Abstract
AIMS: To characterize the pharmacokinetics (PK) and PK/pharmacodynamics (PD) regarding glycosylated haemoglobin (HbA(1c)) lowering using the paediatric data from DINAMO and to assess differences compared with adults. METHODS: Population PK and PK/PD models previously developed for empagliflozin in adults and adolescents were re-estimated in a Bayesian framework. The PK model included 223 observations from 74 patients receiving empagliflozin 10 mg and 25 mg and the PK/PD model used 394 observations from 103 patients receiving empagliflozin (n = 52) or placebo (n = 51). RESULTS: Empagliflozin PK was well described by a 2-compartment model with sequential zero-order and first-order absorption, with tested covariate effects of sex, age, race and estimated glomerular filtration rate on apparent clearance, and fixed allometric exponents on apparent clearance, apparent central volume of distribution, apparent intercompartmental clearance and apparent peripheral volume of distribution. Simulations of area under the curve at steady state (AUC(ss)) demonstrated that adult and paediatric subjects exhibit similar AUC(ss). The PK/PD data were adequately described by a turnover model with disease progression and AUC(ss) inhibiting the HbA(1c) synthesis through an inhibitory maximum effect relationship. Simulations showed that the placebo-adjusted HbA(1c) decrease at Week 26 in the paediatric population was larger than that in the adult population (-0.699 vs. -0.528%). CONCLUSION: A Bayesian estimation framework enabled the characterization of empagliflozin PK and PK/PD with a limited number of samples in paediatric patients aged 10-17 years. Overall, the results confirm 10 and 25 mg as the appropriate empagliflozin doses in paediatric patients aged 10-17 years with type 2 diabetes mellitus.