Understanding the drug exposure–response relationship of bedaquiline to predict efficacy for novel dosing regimens in the treatment of multidrug‐resistant tuberculosis

AIMS: To externally validate an earlier characterized relationship between bedaquiline exposure and decline in bacterial load in a more difficult‐to‐treat patient population, and to explore the performances of alternative dosing regimens through simulations. METHODS: The bedaquiline exposure–response relationship was validated using time‐to‐positivity data from 233 newly diagnosed or treatment‐experienced patients with drug‐resistant tuberculosis from the C209 open‐label study. The significance of the exposure–response relationship on the bacterial clearance was compared to a constant drug effect model. Tuberculosis resistance type and the presence and duration of antituberculosis pre‐treatment were evaluated as additional covariates. Alternative dosing regimens were simulated for tuberculosis patients with different types of drug resistance. RESULTS: High bedaquiline concentrations were confirmed to be associated with faster bacterial load decline in patients, given that the exposure–effect relationship provided a significantly better fit than the constant drug effect (relative likelihood = 0.0003). The half‐life of bacterial clearance was identified to be 22% longer in patients with pre‐extensively drug‐resistant (pre‐XDR) tuberculosis (TB) and 86% longer in patients with extensively drug‐resistant (XDR) TB, compared to patients with multidrug‐resistant (MDR) TB. Achievement of the same treatment response for (pre‐)XDR TB patients as for MDR TB patients would be possible by adjusting the dose and dosing frequency. Furthermore, daily bedaquiline administration as in the ZeNix regimen, was predicted to be as effective as the approved regimen. CONCLUSION: The confirmed bedaquiline exposure–response relationship offers the possibility to predict efficacy under alternative dosing regimens, and provides a useful tool for potential treatment optimization.