Beatriz Fernández-Alarcón ¹ Oscar Nolberger ¹ Anna Vidal-Alabró ² Raul Rigo-Bonnin ³ Josep M Grinyó ⁴ Edoardo Melilli ² Nuria Montero ² Anna Manonelles ² Ana Coloma ² Àlex Favà ² Sergi Codina ² Josep M. Cruzado ² Helena Colom ^1* Nuria Lloberas ^2*

• ¹ Department of Pharmacy and Pharmaceutical Technology and Physical Chemistry, Faculty of Pharmacy and Food Sciences, University of Barcelona, Barcelona, Balearic Islands, Spain
• ² Department of Nephrology, Bellvitge University Hospital, Barcelona, Balearic Islands, Spain
• ³ Department of Biochemistry, Bellvitge University Hospital, Barcelona, Balearic Islands, Spain
• ⁴ Department of Clinical Sciences, Faculty of Medicine and Health Sciences, University of Barcelona, Barcelona, Catalonia, Spain

Aims: The once-daily extended-release tacrolimus formulation (ER-Tac) has demonstrated similar efficacy and safety to the twice- daily immediate release formulation (IR-Tac) but few population-based pharmacokinetic models have been developed in de novo kidney transplant patients to optimize doses. Therefore, this study aimed (i) at developing a population pharmacokinetic model for ER-Tac in de novo adult kidney transplant patients (ii) and at identifying genetic factors and time-varying covariates predictive of pharmacokinetic variability so as to guide tacrolimus dosage during the early post-transplant period. Methods: A total of 1067 blood tacrolimus concentrations from 138 kidney transplant patients were analyzed. 29 out of 138 were intensively sampled for 24 h on the fifth day post-transplantation, meanwhile for the remaining patients concentrations were collected on days 5, 10 and 15 after transplantation. Tacrolimus daily doses, genetic and demographic characteristics were retrieved from the medical files. Biochemistry time-varying covariates were obtained on different days over the pharmacokinetic (PK) study. A simultaneous PK analysis of all concentrations was carried out using the non-linear mixed effects approach with NONMEM 7.5. Results: A 2-compartment model with linear elimination and delayed absorption best described the tacrolimus pharmacokinetics. Between-patient variability was associated with oral blood clearance (CL/F) and the central compartment distribution volume (Vc/F). Tacrolimus concentrations standardized to a hematocrit value of 45% significantly improved the model (p<0.001). This method outperformed the standard covariate modelling of the haematocrit-blood clearance relationship. The effect of CYP3A5 genotype was statistically (p<0.001) and clinically significant, on CL/F.The CL/F of patients CYP3A5*1 carriers was a 51% higher than that of CYP3A5*1 non-carriers. Age also influenced CL/F variability (p<0.001). Specifically, CL/F declined 0.0562 units per each increased year from the value estimated in patients of 60 years and younger. Conclusion: The 36% of between-patient variability in CL/F was explained by CYP3A5 genotype, age and haematocrit. Haematocrit standardization to 45% explained variability of tacrolimus whole blood concentrations and this was of utmost importance in order to better interpret whole blood tacrolimus concentrations during therapeutic drug monitoring. Dose requirements of CYP3A5*/1 carriers, 60 years old and younger were higher than those of CYP3A5*/1 non-carriers older than 60 years.