metabolites, likely through direct competition for the essential co-factor S-adenosylmethionine. Furthermore, polymorphism in the ABCB5 gene, which affects the nucleotide-binding domain of this transporter, is associated with thiopurine hypermethylation, suggesting reduced cellular efflux of methylated metabolites. Further studies are now indicated to establish the role of these genetic markers in clinical practice. Competing interests None declared.