Effect of CYP3A4*22, CYP3A5*3, POR*28, and PPARA RS4253728 on Tacrolimus Exposure and Neurotoxicity in Kidney Transplant Recipients

Abstract

Tacrolimus (TAC) is an immunosuppressive drug, substrate of CYP3A4/5. Studies have identified candidate polymorphisms in the genes related to CYP3A activity. Neurotoxicity is one of the major adverse effects of TAC. The aim of this study was to investigate in renal transplant recipients the relationship between TAC neurotoxicity, its trough (C0) and maximum (Cmax) concentrations or area-under the curve (AUC), and the CYP3A4*22, CYP3A5*3, POR*28 and PPARA rs4253728 variants. Patients of the EPHEGREN cohort were included in this study. Neurotoxicity events were recorded at each visit using a validated questionnaire. C0 were routinely measured and some patients had AUC and Cmax Bayesian estimation for TAC dose adjustment (https://pharmaco.chu-limoges.fr). Association between SNPs and neurotoxicity was investigated using Cox models for repeated events. Association between exposure parameters and SNPs was performed using generalized estimating equation. 7491 TAC C0 from 284 patients followed-up for 4.7[0.0-150.6] months were available (median[min-max]=8.4[1-58] µg/L). Significant associations were observed between C0 and CYP3A5*1/*3 (intercept(expressors)=8.10±0.20 µg/L, βnon-expressors=+1.10±0.24 µg/L, p<0.0001), and PPAR rs4253728 (intercept(AA)=9.70±0.24 µg/L, βGA = -0.86±0.29 µg/L, p=0.0034, βGG = -0.78±0.29 µg/L, p=0.0081). 305 TAC Cmax and AUC each were available in 69 patients followed-up for 61.8[0-302.9] months: AUC0-24h=281[38-728] µg*h/L, Cmax=19.4[2.5-65.5] µg/L. Significant association was observed between Cmax and CYP3A5*1/*3 (intercept(expressors)=25.7±2.2 µg/L, βnon-expressors=-4.6±2.3 µg/L, p=0.045), and POR*28 (intercept(CC)=22.5±1.0µg/L, βCT=-2.0±1.7µg/L, p=0.235, βTT=2.9±1.1µg/L, p=0.012). The questionnaires were filled in by 288 patients followed up for 6[0-84] months, corresponding to 2215 formularies and 52 neurotoxicity events observed in 27 patients occurring at 9[0.0-84.0] months. No significant association was observed between SNPs and neurotoxicity. A higher Cmax was observed in CYP3A5 expressors compared to non-expressors, while no significant increase in neurotoxicity risk was observed. This may be explained by the actual absence of such a relation or a lack of statistical power.