Development of pharmacogenomic algorithm to optimize nateglinide dose for the treatment of type 2 diabetes mellitus

Background Nateglinide is a meglitinide used for the treatment of type 2 diabetes mellitus. Individual studies demonstrated the association of CYP2C9, SLCO1B1, and MTNR1B variants with the safety and efficacy of nateglinide. The current study aimed to develop a pharmacogenomic algorithm to optimize nateglinide therapy. Methods Multiple linear regression (MLR) and classification and regression tree (CART) were used to develop a pharmacogenomic algorithm for nateglinide dosing based on the published nateglinide pharmacokinetic data on the area under the curve data (AUC) and C max ( n = 143). CYP2C9 metabolizer phenotype, SLCO1B1, MTNR1B genotypes, and CYP2C9 inhibitor usage were used as the input variables. The results and associations were further confirmed by meta-analysis and in silico studies. Results The MLR models of AUC and C max explain 87.4% and 59% variability in nateglinide pharmacokinetics. The Bland and Altman analysis of the nateglinide dose predicted by these two MLR models showed a bias of ± 26.28 mg/meal. The CART algorithm was proposed based on these findings. This model is further justified by the meta-analysis showing increased AUCs in CYP2C9 intermediate metabolizers and SLCOB1 TC and CC genotypes compared to the wild genotypes. The increased AUC in SLCO1B1 mutants is due to decreased binding affinity of nateglinide to the mutant affecting the influx of nateglinide into hepatocytes. MTNR1B rs10830963 G-allele-mediated poor response to nateglinide is attributed to increased transcriptional factor binding causing decreased insulin secretion. Conclusion CYP2C9, SLCO1B1, and MTNR1B genotyping help in optimizing nateglinide therapy based on this algorithm and ensuring safety and efficacy. Graphical abstract