Drug Exposure Dictates Evolutionary Dynamics of Drug-Resistant Tuberculosis: A Longitudinal Cohort Study in Beijing, China

Background Despite global efforts to combat drug-resistant tuberculosis (DR-TB), key gaps remain in understanding how resistance amplifies during treatment and the role of collateral drug effects in this process. Resistance amplification complicates management and undermines outcomes. This study aimed to identify the risk factors and collateral drug effects contributing to resistance amplification. Methods We conducted a retrospective, longitudinal cohort study of 133 DR-TB patients with paired phenotypic drug susceptibility tests (pDSTs) at Beijing Chest Hospital in Beijing, China, from January 2018 to January 2023. Patients were classified into "progression" and "non-progression" groups based on acquisition of new drug resistance. Multivariable logistic regression identified risk factors, and a restrictive cohort analysis assessed collateral drug effects. Results The median number of resistances per isolate increased significantly from 7.0 (IQR 5-9) to 8.0 (IQR 6-10) ( P P =0.032), while pyrazinamide (PZA) use correlated with increased progression ( P =0.035), though neither remained significant in multivariable models. Importantly, collateral effect analysis revealed that PZA exposure was strongly associated with acquired rifabutin (Rfb) resistance (OR, 5.94; 95% CI, 1.90-18.53; P =0.001). Conclusion Resistance amplification during DR-TB treatment is frequent, with INH-Mono representing a high-risk transitional state. While BDQ may offer a protective effect, the strong collateral association between PZA and Rfb resistance highlights the need for careful regimen design. These findings underscore the importance of individualized, resistance-informed therapy to mitigate further resistance amplification in DR-TB patients.