Towards model-informed precision dosing of clofazimine, moxifloxacin, and terizidone/cycloserine in the treatment of drug-resistant tuberculosis: An external model evaluation study

Clofazimine, moxifloxacin and terizidone/cycloserine play an important role in the treatment of drug-resistant tuberculosis (DR-TB). Personalized therapy guided by model-informed precision dosing (MIPD) can be a powerful tool to improve treatment outcomes, minimize adverse effects and combat the emergence of resistance. To set up an MIPD workflow, a population pharmacokinetic model (popPK model) is required. In this study, an external evaluation of popPK models of the three aforementioned drugs was carried out, using pharmacokinetic data from a cohort of patients with DR-TB, in order to identify the model with the best predictive performance. The best performing models (Abdelwahab et al. for clofazimine, Chirehwa et al. for moxifloxacin and Mulubwa and Mugabo for terizidone/cycloserine) were selected to calculate the area under the concentration-time curve (AUC, total exposure). An interoccasion variability (IOV, variability across dosing occasions) of AUC was quantified (13.4%CV (clofazimine), 16.1%CV (moxifloxacin), 14.5%CV (cycloserine)) indicating that using samples from one dosing occasion for AUC calculations may be sufficient to guide potential dose adjustment. Various single sampling schemes to estimate AUC were evaluated, but a unified timepoint for all drugs could not be determined. Known pharmacodynamic targets (AUC 0-24h /MIC, or T>MIC) were attained in almost all patients and dosing occasions.