Background: The clinical utility and cost-effectiveness of pre-thiopurine NUDT15 pharmacogenetic testing in European and admixed populations are unknown.
Aims: To report the prevalence, penetrance, expressivity, and pathogenicity of NUDT15 variant allele carriage and the diagnostic accuracy and cost-effectiveness of an inexpensive loop-mediated isothermal amplification (LAMP) test.
Methods: Retrospective case-matched cohort study of rates of severe myelosuppression in patients with and without loss-of-function NUDT15 variant allele(s) treated with a thiopurine.
Results: Overall, 1.3% of Europeans and 11.7% of South Asians carried a variant allele. Severe myelosuppression was associated with NUDT15 variant allele carriage (11.3% vs 0.95%; p < 0.001). Carriage of a single *3, *6 or *9 variant allele was associated with a shorter time to severe myelosuppression. Numbers needed to genotype to prevent myelosuppression in Europeans and South Asians were 786 and 23. Variant calling using the LAMP assay was fully concordant with Sanger sequencing (n = 154). It improved the safety of thiopurine dosing and was cost-effective when used to reduce the frequency and cost of thiopurine blood monitoring for patients without risk variants.
Conclusion: We recommend TPMT and NUDT15 genetic testing in patients of Asian and admixed ancestry. In Europeans, reflex NUDT15 testing should be considered in patients with reduced TPMT activity or loss-of-function genotype. Thiopurines should be avoided in patients with > 1 NUDT15 variant allele and in patients with both NUDT15 and TPMT variant alleles. In patients with a single NUDT15 variant allele, we recommend thiopurine dose reduction (< 1 mg/kg/day) and intensified blood test monitoring.
Keywords: NUDT15; cost‐effectiveness; expressivity; inflammatory bowel disease; myelosuppression; penetrance; pharmacogenetic testing; variant pathogenicity.
© 2025 The Author(s). Alimentary Pharmacology & Therapeutics published by John Wiley & Sons Ltd.