Estelle Chasseloup 1, Bart Remmerie 2, Vincent Duval 3
1 Johnson & Johnson (, France), 2 Johnson & Johnson (, Belgium), 3 Johnson & Johnson (, Switzerland)
Background
Bedaquiline (BDQ), is used to treat pulmonary multidrug -resistant (MDR) tuberculosis (TB) as part of a multi-drug regimen.
BDQ received accelerated Food and Drug Administration (FDA) approval in 2012 and conditional European Medicines Agency (EMA) approval in 2014 for treatment of MDR-TB in adults at the BDQ dosage of 400 mg QD for 2 weeks, followed by 200 mg three times per week for 22 weeks.
Traditional/full approval was granted in 2024 by the FDA and the EMA based on the phase 3 trial – STREAM Stage2.
In parallel to the phase 3 trial in adults, a partly sequential age de-escalating pediatric trial was conducted with the following cohorts (Cohort 1: 12 to <18 years of age approved in 2019, Cohort 2: 5 to <12 years of age approved in 2020, Cohort 3: 2 to <5 years of age approved in 2025, Cohort 4: 0 to <2 years of age ongoing).
The stepwise age de-escalation strategy of BDQ for pediatrics was undertaken to establish the PK (and safety) in the four cohorts, such that the adult-equivalent pediatric dosages (exposure-matching) could be determined. To this end, initially a population PK model consisting of the adults popPK model was modified by adding the standard allometric scaling factors on clearances (CLs) and volumes (Vs). For Cohort 3 specifically, the model was also used to simplify the mg/kg dosing regimen from the clinical trial to a flatter dosage using time-varying simulations covering the weight-for-age changes expected over the 24-week clinical trial duration. While the approach was accepted by EMA, FDA - during Scientific Review - requested additional sensitivity analyses using a popPK model after implementing allometric scaling factors and re-estimating the popPK parameters.
Objectives
Conduct a sensitivity analysis using the pediatric popPK model, re-estimating the popPK parameters and investigating the impact on the proposed dosage for pediatric patients with MDR-TB, aged 2 to <5 years.
Method
Adult data: 5222 BDQ concentrations and 4717 M2 concentrations from 479 adult participants (111 HVs and 368 patients with TB) with weight range of 30 kg – 113 kg1.
Pediatric data 2120 BDQ and M2 concentrations from 45 pediatrics (15 for cohorts 1 to 3).
Adult BQDM2 model1,2: model developed previously for BDQ and its main metabolite M2, developed to support the STREAM Stage 2 submission, including two demographic covariates: a 71% increase in CL/F for Black participants and a 26% decrease in Vc/F for females.
Allometric pediatric BDQM2 model: Allometric scaling with fixed exponents (0.75 for CLs, 1 for Vs) was added to the Adult BDQM2 model to account for growth related body weight increase in pediatrics. PopPK parameters were not re-estimated.
Joint Allometric BDQM2 model: Based on FDA feedback, a sensitivity analysis was performed, re-estimating the popPK parameters based on the pooled adults and pediatrics data to investigate the impact of the fixed allometric scaling factors on the parameter estimates, using NONMEM 7.5.
The Allometric pediatric BDQM2 model was compared to the Joint Allometric BDQM2 model by computing the relative difference in parameter estimates, and the geometric mean of the AUCs (AUC168h at week 12 and week 24).
Results
The Pediatric allometric BDQM2 model and the Joint allometric BDQM2 model were both able to describe the pediatric C1 to C3 data (2 to <18 years old) appropriately.
The Joint allometric BDQM2 model resulted in a relative difference in the estimates of the popPK parameters of 14% for the apparent volume of the BDQ central compartment and <3% for the other parameters. Regarding the BSV and RUV parameters, the maximum relative difference was 10%.
The BDQ and M2 geometric mean of the AUCs obtained in Cohort 3 with the joint allometric BDQM2 model differed maximally of 3.9% from those obtained with the allometric pediatric BDQM2 model, confirming the results with the allometric BDQM2 model, and supporting the simplification of the mg/kg bedaquiline dosage in the clinical trial to the proposed flatter label dosage.
Conclusion
Re-estimating the popPK parameters after implementing standard allometric scaling factors on CLs and Vs based on the pooled adults and pediatrics data resulted in negligible differences in parameter estimates and individual predictions in pediatrics compared to the allometric pediatric BDQ popPK model without re-estimation.
The sensitivity analyses supported the flatter BDQ dosage.
References:
1 McLeay SC, Vis P, van Heeswijk RPG, and Green B. The population pharmacokinetics of bedaquiline (TMC207), a novel anti-tuberculosis drug. Antimicrob Agents Chemother, Jun 2014.
2 32 (2024) Abstr xxxxx [www.page-meeting.org/?abstract=xxxxx], title : Population PK Parent-Metabolite Model development for bedaquiline and M2
Reference: PAGE 34 (2026) Abstr 11864 [www.page-meeting.org/?abstract=11864]
Poster: Drug/Disease Modelling - Paediatrics