When TB Becomes Drug-Resistant: Tracking Bedaquiline's Threat

Bedaquiline (BDQ) has become a cornerstone antibiotic for treating multidrug- or rifampicin-resistant tuberculosis (MDR/RR-TB), but resistance to BDQ (BDQR) is an increasing worry. Resistance can appear in three ways: by spontaneous mutation, by being acquired during a patient’s treatment, or by the transmission of an already resistant strain. To untangle how much each route contributes to the overall burden of BDQR, Abel Kjaersgaard and colleagues developed a transmission threshold model. The team combined four sources of data: novel estimates of the probability that BDQR arises spontaneously; updated estimates of the probability that resistance is acquired during BDQ treatment; a systematic review of country-level BDQR prevalence in MDR/RR-TB (BR-MDR/RR-TB); and WHO MDR/RR-TB notification and treatment data. By comparing what would be expected from mutation and acquisition alone to what is actually observed in prevalence surveys, the model aims to identify when observed resistance is likely driven by transmission rather than only by on-treatment emergence or rare spontaneous events.

The transmission threshold model defines a threshold value as the ratio of observed BDQR in prevalence studies to the amount expected from spontaneous mutation and acquired resistance combined. Analyses were run across hypothetical BDQR prevalence and treatment coverage scenarios and on country-specific data. The researchers estimated spontaneous resistance among all MDR/RR-TB at 0.073% (95% CrI: 0.062-0.085%) and estimated acquired resistance among MDR/RR-TB patients receiving BDQ at 3.6% (95% CrI: 2.8-4.4%). BDQR prevalence among MDR/RR-TB varied between 0-20% across 18 countries from 2015 to 2024. Scenario modelling indicated that spontaneous plus acquired resistance alone could not explain BDQR prevalence above 1.9% in settings where ≤50% of MDR/RR-TB patients had received BDQ in the previous year—pointing to the likely role of transmission. In four countries—Brazil, China, Mozambique, and South Africa—the threshold values and their uncertainty were consistent with ongoing BR-MDR/RR-TB transmission since 2020, while other countries’ values were consistent with resistance mainly arising spontaneously or during treatment.

These findings matter because the mix of how resistance appears determines the right public health response. If most BDQR comes from spontaneous mutation or is acquired during treatment, efforts should focus on protecting BDQ through careful treatment practices, monitoring, and stewardship to prevent resistance from emerging. If a substantial share of BDQR is due to transmission of resistant strains, then measures that interrupt spread—such as infection control, contact tracing, and rapid detection of resistant cases—become priorities. The transmission threshold model offers a practical way to interpret prevalence studies and routine data to estimate the contribution of transmission in a given country and year. Determining country- and year-specific threshold values can therefore help health programs decide when to prioritize preventing resistance acquisition and when to invest in interrupting transmission to protect the effectiveness of BDQ for treating MDR/RR-TB.