Pre-treatment NUDT15 testing to inform thiopurine prescribing in patients with inflammatory bowel disease: an economic evaluation

Abstract Background Pre-treatment NUDT15 genotype is effective at identifying individuals likely to have severe thiopurine-induced myelosuppression (TIM). We evaluate the cost-effectiveness of pre-treatment NUDT15 pharmacogenetic testing to inform thiopurine prescribing for people with inflammatory bowel disease (IBD). Methods Three strategies were compared to current practice (TPMT enzyme activity testing, not TPMT genotyping) using a decision tree model with a one-year time-horizon: • TPMT enzyme activity and NUDT15 genotyping. • Genotyping of both TPMT and NUDT15 . • No testing, thiopurines are avoided, alternative treatments prescribed. Patients with IBD considered for thiopurine treatment were modelled from the UK National Health Service perspective. Primary health outcome is avoidance of severe TIM. Costs include testing, treatment for IBD and TIM. Analyses were conducted for the general population and sub-groups depending upon genetic ancestry. Parameter uncertainty is explored via deterministic sensitivity analyses. Results Compared to current practice, all strategies were associated with fewer severe TIM events. Avoidance of thiopurines was the most expensive strategy. In the general population, avoidance of thiopurines had a cost per event avoided of £227,610 compared to current practice. Adding NUDT15 genotyping to TPMT testing was associated with a cost per event avoided of £6,447 compared to current practice. Genotyping of NUDT15 and TPMT dominated current practice. Differences in cost-effectiveness estimates were seen between the two sub-populations. Depending on assumptions, NUDT15 testing could be considered less costly than current practice in both groups. Genotyping costs and specific clinical management decisions had the greatest impact on results. Conclusions Notwithstanding limitations in the model, NUDT15 genotyping could be cost-effective to inform thiopurine use in patients with IBD in the UK.