Death, Cardiac Event and Acute Liver Injury Associated with Bedaquiline or Delamanid in Patients with Multidrug Resistant Tuberculosis: A Nationwide Cohort Study

Background: Bedaquiline and delamanid are newly approved drugs for the treatment of multidrug resistant tuberculosis (MDT-TB), yet there is limited evidence on their safety beyond the trial setting. Bedaquiline carries a black box warning of increased risk of death compared to the placebo arm, and there is a need to establish the magnitudes of additive QT prolonging effect and hepatotoxicity of bedaquiline and delamanid.

Objectives: To assess risk of mortality and serious adverse events for bedaquiline- or delamanid-containing regimen, compared with the standard regimen.

Methods: We used the South Korea National Health Insurance System database to identify individuals aged 18 years or older diagnosed with pulmonary MDR-TB (ICD-10: A15-A16 and U84.3) and received at least 4 second-line anti-TB drugs (i.e., minimum number of drugs required for MDR-TB treatment) between 2016 and 2019. Bedaquiline and delamanid groups included those who initiated treatment with bedaquiline- or delamanid-containing regimen, respectively, and standard regimen group with fluoroquinolone and/or aminoglycoside-containing regimen. Follow-up began on the day after treatment initiation. Primary outcome was all-cause mortality at 12-month of follow-up, and secondary outcomes were long QT-related cardiac event and acute liver injury assessed until discontinuation of the index drugs or end of study period (12/31/2020). We estimated the propensity score-based inverse probability of treatment weighted (IPTW) hazard ratios (HR) with 95% confidence interval (CI) for bedaquiline and delamanid, compared with the standard regimen group, adjusted for the covariates assessed on or before MDR-TB diagnosis.

Results: Among 1,998 MDR-TB patients, 315 (15.8%) and 292 (14.6%) received bedaquiline and delamanid, respectively, and all baseline covariates were balanced after IPTW. Bedaquiline showed significantly lower risk of mortality at 12-month (IPTW-adjusted HR 0.70, 95% CI: 0.55-0.89), whereas delamanid did not reach statistical significance (0.91, 0.70-1.19). IPTW-adjusted HRs of long QT-related cardiac events and acute liver injury were 1.10 (0.68-1.77) and 1.58 (1.34-1.86) for bedaquiline and 1.97 (1.27-3.05) and 1.17 (0.98-1.39) for delamanid, respectively.

Conclusions: Contrary to the black box warning, bedaquiline was associated with a mortality risk reduction during the early treatment phase of MDR-TB in the real-world population. Association between the serious adverse events with bedaquiline and delamanid warrant further investigations by considering for other concomitantly administered anti-TB drugs.