A systematic review of aspects of NUDT15 pharmacogenomic variants and thiopurine-induced myelosuppression

Abstract

Abstract Objectives Evidence for NUDT15 pharmacogenomic variants and thiopurine-induced myelosuppression (TIM), consists predominantly of association data in Asian, mixed variant homozygote/heterozygote populations. We therefore sought evidence on; (i) NUDT15 genotype-guided thiopurine dosing. (ii) Association data for TIM in NUDT15 variant heterozygotes with inflammatory bowel disease. (iii) Association data for NUDT15 variants with TIM in Europeans. (iv) Health economic data for NUDT15 genotyping in inflammatory bowel disease. Methods A systematic review was conducted, consisting of database searches, screening against pre-defined inclusion/exclusion criteria, and assessment of risk of bias using study-specific appraisal tools. Key findings Titles/abstracts of 493 articles were screened, with 29 studies included. (i) Significant reductions in TIM with genotype-guided thiopurine dosing were reported by both trials and a cohort study. (ii) TIM rates were significantly higher in NUDT15*3 heterozygotes vs. wild type. Data were conflicting for rarer variants. (iii) Four of five studies reported an association with TIM for at least one or a combination of NUDT15 variants in Europeans (OR 9.5–38.2), but data were conflicting. (iv) Both health economic analyses found TPMT/NUDT15 genotyping cost-effective in Asian populations, but not when a European population was considered. Conclusion Limited data showed an association with TIM in NUDT15 variant heterozygotes and Europeans and the potential for genotype-guided dosing to reduce TIM. Studies were generally small, heterogenous, and of variable quality. The low prevalence of rarer NUDT15 variants/variants in Europeans likely contributed to contradictory findings. Further research on the clinical utility of genotyping in diverse populations will help inform future economic analyses.