# Empagliflozin in paediatric heart failure: model-based optimisation of a pharmacokinetic bridging study

## TL;DR

This paper proposes a model-based approach to optimize a clinical trial for empagliflozin in children with heart failure, aiming to determine the appropriate dosage and improve trial success.

## Contribution

The study introduces a model-based design to optimize a pharmacokinetic bridging trial for empagliflozin in pediatric heart failure patients.

## Key findings

- A two-compartment pharmacokinetic model was identified for empagliflozin with parameters varying allometrically with body weight.
- A median AUC ratio of 1.03 was achieved in children ≥15 kg using 10 mg tablets, comparable to adult exposure.
- An optimized sampling scheme for 12 patients was selected to characterize exposure and collect safety and efficacy data.

## Abstract

Current therapy for paediatric heart failure is still unsatisfactory, and trials in this population have often failed. Here we apply a model-based approach to optimise the study design and increase the probability of success of a prospective trial aimed at establishing the dose rationale for empagliflozin in children with heart failure. The proposed prospective protocol is based on the assumption that the cardioprotective mechanisms and efficacy of SGLT-2 inhibitors in the paediatric population can be extrapolated from adults.

A nonlinear mixed effects modelling approach incorporating prior information from pharmacokinetics (PK) in adults was used to extrapolate empagliflozin disposition parameters to children with ≥15 kg body weight. Protocol elements of interest were sampling schedule, dose, and sample size. These features were explored using an optimization algorithm ($DESIGN) and a simulation re-estimation procedure (SSE) in a large virtual paediatric cohort, overcoming some of the difficulties associated with low parameter precision in small populations.

A two-compartment pharmacokinetic model with sequential zero- and first-order absorption, absorption lag time and first-order elimination was identified. Clearance and distribution parameters were assumed to vary allometrically with body weight. We defined the lowest weight of 15 kg as inclusion criterion for the prospective trial, achieving, with the lowest commercially available tablet of 10 mg, a median AUC ratio of 1.03 (interquartile range 0.82–1.30) relative to the systemic exposure observed in a 50 kg adult receiving the 25 mg dose (median 7,163, IQR 6115–8,338 nmol*h/L). An optimised sampling scheme for a study with 12 patients was selected, which includes a sampling matrix with 4 different groups. Such a design enables the characterisation of empagliflozin exposure, along with exploratory safety and efficacy data collection in the population of interest.

Repurposing of drugs for paediatric rare diseases is fraught with challenges. Our results indicate that a weight-banded regimen with commercially available tablets of 10 mg empagliflozin can be used in a prospective protocol including paediatric patients with body weight ≥15 kg. In addition, this study illustrates the importance of optimising evidence generation in paediatric clinical trials through model-based approaches, ensuring that available knowledge is used to maximise the information content and reduce patient burden.