Abstract

Short-course regimens for multi-drug resistant tuberculosis (MDR-TB) are urgently needed. Limited data suggest that the new drug, bedaquiline (BDQ), may have the potential to shorten MDR-TB treatment to less than six months when used in conjunction with standard anti-TB drugs. However, the feasibility of BDQ in shortening MDR-TB treatment duration remains to be established. Mathematical modelling provides a platform to investigate different treatment regimens and predict their efficacy. We developed a mathematical model to capture the immune response to TB inside a human host environment. This model was then combined with a pharmacokinetic-pharmacodynamic model to simulate various short-course BDQ-containing regimens. Our modelling suggests that BDQ could reduce MDR-TB treatment duration to just 18 weeks (four months) while still maintaining a very high treatment success rate (100% for daily BDQ for two weeks, or 95% for daily BDQ for one week during the intensive phase). The estimated time to bacterial clearance of these regimens ranges from 27 to 33 days. Our findings provide the justification for empirical evaluation of short-course BDQ-containing regimens. If short-course BDQ-containing regimens are found to improve outcomes then we anticipate clear cost-savings and a subsequent improvement in the efficiency of national TB programs.