P440 Prevalence of NUDT15 polymorphisms and incidence of thiopurine-induced leukopenia in Inflammatory Bowel Disease patients: A systematic review and meta-analysis

Abstract Background Nudix hydrolase 15 (NUDT15) genetic polymorphisms confer an increased risk of thiopurine-induced myelosuppression (TIM). Thiopurines continue to be widely used as maintenance therapy for inflammatory bowel disease (IBD) in patients worldwide. However, the global prevalence of NUDT15 variants and incidence of TIM in IBD remains unclear. We aimed to evaluate the global prevalence of NUDT15 polymorphisms in IBD patients and incidence of thiopurine-induced leukopenia in patients with these polymorphisms. Methods We searched Medline, Embase, Emcare, CINAHL, APA PsycInfo and the Cochrane Library databases from inception until 7 July 2022. Full-text articles and conference abstracts reporting the frequency of any NUDT15 polymorphism and/or frequency of leukopenia in adult IBD patients with these variants, were included. Studies that examined paediatric populations or mixed disease populations, or reported duplicate data, were excluded. A random effects model was performed to estimate the pooled prevalence of NUDT15 variants, pooled prevalence of early (≤8 weeks) and late (>8 weeks) leukopenia and relative risk (RR) of developing leukopenia. Results 20 studies comprising 5232 IBD patients were included. The pooled prevalence of the *1/*1 wild type genotype was 83% (95% CI: 0.78-0.88). The pooled prevalence of the *1/*3 c.415C>T C/T variant was 13% (95% CI: 0.10-0.18), *3/*3 c.415C>T T/T variant was 2% (95% CI: 0.01-0.02), *1/*5 c.52G>A G/A variant was 2% (95% CI: 0.01-0.03) and *1/*6 c.36_37insGGAGTC ins/- variant was 7% (95% CI: 0.04-0.12). Subgroup analyses showed a higher pooled prevalence of the *1/*3 c.415C>T C/T variant in Japanese (20%, 95% CI: 0.16-0.24) and Chinese patients (18%, 95% CI: 0.12-0.27). Patients with the *1/*3 c.415C>T C/T variant had a RR of 4.12 (95% CI: 2.87-5.91) of developing leukopenia relative to the *1/*1 genotype, and the pooled prevalence of early leukopenia was 20% (95% CI: 0.16-0.26) and late leukopenia was 36% (95% CI: 0.26-0.49). Patients with the *3/*3 c.415C>T T/T variant had a RR of 9.38 (95% CI: 5.17-17.01) of developing leukopenia relative to the *1/*1 genotype, and the pooled prevalence of early leukopenia was 99% (95% CI: 0.07-1.00). The pooled prevalence of early leukopenia was 49% (95% CI: 0.29-0.69) in patients with the *1/*6 c.36_37insGGAGTC ins/- variant. Conclusion NUDT15 polymorphisms are common in IBD patients and strongly predict thiopurine-induced leukopenia. The most common variant was the *1/*3 c.415C>T C/T variant and the *3/*3 c.415C>T T/T variant was associated with the highest risk of leukopenia. Pre-treatment genotyping for NUDT15 variants should be considered particularly in Asian populations, to guide individualised thiopurine dosing and avoid myelotoxicity.