Genetic and clinical predictors of rifapentine and isoniazid pharmacokinetics in paediatrics with tuberculosis infection

Objectives Twelve weekly doses of rifapentine and isoniazid (3HP regimen) are recommended for TB preventive therapy in children with TB infection. However, they present with variability in the pharmacokinetic profiles. The current study aimed to develop a pharmacokinetic model of rifapentine and isoniazid in 12 children with TB infection using NONMEM.Methods Ninety plasma and 41 urine samples were collected at Week 4 of treatment. Drug concentrations were measured using a validated HPLC-UV method. MassARRAY (R) SNP genotyping was used to investigate genetic factors, including P-glycoprotein (ABCB1), solute carrier organic anion transporter B1 (SLCO1B1), arylacetamide deacetylase (AADAC) and N-acetyl transferase (NAT2). Clinically relevant covariates were also analysed.Results A two-compartment model for isoniazid and a one-compartment model for rifapentine with transit compartment absorption and first-order elimination were the best models for describing plasma and urine data. The estimated (relative standard error, RSE) of isoniazid non-renal clearance was 3.52 Lh-1 (23.1%), 2.91 Lh-1 (19.6%), and 2.58 Lh-1 (20.0%) in NAT2 rapid, intermediate and slow acetylators. A significant proportion of the unchanged isoniazid was cleared renally (2.7 Lh-1; 8.0%), while the unchanged rifapentine was cleared primarily through non-renal routes (0.681 Lh-1; 3.6%). Participants with the ABCB1 mutant allele had lower bioavailability of rifapentine, while food prolonged the mean transit time of isoniazid.Conclusions ABCB1 mutant allele carriers may require higher rifapentine doses; however, this must be confirmed in larger trials. Food did not affect overall exposure to isoniazid and only delayed absorption time.